Top Read Articles

Published in last 1 year |

In last 2 years |

In last 3 years |

All

Please wait a minute...


For Selected: Download Citations EndNote Ris BibTeX Toggle Thumbnails

Select

Multi-Robot Task Allocation Using Multimodal Multi-Objective Evolutionary Algorithm Based on Deep Reinforcement Learning MIAO Zhenhua(苗镇华), HUANG Wentao(黄文焘), ZHANG Yilian(张依恋), FAN Qinqin(范勤勤) J Shanghai Jiaotong Univ Sci    2024, 29 (3): 377-387.   DOI: 10.1007/s12204-023-2679-7 Abstract （584）      PDF（pc） （975KB）（229）       Knowledge map    Save The overall performance of multi-robot collaborative systems is significantly affected by the multirobot task allocation. To improve the effectiveness, robustness, and safety of multi-robot collaborative systems,a multimodal multi-objective evolutionary algorithm based on deep reinforcement learning is proposed in this paper. The improved multimodal multi-objective evolutionary algorithm is used to solve multi-robot task allocation problems. Moreover, a deep reinforcement learning strategy is used in the last generation to provide a high-quality path for each assigned robot via an end-to-end manner. Comparisons with three popular multimodal multi-objective evolutionary algorithms on three different scenarios of multi-robot task allocation problems are carried out to verify the performance of the proposed algorithm. The experimental test results show that the proposed algorithm can generate sufficient equivalent schemes to improve the availability and robustness of multirobot collaborative systems in uncertain environments, and also produce the best scheme to improve the overall task execution efficiency of multi-robot collaborative systems. Reference | Related Articles | Metrics | Comments（0）

Select

Multi-Agent Path Planning Method Based on Improved Deep Q-Network in Dynamic Environments LI Shuyi (李舒逸), LI Minzhe (李旻哲), JING Zhongliang∗ (敬忠良) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 601-612.   DOI: 10.1007/s12204-024-2732-1 Abstract （580）      PDF（pc） （1213KB）（245）       Knowledge map    Save The multi-agent path planning problem presents significant challenges in dynamic environments, primarily due to the ever-changing positions of obstacles and the complex interactions between agents’ actions. These factors contribute to a tendency for the solution to converge slowly, and in some cases, diverge altogether. In addressing this issue, this paper introduces a novel approach utilizing a double dueling deep Q-network (D3QN), tailored for dynamic multi-agent environments. A novel reward function based on multi-agent positional constraints is designed, and a training strategy based on incremental learning is performed to achieve collaborative path planning of multiple agents. Moreover, the greedy and Boltzmann probability selection policy is introduced for action selection and avoiding convergence to local extremum. To match radar and image sensors, a convolutional neural network - long short-term memory (CNN-LSTM) architecture is constructed to extract the feature of multi-source measurement as the input of the D3QN. The algorithm’s efficacy and reliability are validated in a simulated environment, utilizing robot operating system and Gazebo. The simulation results show that the proposed algorithm provides a real-time solution for path planning tasks in dynamic scenarios. In terms of the average success rate and accuracy, the proposed method is superior to other deep learning algorithms, and the convergence speed is also improved. Reference | Related Articles | Metrics | Comments（0）

Select

Video-Based Detection of Epileptic Spasms in IESS: Modeling, Detection, and Evaluation DING Lihui1, 2(丁黎辉), FU Lijun1, 3 (付立军), YANG Guang4(杨光), WAN Lin4, 5 (万林), CHANG Zhijun7(常志军) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 1-9.   DOI: 10.1007/s12204-024-2789-x Abstract （350）      PDF（pc） （712KB）（46）       Knowledge map    Save Behavioral scoring based on clinical observations remains the gold standard for screening, diagnosing,and evaluating infantile epileptic spasm syndrome (IESS). The accurate identification of seizures is crucial for clinical diagnosis and assessment. In this study, we propose an innovative seizure detection method based on video feature recognition of patient spasms. To capture the temporal characteristics of the spasm behavior presented in the videos effectively, we incorporate asymmetric convolution and convolution–batch normalization–ReLU (CBR) modules. Specifically within the 3D-ResNet residual blocks, we split the larger convolutional kernels into two asymmetric 3D convolutional kernels. These kernels are connected in series to enhance the ability of the convolutional layers to extract key local features, both horizontally and vertically. In addition, we introduce a 3D convolutional block attention module to enhance the spatial correlations between video frame channels efficiently. To improve the generalization ability, we design a composite loss function that combines cross-entropy loss with triplet loss to balance the classification and similarity requirements. We train and evaluate our method using the PLA IESS-VIDEO dataset, achieving an average seizure recognition accuracy of 90.59%, precision of 90.94%, and recall of 87.64%. To validate its generalization capability further, we conducted external validation using six different patient monitoring videos compared with assessments by six human experts from various medical centers. The final test results demonstrate that our method achieved a recall of 0.647 6, surpassing the average level achieved by human experts (0.559 5), while attaining a high F1-score of 0.721 9. These findings have substantial significance for the long-term assessment of patients with IESS. Reference | Related Articles | Metrics | Comments（0）

Select

Applications of Artificial Intelligence in Cardiac Electrophysiology and Clinical Diagnosis with Magnetic Resonance Imaging and Computational Modeling Techniques ZHAN Heqing1 (詹何庆), HAN Guilai1 (韩贵来), WEI Chuan’an1 (魏传安), LI Zhiqun2* (李治群) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 53-65.   DOI: 10.1007/s12204-023-2628-5 Abstract （323）      PDF（pc） （232KB）（23）       Knowledge map    Save The underlying electrophysiological mechanisms and clinical treatments of cardiovascular diseases, which are the most common cause of morbidity and mortality worldwide, have gotten a lot of attention and been widely explored in recent decades. Along the way, techniques such as medical imaging, computing modeling, and artificial intelligence (AI) have always played significant roles in above studies. In this article, we illustrated the applications of AI in cardiac electrophysiological research and disease prediction. We summarized general principles of AI and then focused on the roles of AI in cardiac basic and clinical studies incorporating magnetic resonance imaging and computing modeling techniques. The main challenges and perspectives were also analyzed. Reference | Related Articles | Metrics | Comments（0）

Select

Medical Image Encryption Based on Fisher-Yates Scrambling and Filter Diffusion HUANG Jiaxin (黄佳鑫), GUO Yali (郭亚丽), GAO Ruoyun (高若云),LI Shanshan∗ (李珊珊) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 136-152.   DOI: 10.1007/s12204-023-2618-7 Abstract （301）      PDF（pc） （8076KB）（13）       Knowledge map    Save A medical image encryption is proposed based on the Fisher-Yates scrambling, filter diffusion and S-box substitution. First, chaotic sequence associated with the plaintext is generated by logistic-sine-cosine system, which is used for the scrambling, substitution and diffusion processes. The three-dimensional Fisher-Yates scrambling, S-box substitution and diffusion are employed for the first round of encryption. The chaotic sequence is adopted for secondary encryption to scramble the ciphertext obtained in the first round. Then, three-dimensional filter is applied to diffusion for further useful information hiding. The key to the algorithm is generated by the combination of hash value of plaintext image and the input parameters. It improves resisting ability of plaintext attacks. The security analysis shows that the algorithm is effective and efficient. It can resist common attacks. In addition, the good diffusion effect shows that the scheme can solve the differential attacks encountered in the transmission of medical images and has positive implications for future research. Reference | Related Articles | Metrics | Comments（0）

Select

Augmented Reality Navigation Using Surgical Guides Versus Conventional Techniques in Pedicle Screw Placement KONG Huiyang1 (孔会扬), WANG Shuyi1 (王殊轶), ZHANG Can2 (张璨), CHEN Zan2, 3 (陈赞) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 10-17.   DOI: 10.1007/s12204-023-2689-5 Abstract （296）      PDF（pc） （1106KB）（26）       Knowledge map    Save The aim of this study was to assess the potential of surgical guides as a complementary tool to augmented reality (AR) in enhancing the safety and precision of pedicle screw placement in spinal surgery. Four trainers were divided into the AR navigation group using surgical guides and the free-hand group. Each group consisted of a novice and an experienced spine surgeon. A total of 80 pedicle screws were implanted. First, the AR group reconstructed the 3D model and planned the screw insertion route according to the computed tomography data of L2 lumbar vertebrae. Then, the Microsoft HoloLensTM 2 was used to identify the vertebral model, and the planned virtual path was superimposed on the real cone model. Next, the screw was placed according to the projected trajectory. Finally, Micron Tracker was used to measure the deviation of screws from the preoperatively planned trajectory, and pedicle screws were evaluated using the Gertzbein-Robbins scale. In the AR group, the linear deviations of the experienced doctor and the novice were (1.59±0.39) mm and (1.73±0.52) mm respectively, and the angle deviations were 2.72◦ ± 0.61◦ and 2.87◦ ± 0.63◦ respectively. In the free-hand group, the linear deviations of the experienced doctor and the novice were (2.88 ± 0.58) mm and (5.25 ± 0.62) mm respectively, and the angle deviations were 4.41◦ ± 1.18◦ and 7.15◦ ± 1.45◦ respectively. Both kinds of deviations between the two groups were significantly different (P < 0.05). The screw accuracy rate was 95% in the AR navigation group and 77.5% in the free-hand group. The results of this study indicate that the integration of surgical guides and AR is an innovative technique that can substantially enhance the safety and precision of spinal surgery and assist inexperienced doctors in completing the surgery. Reference | Related Articles | Metrics | Comments（0）

Select

Anti-Occlusion Object Tracking Algorithm Based on Filter Prediction CHEN Kun(陈坤), ZHAO Xu(赵旭), DONG Chunyu(董春玉), DI Zichao(邸子超), CHEN Zongzhi(陈宗枝) J Shanghai Jiaotong Univ Sci    2024, 29 (3): 400-413.   DOI: 10.1007/s12204-022-2484-8 Abstract （287）      PDF（pc） （5510KB）（92）       Knowledge map    Save Visual object tracking is an important issue that has received long-term attention in computer vision.The ability to effectively handle occlusion, especially severe occlusion, is an important aspect of evaluating theperformance of object tracking algorithms in long-term tracking, and is of great significance to improving therobustness of object tracking algorithms. However, most object tracking algorithms lack a processing mechanism specifically for occlusion. In the case of occlusion, due to the lack of target information, it is necessary to predict the target position based on the motion trajectory. Kalman filtering and particle filtering can effectively predict the target motion state based on the historical motion information. A single object tracking method, called probabilistic discriminative model prediction (PrDiMP), is based on the spatial attention mechanism in complex scenes and occlusions. In order to improve the performance of PrDiMP, Kalman filtering, particle filtering and linear filtering are introduced. First, for the occlusion situation, Kalman filtering and particle filtering are respectively introduced to predict the object position, thereby replacing the detection result of the original tracking algorithm and stopping recursion of target model. Second, for detection-jump problem of similar objects in complex scenes, a linear filtering window is added. The evaluation results on the three datasets, including GOT-10k, UAV123 and LaSOT, and the visualization results on several videos, show that our algorithms have improved tracking performance under occlusion and the detection-jump is effectively suppressed. Reference | Related Articles | Metrics | Comments（0）

Select

Positional Information is a Strong Supervision for Volumetric Medical Image Segmentation ZHAO Yinjie1 (赵寅杰), HOU Runpingg1 (侯润萍), ZENG Wanqin2 (曾琬琴), QIN Yulei1 (秦玉磊), SHEN Tianle2 (沈天乐), XU Zhiyong2 (徐志勇), FU Xiaolong2* (傅小龙), SHEN Hongbin1* (沈红斌) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 121-129.   DOI: 10.1007/s12204-023-2614-y Abstract （261）      PDF（pc） （481KB）（3）       Knowledge map    Save Medical image segmentation is a crucial preliminary step for a number of downstream diagnosis tasks. As deep convolutional neural networks successfully promote the development of computer vision, it is possible to make medical image segmentation a semi-automatic procedure by applying deep convolutional neural networks to finding the contours of regions of interest that are then revised by radiologists. However, supervised learning necessitates large annotated data, which are difficult to acquire especially for medical images. Self-supervised learning is able to take advantage of unlabeled data and provide good initialization to be finetuned for downstream tasks with limited annotations. Considering that most self-supervised learning especially contrastive learning methods are tailored to natural image classification and entail expensive GPU resources, we propose a novel and simple pretext-based self-supervised learning method that exploits the value of positional information in volumetric medical images. Specifically, we regard spatial coordinates as pseudo labels and pretrain the model by predicting positions of randomly sampled 2D slices in volumetric medical images. Experiments on four semantic segmentation datasets demonstrate the superiority of our method over other self-supervised learning methods in both semisupervised learning and transfer learning settings. Codes are available at https://github.com/alienzyj/PPos. Reference | Related Articles | Metrics | Comments（0）

Select

Semi-Autonomous Navigation Based on Local Semantic Map for Mobile Robot ZHAO Yanfei1，2，3(赵艳飞), XIAO Peng4 (肖鹏), WANG Jingchuan1，2，3* (王景川), GUO Rui4*(郭锐） J Shanghai Jiaotong Univ Sci    2025, 30 (1): 27-33.   DOI: 10.1007/s12204-023-2678-8 Abstract （256）      PDF（pc） （995KB）（12）       Knowledge map    Save Mobile robots represented by smart wheelchairs can assist elderly people with mobility difficulties. This paper proposes a multi-mode semi-autonomous navigation system based on a local semantic map for mobile robots, which can assist users to implement accurate navigation (e.g., docking) in the environment without prior maps. In order to overcome the problem of repeated oscillations during the docking of traditional local path planning algorithms, this paper adopts a mode-switching method and uses feedback control to perform docking when approaching semantic goals. At last, comparative experiments were carried out in the real environment. Results show that our method is superior in terms of safety, comfort and docking accuracy. Reference | Related Articles | Metrics | Comments（0）

Select

Adjacent Segment Biomechanical Changes After Implantation of Cage Plus Plate or Zero-Profile Device in Different Segmental Anterior Cervical Discectomy and Fusion YE Peng (叶鹏), FU Rongchang∗ (富荣昌), WANG Zhaoyao (王召耀) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 166-174.   DOI: 10.1007/s12204-023-2633-8 Abstract （254）      PDF（pc） （929KB）（17）       Knowledge map    Save Cage plus plate (CP) and zero-profile (Zero-P) devices are widely used in anterior cervical discectomy and fusion (ACDF). This study aimed to compare adjacent segment biomechanical changes after ACDF when using Zero-P device and CP in different segments. First, complete C1—C7 cervical segments were constructed and validated. Meanwhile, four surgery models were developed by implanting the Zero-P device or CP into C4—C5 or C5—C6 segments based on the intact model. The segmental range of motion (ROM) and maximum value of the intradiscal pressure of the surgery models were compared with those of the intact model. The implantation of CP and Zero-P devices in C4—C5 segments decreased ROM by about 91.6% and 84.3%, respectively, and increased adjacent segment ROM by about 8.3% and 6.82%, respectively. The implantation of CP and Zero-P devices in C5—C6 segments decreased ROM by about 93.3% and 89.9%, respectively, while increasing adjacent segment ROM by about 4.9% and 4%, respectively. Furthermore, the implantation of CP and Zero-P devices increased the intradiscal pressure in the adjacent segments of C4—C5 segments by about 4.5% and 6.7%, respectively. The implantation of CP and Zero-P devices significantly increased the intradiscal pressure in the adjacent segments of C5—C6 by about 54.1% and 15.4%, respectively. In conclusion, CP and Zero-P fusion systems can significantly reduce the ROM of the fusion implant segment in ACDF while increasing the ROM and intradiscal pressure of adjacent segments. Results showed that Zero-P fusion system is the best choice for C5—C6 segmental ACDF. However, further studies are needed to select the most suitable cervical fusion system for C4—C5 segmental ACDF. Therefore, this study provides biomechanical recommendations for clinical surgery. Reference | Related Articles | Metrics | Comments（0）

Select

Ship Pipe Layout Optimization Based on Improved Particle Swarm Optimization LIN Yan1, 2(林焰), BIAN Xuanyi1(卞璇屹), DONG Zongran3(董宗然) J Shanghai Jiaotong Univ Sci    2024, 29 (5): 737-746.   DOI: 10.1007/s12204-022-2530-6 Abstract （251）      PDF（pc） （1456KB）（100）       Knowledge map    Save Ship pipe layout optimization is one of the difficulties and hot spots in ship intelligent production design. A high-dimensional vector coding is proposed based on the research of related pipe coding and ship pipe route features in this paper. The advantages of this coding method are concise structure, strong compatibility, and independence from the gridding space. Based on the proposed coding, the particle swarm optimization algorithm is implemented, and the algorithm is improved by the pre-selected path strategy and the branch-pipe processing strategy. Finally, two simulation results reveal that the proposed coding and algorithm have feasibility and engineering practicability. Reference | Related Articles | Metrics | Comments（0）

Select

Vibration Transmission Characteristics of Shoe Sole Based on Mechanical Mobility and Vibration Transmissibility WU Xuyang1 (吴旭阳), LIU Xiaoying1 (刘晓颖), HAO Yanhua1 (郝艳华), LIU Changhuang1 (刘长煌), HUANG Xianwei2 (黄贤伟) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 175-186.   DOI: 10.1007/s12204-023-2587-x Abstract （247）      PDF（pc） （2284KB）（9）       Knowledge map    Save It is particularly important to explore the response and transmission characteristics of shoe sole when exposed to foot-transmitted vibration (FTV) in daily life. In this study, based on mechanical mobility and vibration transmissibility, the vibration response and transmission characteristics of ordinary sole and multicellular structure sole under three excitation modes were analyzed with finite element analysis. The analysis results of the ordinary sole are as follows: The distribution and transmission of vibration energy of ordinary sole are more related to the excitation position and mode-shape; the phalange region is more violent in vibration response to vibration and transmission of vibration. In addition, the analysis results of multi-cellular structure soles show that different types of multi-cellular structure soles have different effects on the equivalent mechanical mobility and the equivalent vibration transmissibility, among which Grid type has the greatest influence. So, this study can help prevent foot injury and provide guidance for the optimal design of the sole. Reference | Related Articles | Metrics | Comments（0）

Select

Federated Approach for Privacy-Preserving Traffic Prediction Using Graph Convolutional Network LONARE Savita1，2* , BHRAMARAMBA Ravi2 J Shanghai Jiaotong Univ Sci    2024, 29 (3): 509-517.   DOI: 10.1007/s12204-021-2382-5 Abstract （243）      PDF（pc） （525KB）（47）       Knowledge map    Save Existing traffic flow prediction frameworks have already achieved enormous success due to large traffic datasets and capability of deep learning models. However, data privacy and security are always a challenge in every field where data need to be uploaded to the cloud. Federated learning (FL) is an emerging trend for distributed training of data. The primary goal of FL is to train an efficient communication model without compromising data privacy. The traffic data have a robust spatio-temporal correlation, but various approaches proposed earlier have not considered spatial correlation of the traffic data. This paper presents FL-based traffic flow prediction with spatio-temporal correlation. This work uses a differential privacy (DP) scheme for privacy preservation of participant’s data. To the best of our knowledge, this is the first time that FL is used for vehicular traffic prediction while considering the spatio-temporal correlation of traffic data with DP preservation. The proposed framework trains the data locally at the client-side with DP. It then uses the model aggregation mechanism federated graph convolutional network (FedGCN) at the server-side to find the average of locally trained models. The results of the proposed work show that the FedGCN model accurately predicts the traffic. DP scheme at client-side helps clients to set a budget for privacy loss. Reference | Related Articles | Metrics | Comments（0）

Select

Online Multi-Object Tracking Under Moving Unmanned Aerial Vehicle Platform Based on Object Detection and Feature Extraction Network LIU Zengmin (刘增敏), WANG Shentao(王申涛), YAO Lixiu(姚莉秀), CAI Yunze(蔡云泽) J Shanghai Jiaotong Univ Sci    2024, 29 (3): 388-399.   DOI: 10.1007/s12204-022-2540-4 Abstract （238）      PDF（pc） （1105KB）（81）       Knowledge map    Save In order to solve the problem of small object size and low detection accuracy under the unmanned aerial vehicle (UAV) platform, the object detection algorithm based on deep aggregation network and high-resolution fusion module is studied. Furthermore, a joint network of object detection and feature extraction is studied to construct a real-time multi-object tracking algorithm. For the problem of object association failure caused by UAV movement, image registration is applied to multi-object tracking and a camera motion discrimination model is proposed to improve the speed of the multi-object tracking algorithm. The simulation results show that the algorithm proposed in this study can improve the accuracy of multi-object tracking under the UAV platform, and effectively solve the problem of association failure caused by UAV movement. Reference | Related Articles | Metrics | Comments（0）

Select

Electrocardiogram Signal Denoising Using Optimized Adaptive Hybrid Filter with Empirical Wavelet Transform BALASUBRAMANIAN S1*, NARUKA Mahaveer Singh2, TEWARI Gaurav3 J Shanghai Jiaotong Univ Sci    2025, 30 (1): 66-80.   DOI: 10.1007/s12204-023-2591-1 Abstract （235）      PDF（pc） （1496KB）（3）       Knowledge map    Save Cardiovascular diseases are the world’s leading cause of death; therefore cardiac health of the human heart has been a fascinating topic for decades. The electrocardiogram (ECG) signal is a comprehensive noninvasive method for determining cardiac health. Various health practitioners use the ECG signal to ascertain critical information about the human heart. In this article, swarm intelligence approaches are used in the biomedical signal processing sector to enhance adaptive hybrid filters and empirical wavelet transforms (EWTs). At first, the white Gaussian noise is added to the input ECG signal and then applied to the EWT. The ECG signals are denoised by the proposed adaptive hybrid filter. The honey badge optimization (HBO) algorithm is utilized to optimize the EWT window function and adaptive hybrid filter weight parameters. The proposed approach is simulated by MATLAB 2018a using the MIT-BIH dataset with white Gaussian, electromyogram and electrode motion artifact noises. A comparison of the HBO approach with recursive least square-based adaptive filter, multichannel least means square, and discrete wavelet transform methods has been done in order to show the efficiency of the proposed adaptive hybrid filter. The experimental results show that the HBO approach supported by EWT and adaptive hybrid filter can be employed efficiently for cardiovascular signal denoising. Reference | Related Articles | Metrics | Comments（0）

Select

Tree Detection Algorithm Based on Embedded YOLO Lightweight Network LV Feng(吕峰), WANG Xinyan* (王新彦), LI Lei(李磊), JIANG Quan(江泉), YI Zhengyang(易政洋) J Shanghai Jiaotong Univ Sci    2024, 29 (3): 518-527.   DOI: 10.1007/s12204-022-2451-4 Abstract （234）      PDF（pc） （2443KB）（38）       Knowledge map    Save To avoid colliding with trees during its operation, a lawn mower robot must detect the trees. Existing tree detection methods suffer from low detection accuracy (missed detection) and the lack of a lightweight model. In this study, a dataset of trees was constructed on the basis of a real lawn environment. According to the theory of channel incremental depthwise convolution and residual suppression, the Embedded-A module is proposed, which expands the depth of the feature map twice to form a residual structure to improve the lightweight degree of the model. According to residual fusion theory, the Embedded-B module is proposed, which improves the accuracy of feature-map downsampling by depthwise convolution and pooling fusion. The Embedded YOLO object detection network is formed by stacking the embedded modules and the fusion of feature maps of different resolutions. Experimental results on the testing set show that the Embedded YOLO tree detection algorithm has 84.17% and 69.91% average precision values respectively for trunk and spherical tree, and 77.04% mean average precision value. The number of convolution parameters is 1.78 × 106, and the calculation amount is 3.85 billion float operations per second. The size of weight file is 7.11 MB, and the detection speed can reach 179 frame/s. This study provides a theoretical basis for the lightweight application of the object detection algorithm based on deep learning for lawn mower robots. Reference | Related Articles | Metrics | Comments（0）

Select

Direct Ink Writing Method of Fractal Wearable Flexible Sensor Based on Conductive Graphene/Polydimethylsiloxane Ink CHEN Junling1, 2, 3 (陈俊伶), GAO Feiyang1, 3 (高飞扬), ZHANG Liming1, 3 (张黎明), ZHENG Xiongfei1, 3(郑雄飞) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 18-26.   DOI: 10.1007/s12204-023-2687-7 Abstract （219）      PDF（pc） （1712KB）（13）       Knowledge map    Save Flexible electronic technology has laid the foundation for complex human-computer interaction system, and has attracted great attention in the field of human motion detection and soft robotics. Graphene has received an extensive attention due to its excellent electrical conductivity; however, how to use it to fabricate wearable flexible sensors with complex structures remains challenging. In this study, we studied the rheological behavior of graphene/polydimethylsiloxane ink and proposed an optimal graphene ratio, which makes the ink have a good printability and conductivity at the same time. Then, based on the theory of Peano fractal layout, we proposed a two-dimensional structure that can withstand multi-directional tension by replacing the traditional arris structure with the arc structure. After that, we manufactured circular arc fractal structure sensor by adjusting ink composition and printing structure through direct ink writing method. Finally, we evaluated the detection performance and repeatability of the sensor. This method provides a simple and effective solution for fabricating wearable flexible sensors and exhibits the potential to fabricate 3D complex flexible electronic devices. Reference | Related Articles | Metrics | Comments（0）

Select

TshFNA-Examiner: A Nuclei Segmentation and Cancer Assessment Framework for Thyroid Cytology Image KE Jing1(柯晶), ZHU Junchao2 (朱俊超), YANG Xin1(杨鑫), ZHANG Haolin3 (张浩林), SUN Yuxiang1(孙宇翔), WANG Jiayi1(王嘉怡), LU Yizhou4(鲁亦舟), SHEN Yiqing5(沈逸卿), LIU Sheng6(刘晟), JIANG Fusong7(蒋伏松), HUANG Qin8(黄琴) J Shanghai Jiaotong Univ Sci    2024, 29 (6): 945-957.   DOI: 10.1007/s12204-024-2743-y Abstract （218）      PDF（pc） （2836KB）（77）       Knowledge map    Save Examining thyroid fine-needle aspiration (FNA) can grade cancer risks, derive prognostic information, and guide follow-up care or surgery. The digitization of biopsy and deep learning techniques has recently enabled computational pathology. However, there is still lack of systematic diagnostic system for the complicated gigapixel cytopathology images, which can match physician-level basic perception. In this study, we design a deep learning framework, thyroid segmentation and hierarchy fine-needle aspiration (TshFNA)-Examiner to quantitatively profile the cancer risk of a thyroid FNA image. In the TshFNA-Examiner, cellular-intensive areas strongly correlated with diagnostic medical information are detected by a nuclei segmentation neural network; cell-level image patches are catalogued following The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) system, by a classification neural network which is further enhanced by leveraging unlabeled data. A cohort of 333 thyroid FNA cases collected from 2019 to 2022 from I to VI is studied, with pixel-wise and image-wise image patches annotated. Empirically, TshFNA-Examiner is evaluated with comprehensive metrics and multiple tasks to demonstrate its superiority to state-of-the-art deep learning approaches. The average performance of cellular area segmentation achieves a Dice of 0.931 and Jaccard index of 0.871. The cancer risk classifier achieves a macro-F1-score of 0.959, macro-AUC of 0.998, and accuracy of 0.959 following TBSRTC. The corresponding metrics can be enhanced to a macro-F1-score of 0.970, macro-AUC of 0.999, and accuracy of 0.970 by leveraging informative unlabeled data. In clinical practice, TshFNA-Examiner can help cytologists to visualize the output of deep learning networks in a convenient way to facilitate making the final decision. Reference | Related Articles | Metrics | Comments（0）

Select

Multi-Channel Based on Attention Network for Infrared Small Target Detection ZHANG Yanjun(张彦军), WANG Biyun(王碧云)，CAI Yunze (蔡云泽) J Shanghai Jiaotong Univ Sci    2024, 29 (3): 414-427.   DOI: 10.1007/s12204-023-2616-9 Abstract （217）      PDF（pc） （1697KB）（53）       Knowledge map    Save Infrared detection technology has the advantages of all-weather detection and good concealment,which is widely used in long-distance target detection and tracking systems. However, the complex background,the strong noise, and the characteristics of small scale and weak intensity of targets bring great difficulties to the detection of infrared small targets. A multi-channel based on attention network is proposed in this paper, aimed at the problem of high missed detection rate and false alarm rate of traditional algorithms and the problem of large model, high complexity and poor detection performance of deep learning algorithms. First, given the difficulty in extracting the features of infrared multiscale and small dim targets, the multiple channels are designed based on dilated convolution to capture multiscale target features. Second, the coordinate attention block is incorporated in each channel to suppress background clutters adaptively and enhance target features. In addition, the fusion of shallow detail features and deep abstract semantic features is realized by synthesizing the contextual attention fusion block. Finally, it is verified that, compared with other state-of-the-art methods based on the datasets SIRST and MDFA, the proposed algorithm further improves the detection effect, and the model size and computational complexity are smaller. Reference | Related Articles | Metrics | Comments（0）

Select

Biomechanical Analysis of Scoliosis Orthopedic Force Loading with Human Avoidance Effect ZHU Ye1 (朱晔), REN Dong1 (任东), ZHANG Shuang2 (张爽), CAO Qian3 (曹倩) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 187-196.   DOI: 10.1007/s12204-023-2620-0 Abstract （215）      PDF（pc） （1836KB）（2）       Knowledge map    Save Due to the lack of human avoidance analysis, the orthosis cannot accurately apply orthopedic force during orthopedic, resulting in poor orthopedic effect. Therefore, the relationship between the human body’s active avoidance ability and force application is studied to achieve accurate loading of orthopedic force. First, a high-precision scoliosis model was established based on computed tomography data, and the relationship between orthopedic force and Cobb angle was analyzed. Then 9 subjects were selected for avoidance ability test grouped by body mass index calculation, and the avoidance function of different groups was fitted. The avoidance function corrected the application of orthopedic forces. The results show that the optimal correction force calculated by the finite element method was 60 N. The obese group had the largest avoidance ability, followed by the standard group and the lean group. When the orthopedic force was 60 N, the Cobb angle was reduced from 33.77◦ to 20◦, the avoidance ability of the standard group at 50 N obtained from the avoidance function was 20.28% and 10.14 N was actively avoided. Therefore, when 50 N was applied, 60.14 N was actually generated, which can achieve the orthopedic effect of 60 N numerical simulation analysis. The avoidance effect can take the active factors of the human body into consideration in the orthopedic process, so as to achieve a more accurate application of orthopedic force, and provide data reference for clinicians in the orthopedic process. Reference | Related Articles | Metrics | Comments（0）

Select

Fast Four-Stage Local Motion Planning Method for Mobile Robot HUANG Shan(黄山), HUANG Hongzhong(黄洪钟), ZENG Qi(曾奇) J Shanghai Jiaotong Univ Sci    2024, 29 (3): 428-435.   DOI: 10.1007/s12204-022-2423-8 Abstract （214）      PDF（pc） （1810KB）（36）       Knowledge map    Save Mobile robot local motion planning is responsible for the fast and smooth obstacle avoidance, which is one of the main indicators for evaluating mobile robots’ navigation capabilities. Current methods formulate local motion planning as a unified problem; therefore it cannot satisfy the real-time requirement on the platform with limited computing ability. In order to solve this problem, this paper proposes a fast local motion planning method that can reach a planning frequency of 500 Hz on a low-cost CPU. The proposed method decouples the local motion planning as the front-end path searching and the back-end optimization. The front-end is composed of the environment topology analysis and graph searching. The back-end includes dynamically feasible trajectory generation and optimal trajectory selection. Different from the popular methods, the proposed method decomposes the local motion planning into four sub-modules, each of which aims to solve one problem. Combining four submodules, the proposed method can obtain the complete local motion planning algorithm which can fast generate a smooth and collision-free trajectory. The experimental results demonstrate that the proposed method has the ability to obtain the smooth, dynamically feasible and collision-free trajectory and the speed of the planning is fast. Reference | Related Articles | Metrics | Comments（0）

Select

New Lite YOLOv4-Tiny Algorithm and Application on Crack Intelligent Detection SONG Liboa (宋立博), FEI Yanqiongb (费燕琼) J Shanghai Jiaotong Univ Sci    2024, 29 (3): 528-536.   DOI: 10.1007/s12204-022-2504-8 Abstract （212）      PDF（pc） （1358KB）（41）       Knowledge map    Save Conforming to the rapidly increasing market demand of crack detection for tall buildings, the idea of integrating deep network technology into wall climbing robot for crack detection is put forward in this paper. Taking the dependence and hardware requirements when deployed on such edge devices as Raspberry Pi into consideration, the Darknet neural network is selected as the basic framework for detection. In order to improve the inference efficiency on edge devices and avoid the possible premature over-fitting of deep networks, the lite YOLOv4-tiny algorithm is then improved from the original YOLOv4-tiny algorithm and its structure is illustrated using Netron accordingly. The images downloaded from Internet and taken from the buildings in campus are processed to form crack detection data sets, which are trained on personal computer with the AlexeyAB version of Darknet to generate weight files. Meanwhile, the AlexeyAB version of Darknet accelerated by NNpack package is deployed on Raspberry Pi 4B, and the crack detection experiments are carried out. Some characteristics, e.g., fast speed and lower false detection rate of the lite YOLOv4-tiny algorithm, are confirmed by comparison with those of original YOLOv4-tiny algorithm. The innovations of this paper focus on the simple network structure, fewer network layers, and earlier forward transmission of features to prevent over-fitting, showing the new lite neural network exceeds the original YOLOv4-tiny network significantly. Reference | Related Articles | Metrics | Comments（0）

Select

Event-Triggered Fixed-Time Consensus of Second-Order Nonlinear Multi-Agent Systems with Delay and Switching Topologies XING Youjing1 (邢优靖), GAO Jinfeng1∗ (高金凤), LIU Xiaoping1, 2 (刘小平), WU Ping1 (吴平) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 625-639.   DOI: 10.1007/s12204-024-2695-2 Abstract （210）      PDF（pc） （1059KB）（63）       Knowledge map    Save To address fixed-time consensus problems of a class of leader-follower second-order nonlinear multiagent systems with uncertain external disturbances, the event-triggered fixed-time consensus protocol is proposed. First, the virtual velocity is designed based on the backstepping control method to achieve the system consensus and the bound on convergence time only depending on the system parameters. Second, an event-triggered mechanism is presented to solve the problem of frequent communication between agents, and triggered condition based on state information is given for each follower. It is available to save communication resources, and the Zeno behaviors are excluded. Then, the delay and switching topologies of the system are also discussed. Next, the system stabilization is analyzed by Lyapunov stability theory. Finally, simulation results demonstrate the validity of the presented method. Reference | Related Articles | Metrics | Comments（0）

Select

Coordination Design of a Power-Assisted Ankle Exoskeleton Robot Based on Active-Passive Combined Drive HE Guisong (贺贵松), HUANG Xuegong* (黄学功),LI Feng（李峰) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 197-208.   DOI: 10.1007/s12204-023-2589-8 Abstract （205）      PDF（pc） （2026KB）（14）       Knowledge map    Save With the continuous escalation of modern war, soldiers need to transport more combat materials to the combat area. The limited load-bearing capacity of soldiers seriously restricts their carrying capacity and mobility. It is urgent to develop a power-assisted exoskeleton robot suitable for individual combat. In the past, most power-assisted exoskeleton robots were driven by motors. This driving method has an excellent powerassisted effect, but the endurance is often insufficient. In view of this shortcoming, this study designed an ankle exoskeleton robot based on an active-passive combined drive through simulation analysis of human motion. It used OpenSim software to simulate and verify that the addition of spring could achieve a good effect. At the same time, according to the gait characteristics of the human body, the gait planning of an exoskeleton robot was carried out. Afterwards, theoretical analysis explained that the cooperation among spring, motor and wearer could be realized in this gait. Finally, the assisting ability and driving coordination of the active-passive combination driven ankle exoskeleton robot were verified through experiments. Reference | Related Articles | Metrics | Comments（0）

Select

Histological Image Diagnosis of Breast Cancer Based on Multi-Attention Convolution Neural Network XU Wangwang1，2 (徐旺旺), XU Liangfeng1，2 (许良凤), LIU Ninghui3(刘宁徽), LU Na3(律娜) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 91-106.   DOI: 10.1007/s12204-024-2705-4 Abstract （205）      PDF（pc） （1715KB）（10）       Knowledge map    Save Breast cancer is a serious and high morbidity disease in women, and it is the main cause of cancer death in China. However, getting tested and diagnosed early can reduce the risk of cancer. At present, there are clinical examinations, imaging screening and biopsies, among which histopathological examination is the gold standard. However, the process is complicated and time-consuming, and misdiagnosis may exist. This paper puts forward a classification framework based on deep learning, introducing multi-attention mechanism, selecting kernel convolution instead of ordinary convolution, and using different weights and combinations to pay attention to the accuracy index and growth rate of the model. In addition, we also compared the learning rate regulators. Error function can fine-tune the learning rate to achieve good performance, using label softening to reduce the loss error caused by model error recognition in the label, and assigning different category weights in the loss function to balance the positive and negative samples. We used the BreakHis data set to automatically classify histological images into benign and malignant, four categories and eight subtypes. Experimental results showed that the accuracy of binary classifications ranged from 98.23% to 99.50%, and that of multipl classifications ranged from 97.89% to 98.11%. Reference | Related Articles | Metrics | Comments（0）

Select

Brain Age Detection of Alzheimer’s Disease Magnetic Resonance Images Based on Mutual Information - Support Vector Regression LIU Yuchuan1 (刘玉川), LI Hao1 (李浩), TANG Yulong1 (唐宇龙), LIANG Dujuan2 (梁杜娟), TAN Jia3 (谭佳), FU Yue1 (符玥), LI Yongming4∗ (李勇明) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 130-135.   DOI: 10.1007/s12204-023-2590-2 Abstract （198）      PDF（pc） （635KB）（2）       Knowledge map    Save Brain age is an effective biomarker for diagnosing Alzheimer’s disease (AD). Aimed at the issue that the existing brain age detection methods are inconsistent with the biological hypothesis that AD is the accelerated aging of the brain, a mutual information - support vector regression (MI-SVR) brain age prediction model is proposed. First, the age deviation is introduced according to the biological hypothesis of AD. Second, fitness function is designed based on mutual information criterion. Third, support vector regression and fitness function are used to obtain the predicted brain age and fitness value of the subjects, respectively. The optimal age deviation is obtained by maximizing the fitness value. Finally, the proposed method is compared with some existing brain age detection methods. Experimental results show that the brain age obtained by the proposed method has better separability, can better reflect the accelerated aging of AD, and is more helpful for improving the diagnostic accuracy of AD. Reference | Related Articles | Metrics | Comments（0）

Select

Significant Retest Effects in Spatial Working Memory Task MA Xianda1,2,3‡ (马显达), LAN Zhaohui1,2,3‡ (兰兆辉),CHEN Zhitang1,2,3 (陈志堂), MONISHA M L4, HE Xinyi1,2,3 (何欣怡), LI Weidong1,2,3* (李卫东) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 115-120.   DOI: 10.1007/s12204-023-2585-z Abstract （197）      PDF（pc） （566KB）（2）       Knowledge map    Save Working memory is a core cognitive function that supports goal-directed behavior and complex thought. We developed a spatial working memory and attention test on paired symbols (SWAPS) which has been proved to be a useful and valid tool for spatial working memory and attention studies in the fields of cognitive psychology, education, and psychiatry. The repeated administration of working memory capacity tests is common in clinical and research settings. Studies suggest that repeated cognitive tests may improve the performance scores also known as retest effects. The systematic investigation of retest effects in SWAPS is critical for interpreting scientific results, but it is still not fully developed. To address this, we recruited 77 college students aged 18—21 years and used SWAPS comprising 72 trials with different memory loads, learning time, and delay span. We repeated the test once a week for five weeks to investigate the retest effects of SWAPS. There were significant retest effects in the first two tests: the accuracy of the SWAPS tests significantly increased, and then stabilized. These findings provide useful information for researchers to appropriately use or interpret the repeated working memory tests. Further experiments are still needed to clarify the factors that mediate the retest effects, and find out the cognitive mechanism that influences the retest effects. Reference | Related Articles | Metrics | Comments（0）

Select

Reward Function Design Method for Long Episode Pursuit Tasks Under Polar Coordinate in Multi-Agent Reinforcement Learning DONG Yubo1 (董玉博), CUI Tao1 (崔涛), ZHOU Yufan1 (周禹帆), SONG Xun2 (宋勋), ZHU Yue2 (祝月), DONG Peng1∗ (董鹏) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 646-655.   DOI: 10.1007/s12204-024-2713-4 Abstract （194）      PDF（pc） （567KB）（73）       Knowledge map    Save Multi-agent reinforcement learning has recently been applied to solve pursuit problems. However, it suffers from a large number of time steps per training episode, thus always struggling to converge effectively, resulting in low rewards and an inability for agents to learn strategies. This paper proposes a deep reinforcement learning (DRL) training method that employs an ensemble segmented multi-reward function design approach to address the convergence problem mentioned before. The ensemble reward function combines the advantages of two reward functions, which enhances the training effect of agents in long episode. Then, we eliminate the non-monotonic behavior in reward function introduced by the trigonometric functions in the traditional 2D polar coordinates observation representation. Experimental results demonstrate that this method outperforms the traditional single reward function mechanism in the pursuit scenario by enhancing agents’ policy scores of the task. These ideas offer a solution to the convergence challenges faced by DRL models in long episode pursuit problems, leading to an improved model training performance. Reference | Related Articles | Metrics | Comments（0）

Select

Ensemble Learning-Based Mortality Prediction After Acute Myocardial Infarction YAN Mingxuan1 (颜铭萱), MIAO Yutong2，3 (苗雨桐), SHENG Shuqian1 (盛淑茜), GAN Xiaoying1 (甘小莺), HE Ben2 (何 奔), SHEN Lan2，3* (沈 兰) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 153-165.   DOI: 10.1007/s12204-023-2611-1 Abstract （190）      PDF（pc） （710KB）（10）       Knowledge map    Save A mortality prediction model based on small acute myocardial infarction (AMI) patients coherent with low death rate is established. In total, 1 639 AMI patients are selected as research objects who received treatment in seven tertiary and secondary hospitals in Shanghai between January 1, 2016 and January 1, 2018. Among them, 72 patients deceased during the two-year follow-up. Models are established with ensemble learning framework and machine learning algorithms based on 51 physiological indicators of the patient. Shapley additive explanations algorithm and univariate test with point-biserial and phi correlation coefficients are employed to determine significant features and rank feature importance. Based on 5-fold cross validation experiment and external validation, prediction model with self-paced ensemble framework and random forest algorithm achieves the best performance with area under receiver operating characteristic curve (AUROC) score of 0.911 and recall of 0.864. Both feature ranking methods showed that ejection fractions, serum creatinine (admission), hemoglobin and Killip class are the most important features. With these top-ranked features, the simplified prediction model is capable of achieving a comparable result with AUROC score of 0.872 and recall of 0.818. This work proposes a new method to establish mortality prediction models for AMI patients based on self-paced ensemble framework, which allows models to achieve high performance with small scale of patients coherent with low death rate. It will assist in medical decision and prognosis as a new reference. Reference | Related Articles | Metrics | Comments（0）

Select

Data Augmentation of Ship Wakes in SAR Images Based on Improved CycleGAN YAN Congqiang1,2 (鄢丛强), GUO Zhengyun3,4 (郭正玉), CAI Yunze1,2∗∗ (蔡云泽) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 702-711.   DOI: 10.1007/s12204-024-2746-8 Abstract （189）      PDF（pc） （1418KB）（38）       Knowledge map    Save The study on ship wakes of synthetic aperture radar (SAR) images holds great importance in detecting ship targets in the ocean. In this study, we focus on the issues of low quantity and insufficient diversity in ship wakes of SAR images, and propose a method of data augmentation of ship wakes in SAR images based on the improved cycle-consistent generative adversarial network (CycleGAN). The improvement measures mainly include two aspects: First, to enhance the quality of the generated images and guarantee a stable training process of the model, the least-squares loss is employed as the adversarial loss function; Second, the decoder of the generator is augmented with the convolutional block attention module (CBAM) to address the issue of missing details in the generated ship wakes of SAR images at the microscopic level. The experiment findings indicate that the improved CycleGAN model generates clearer ship wakes of SAR images, and outperforms the traditional CycleGAN models in both subjective and objective aspects. Reference | Related Articles | Metrics | Comments（0）

Select

Augmented Reality Based Navigation System for Endoscopic Transnasal Optic Canal Decompression FU Hang1 (傅航),XU Jiangchang1 (许江长), LI Yinwei2，4* (李寅炜),ZHOU Huifang2，4 (周慧芳),CHEN Xiaojun1，3* (陈晓军) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 34-42.   DOI: 10.1007/s12204-024-2722-3 Abstract （186）      PDF（pc） （2193KB）（12）       Knowledge map    Save Endoscopic transnasal optic nerve decompression surgery plays a crucial role in minimal invasive treatment of complex traumatic optic neuropathy. However, a major challenge faced during the procedure is the inability to visualize the optic nerve intraoperatively. To address this issue, an endoscopic image-based augmented reality surgical navigation system is developed in this study. The system aims to virtually fuse the optic nerve onto the endoscopic images, assisting surgeons in determining the optic nerve’s position and reducing surgical risks. First, a calibration algorithm based on a checkerboard grid of immobile points is proposed, building upon existing calibration methods. Additionally, to tackle accuracy issues associated with augmented reality technology, an optical navigation and visual fusion compensation algorithm is proposed to improve the intraoperative tracking accuracy. To evaluate the system’s performance, model experiments were meticulously designed and conducted. The results confirm the accuracy and stability of the proposed system, with an average tracking error of (0.99 ± 0.46) mm. This outcome demonstrates the effectiveness of the proposed algorithm in improving the augmented reality surgical navigation system’s accuracy. Furthermore, the system successfully displays hidden optic nerves and other deep tissues, thus showcasing the promising potential for future applications in orbital and maxillofacial surgery. Reference | Related Articles | Metrics | Comments（0）

Select

LOBO Optimization-Tuned Deep-Convolutional Neural Network for Brain Tumor Classification Approach A. Sahaya Anselin Nisha1* , NARMADHA R.1 , AMIRTHALAKSHMI T. M.2,BALAMURUGAN V.1, VEDANARAYANAN V.1 J Shanghai Jiaotong Univ Sci    2025, 30 (1): 107-114.   DOI: 10.1007/s12204-023-2625-8 Abstract （184）      PDF（pc） （1079KB）（14）       Knowledge map    Save The categorization of brain tumors is a significant issue for healthcare applications. Perfect and timely identification of brain tumors is important for employing an effective treatment of this disease. Brain tumors possess high changes in terms of size, shape, and amount, and hence the classification process acts as a more difficult research problem. This paper suggests a deep learning model using the magnetic resonance imaging technique that overcomes the limitations associated with the existing classification methods. The effectiveness of the suggested method depends on the coyote optimization algorithm, also known as the LOBO algorithm, which optimizes the weights of the deep-convolutional neural network classifier. The accuracy, sensitivity, and specificity indices, which are obtained to be 92.40%, 94.15%, and 91.92%, respectively, are used to validate the effectiveness of the suggested method. The result suggests that the suggested strategy is superior for effectively classifying brain tumors. Reference | Related Articles | Metrics | Comments（0）

Select

Experimental Study and Numerical Simulation of Evacuation in an Offshore Platform ZHANG Jingjinga (张菁菁), ZHAO Jinchenga, b, c∗(赵金城), SONG Zhensena, b, c (宋振森), DUAN Lipinga, b, c(段立平) J Shanghai Jiaotong Univ Sci    2024, 29 (5): 747-758.   DOI: 10.1007/s12204-023-2629-4 Abstract （170）      PDF（pc） （4007KB）（48）       Knowledge map    Save With the rapid development of marine oil and gas exploitation, the evacuation of offshore platforms has received more attention. First, an experimental investigation of the evacuation process of 120 participants in a real offshore platform is performed, and then simulation results provided by Pathfinder are validated against the measurement results. Second, four typical evacuation scenarios on the platform referring to IMO guidelines are investigated by Pathfinder with the speed values achieved in experiments. The simulation results show that both the utilization of exits and evacuation efficiency of people on the offshore platform need to be further improved. Last, the evacuation routes of people under the four scenarios are optimized, and the improvement of the evacuation performance after the optimization is evaluated by several mathematical indicators. Final results show that the evacuation with the optimized route design prompts the use efficiency of exits and further reduces the evacuation time. The present study provides a useful advice for potentially revising the IMO guidelines in future and provides efficient evacuation strategies for planning the emergency evacuation on offshore platforms. Reference | Related Articles | Metrics | Comments（0）

Select

Multi-Objective Loosely Synchronized Search for Multi-Objective Multi-Agent Path Finding with Asynchronous Actions DU Haikuo1,2 (杜海阔), GUO Zhengyu3，4(郭正玉), ZHANG Lulu1,2(章露露), CAI Yunze1,2∗ (蔡云泽) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 667-677.   DOI: 10.1007/s12204-024-2744-x Abstract （163）      PDF（pc） （1177KB）（41）       Knowledge map    Save In recent years, the path planning for multi-agent technology has gradually matured, and has made breakthrough progress. The main difficulties in path planning for multi-agent are large state space, long algorithm running time, multiple optimization objectives, and asynchronous action of multiple agents. To solve the above problems, this paper first introduces the main problem of the research: multi-objective multi-agent path finding with asynchronous action, and proposes the algorithm framework of multi-objective loose synchronous (MOLS) search. By combining A∗ and M∗, MO LS-A∗ and MO-LS-M∗ algorithms are respectively proposed. The completeness and optimality of the algorithm are proved, and a series of comparative experiments are designed to analyze the factors affecting the performance of the algorithm, verifying that the proposed MO-LS-M∗ algorithm has certain advantages. Reference | Related Articles | Metrics | Comments（0）

Select

Multi-AGVs Scheduling with Vehicle Conflict Consideration in Ship Outfitting Items Warehouse DONG Dejin1,2 (董德金), DONG Shiyin3 (董诗音), ZHANG Lulu1,2 (章露露), CAI Yunze1,2∗ (蔡云泽) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 725-736.   DOI: 10.1007/s12204-024-2731-2 Abstract （162）      PDF（pc） （1452KB）（44）       Knowledge map    Save The path planning problem of complex wild environment with multiple elements still poses challenges. This paper designs an algorithm that integrates global and local planning to apply to the wild environmental path planning. The modeling process of wild environment map is designed. Three optimization strategies are designed to improve the A-Star in overcoming the problems of touching the edge of obstacles, redundant nodes and twisting paths. A new weighted cost function is designed to achieve different planning modes. Furthermore, the improved dynamic window approach (DWA) is designed to avoid local optimality and improve time efficiency compare to traditional DWA. For the necessary path re-planning of wild environment, the improved A-Star is integrated with the improved DWA to solve re-planning problem of unknown and moving obstacles in wild environment with multiple elements. The improved fusion algorithm effectively solves problems and consumes less time, and the simulation results verify the effectiveness of improved algorithms above. Reference | Related Articles | Metrics | Comments（0）

Select

AlgoTime-Varying Formation-Containment Tracking Control for Unmanned Aerial Vehicle Swarm Systems with Switching Topologies and a Non-Cooperative Target WU Xiaojing∗(武晓晶), CAO Tongyao (曹童瑶), ZHEN Ran (甄然), LI Zhijie (李志杰) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 689-701.   DOI: 10.1007/s12204-024-2728-x Abstract （156）      PDF（pc） （1627KB）（62）       Knowledge map    Save This paper studies the time-varying formation-containment tracking control problems for unmanned aerial vehicle (UAV) swarm systems with switching topologies and a non-cooperative target, where the UAV swarm systems consist of one tracking-leader, several formation-leaders, and followers. The formation-leaders are required to accomplish a predefined time-varying formation and track the desired trajectory of the tracking-leader, and the states of the followers should converge to the convex hull spanned by those of the formation-leaders. First, a formation-containment tracking protocol is proposed with the neighboring relative information, and the feasibilit condition for formation-containment tracking and the algebraic Riccati equation are given. Then, the stability of the control system with the designed control protocol is proved by constructing a reasonable Lyapunov function. Finally, the simulation examples are applied to verify the effectiveness of the theoretical results. The simulation results show that both the formation tracking error and the containment error are convergent, so the system can complete the formation containment tracking control well. In the actual battlefield, combat UAVs need to chase and attack hostile UAVs, but sometimes when multiple UAVs work together for military interception, formationcontainment tracking control will occur. Reference | Related Articles | Metrics | Comments（0）

Select

Comfort of Autonomous Vehicles Incorporating Quantitative Indices for Passenger Feeling PENG Shiwei1 (彭诗玮), ZHANG Xi1∗ (张希), ZHU Wangwang1 (朱旺旺), DOU Rui2 (窦瑞) J Shanghai Jiaotong Univ Sci    2024, 29 (6): 1063-1070.   DOI: 10.1007/s12204-022-2531-5 Abstract （156）      PDF（pc） （659KB）（19）       Knowledge map    Save At present, most of the studies on autonomous vehicles mainly focus on improving driving safety and efficiency, while less consideration is given to the comfort of passengers. Therefore, in order to gain and optimize quantitative indices for the ride experience of autonomous vehicles, this paper proposes an evaluation method for the correlation between driving behavior and passenger comfort with bidirectional long short-term memory network and attention mechanism. By collecting subjective feeling scores of passengers under different driving styles, and measuring the pressure level with skin conductance response and heart rate variability, the comprehensive quantitative indices of passenger comfort caused by driving behavior are evaluated. Based on this, a personalized comfort evaluation model for passengers with different driving style preferences is established. The results obtained from experiments in open road and closed test areas have validated the effectiveness and feasibility of the method proposed in this paper. Reference | Related Articles | Metrics | Comments（0）

Select

Iterative Model Predictive Control for Automatic Carrier Landing of Carrier-Based Aircrafts Under Complex Surroundings and Constraints ZHANG Xiaotian1(张啸天), HE Defeng1* (何德峰), LIAO Fei2 (廖飞) J Shanghai Jiaotong Univ Sci    2024, 29 (4): 712-724.   DOI: 10.1007/s12204-023-2690-z Abstract （152）      PDF（pc） （1363KB）（43）       Knowledge map    Save This paper considers the automatic carrier landing problem of carrier-based aircrafts subjected to constraints, deck motion, measurement noises, and unknown disturbances. The iterative model predictive control (MPC) strategy with constraints is proposed for automatic landing control of the aircraft. First, the long shortterm memory (LSTM) neural network is used to calculate the adaptive reference trajectories of the aircraft. Then the Sage-Husa adaptive Kalman filter and the disturbance observer are introduced to design the composite compensator. Second, an iterative optimization algorithm is presented to fast solve the receding horizon optimal control problem of MPC based on the Lagrange’s theory. Moreover, some sufficient conditions are derived to guarantee the stability of the landing system in a closed loop with the MPC. Finally, the simulation results of F/A-18A aircraft show that compared with the conventional MPC, the presented MPC strategy improves the computational efficiency by nearly 56% and satisfies the control performance requirements of carrier landing. Reference | Related Articles | Metrics | Comments（0）

Select

Predictive Simulation of External Truck Operation Time in a Container Terminal Based on Traffic Big Data DU Ye1 (杜晔), ZHAO Yifei2 (赵一飞), GAO Deyi1 (高德毅) J Shanghai Jiaotong Univ Sci    2024, 29 (5): 801-808.   DOI: 10.1007/s12204-022-2415-8 Abstract （151）      PDF（pc） （620KB）（17）       Knowledge map    Save The operation time of external trucks in a container terminal is one of port operation key performance indicators concerned by port operators, external truck operators and related government authorities. With the traffic big data combined with the operation characteristics of the container terminal, the system dynamics method is used to build the simulation model of the operation system for external trucks. The simulation results of the operation time of external trucks are consistent with the actual situation, which provides an effective way to eliminate the “black box” of the operation time of the external trucks. The model can also be applied in multiple scenarios by using the traffic big data, and the simulation results can be adopted by the relevant organizations. Reference | Related Articles | Metrics | Comments（0）

Select

Experimental Investigation of Possibility of Simultaneously Monitoring Lung Perfusion/Cardiomotility and Ventilation via Thoracic Impedance Measurement BAI Zixuan 1(白子轩), MA Yixin1，2∗(马艺馨), KONG Zhibin3(孔志斌),XUE Shan4 (薛珊) J Shanghai Jiaotong Univ Sci    2025, 30 (1): 81-90.   DOI: 10.1007/s12204-023-2639-2 Abstract （150）      PDF（pc） （808KB）（5）       Knowledge map    Save Impedance pneumography has a significant advantage for continuous and noninvasive monitoring of respiration, compared with conventional flowmeter-based ventilation measurement technologies. While thoracic impedance is sensitive to pulmonary ventilation, it is also sensitive to physiological activities such as blood flow and cardiomotility, in addition, body movement/posture. This paper explores the possibility of simultaneously monitoring pulmonary ventilation, blood circulation and cardiomotility by bioimpedance measurement. Respiratory, blood perfusion and cardiomotility signals are extracted using the wavelet method from thoracic impedance data measured in breath-holding and tidal breathing statuses, to investigate signal strength and their dependency. This research provides a foundation for the development of bedside devices to monitor various physiological activities. Reference | Related Articles | Metrics | Comments（0）