Fatigue failure is one of the most important reasons for the failure of engineering application components. However, due to the high cost of fatigue experiments, it is necessary to use mechanical properties to predict fatigue strength. Based on the true stress-strain curves, a novel model for fatigue strength prediction is established and is used to calculate the fatigue strength. The strength predicted is compared with that calculated by the staircase method and the Basquin equation. The results show that the model could obtain the fatigue strength of the materials only by using tensile strength and work-hardening strength, and it is suitable for other steels, which greatly saves costs and increases accuracy.
According to its actual structure, the steam generator is divided into the hot leg, the cold leg, the steam-water separator, and the steam chamber, where the hot leg and the cold leg are further divided into a descending channel and an ascending channel. Based on the mass, energy, and momentum conservation laws, the distribution parameter models for the hot leg, the cold leg, the steam-water separator, and the steam chamber are established. Combining the physical parameter database of the working medium and the data collection system (DCS) real-time measurement data, the model is solved in real time and the estimated mass flow rate of recirculation water is obtained. The verification results using the operating data of a CPR1000 nuclear power unit show that the calculated temperature, the pressure, and the mass flow rate of the outlet saturated steam of the steam generator agree well with the measured values, which indirectly verifies the correctness of the estimation approach of recirculation water proposed in this paper.
To clarify the transient flow characteristics of a high-speed aero-fuel centrifugal pump in variable gas-liquid ratio conditions, numerical simulations for the internal flow field in design flow rate and small flow rate conditions are conducted, focusing on the transient flow characteristics and time-frequency performance of pressure pulsation in the impeller channel. The conversion relationship between gas-liquid ratio and inlet pressure is given to determine the inlet simulation boundary, and then the grid model and length of time step are checked for relevant test. The prediction results between simulations and test are given to verify the effectiveness of the adopted simulation method. Then, the transient characteristics are analyzed through the results of pressure contour and turbulent kinetic energy, and the time-frequency performances of pressure pulsations at impeller inlet and outlet are conducted by fast Fourier transform(FFT). The results show that the flow in the impeller channel is relatively stable under the fuel saturation condition, and the main frequency of pressure amplitude is rotation frequency. With the increase of gas-liquid ratio, the impeller inlet produces a low-pressure zone whose area is significantly enlarged. Besides, a certain wake flow zone is generated at impeller outlet, where the turbulent energy dissipation rate is also demonstrated to be the strongest at these zones. Moreover, the inlet pressure is generally decreased with the increase of gas-liquid ratio, and the main frequency at the design flow rate is rotation frequency, but other frequency multiplication appears at the small flow rate. Meanwhile, the wake flow at the impeller outlet does not seriously affect the main frequencies at the monitoring points, where the main frequency is still rotation frequency.
The cutting performance of the porous polyimide material assisted by cold air trace lubrication was investigated, and the influence law of cutting process parameters on the cutting effect (cutting force, surface finish quality, and oil content) of the porous material was explored. The results show that compared with dry cutting and low-temperature cold air cutting, the cutting temperature and the machined surface roughness of cold air micro-lubrication cutting are the lowest, and the damage to the porous runner is the smallest. The depth of cut and the feed volume are the factors that have the greatest influence on milling force and surface roughness, respectively. Defects such as chip burrs and tears are the main factors that lead to the increase in surface roughness. The existence of drawing, micro-crack, stacking, and tiny debris on the workpiece surface are the main causes for the decrease in the oil content and oil delivery rate of the material. The optimal machining parameters of the porous polyimide material in the orthogonal experiment with the aid of cold air trace lubrication are vc=(100±2) mm/min, fz=(0.3±0.01) mm/r, ap=(0.8±0.1) mm (vc is the cutting speed, fz is the feeding rate, and ap is the cutting depth). With the assist of the cryogenic minimal quantity lubrication technology, the low damage processing, high oil content, and high oil delivery rate of cage products can be obtained.
An experimental study is conducted to find the characteristics of steam plume and pressure oscillation on direct contact condensation by a side-hole sparger. Synchronal measuring of transient pressure of the steam plume is gained from the high-speed camera and high-pressure sensor respectively. The influence of steam mass flux and water temperature on direct contact condensation characteristic are presented and its regime map is plotted. Then, the dynamic connections of transient pressure and steam plume in different condensation regimes are analyzed. It is found that high frequency pressure oscillation and the collapse of detached bubbles occur at the same time. Together with the condensing and disappearing process of the collapse of detached bubbles, the intensity of pressure oscillation decays exponentially in a vibration way. The changing trends of steam plume length in condensation oscillation regime and stable condensation regime are also obtained, which shows that the steam plume increases with the steam mass flux and the temperature in the condensation oscillation regime. When entering the stable condensation regime, the steam plume suddenly decreases and then increases with the temperature and the steam mass flux. The research results are useful for the engineering application of sparger in steam emission devices.
A full height sliced test facility is established based on the modeling process. The heating test section is manufactured by adopting the explosive welding technology, and the surface of carbon steel can simulate the actual surface characteristic of a pressurized reactor. Critical heat flux tests on flow, pressure, subcooling degree, chemical material in water, and surface characteristics are performed on the outer surface of the pressurized reactor vessel. Therefore, the law of heat transfer limit on the outer surface of the CAP1400 pressurized reactor vessel is obtained, and the process of flow and heat transfer for in-vessel retention is validated.
The discharge characteristics of nanosecond pulse discharge at different pulse voltages and different pulse intervals are investigated based on constant volume bomb. The effects of different discharge parameters on the flame success rate and ignition delay of methane/air mixture are compared. The results show that the coupling between pulses can be reasonably utilized and the utilization rate of discharge energy can be improved by decreasing the pulse interval.The ignition limit of methane/air mixture can be enlarged by increasing the pulse voltage, the pulse number, and decreasing the pulse interval. The ignition delay can be reduced by increasing the pulse voltage and decreasing the pulse interval.
The Sydney partially premixed flame FJ200-5GP-Lr75-103 case with the local extinction phenomenon is studied by large eddy simulation (LES) coupled with the dynamic k equation sub-grid scale model and the dynamic thickened flame (DTF) combustion model. To account for the influence of non-uniform equivalence ratio on the laminar flame speed and thickness, two fitting functions are introduced in the DTF model to automatically adjust these two parameters in the wrinkling functions, according to the local value of mixture fraction in the flow field. The results show that the dynamic k equation model can predict the mixture fraction of the non-uniform premixed gas in the flow field well. In the middle and downstream with more flame extinction, the temperature, the species profiles, and the scatter plot of temperature versus mixture fraction calculated by LES agree well with the experiment data. This demonstrates that the improved DTF model can capture the local extinction phenomenon in partially premixed flame, but for the CO mass fraction, discrepancy between the LES results and experiment data is presented.
The profile flow around the wind turbine airfoil has a significant impact on the overall performance and the noise of a wind turbine. In this paper, a combination method of large eddy simulation (LES) and the solution of Ffowcs Williams and Hawkings (FW-H) equation is used to obtain the unsteady flow field and the far-field aeroacoustic noise of a wind turbine airfoil. The dominant modes of the unsteady vorticity at the angle of attack of 8° are extracted by the proper orthogonal decomposition (POD). The modal structure of the dominant modes indicates that the trailing edge vortices and the laminar separation bubble are the main unsteady feature of the airfoil flow. The modal structures most related to the far-field aeracoustics noise are obtained by the extend proper orthogonal decomposition (EPOD). The modal results show that the vortices in the boundary layer and the large-scale vortices in the wake are most relevant to the generation and propagation of the aeracoustics noise.
Surface defects refer to the local undulation of the parts in the forming process. The variable cross-section and variable-curvature aluminum trim parts are prone to local surface defects during the stretch-bending process, which has a great impact on the appearance of the whole vehicle. Research on the formation mechanism of surface defects has important guiding significance for the prevention and control of surface defects. In this paper, the formation mechanism of the deformation defects of the stretch-bending forming surface of aluminum trim is studied, the stretch-bending forming test and the surface defects measurement are conducted, and the simulation model of the stretch-bending forming of aluminum trim is established by using the Abaqus software platform. In addition, the forming process of surface defects is analyzed. The results show that when the ratio of section width to bending radius is larger than 35%, the part will have a warping tendency, and the joint action of section shape, upper die, and padding will cause local surface defects.
Flow spinning is an advantageous technology in forming thin-walled cylindrical parts, but for cylindrical parts with inner ribs,the height of the inner ribs that can be formed is limited. Aimed at the problem that the inner ribs are difficult to fill in such components, a numerical simulation model of staggered spinning of thin-walled cylindrical parts with cross inner ribs is established based on Abaqus, and the influence of the number of rollers on the filling of inner ribs is analyzed. The process test of the cylinder part with orthogonal inner ribs is conducted. The results show that in the staggered spinning process, increasing the number of rollers will reduce the diameter expansion of the aluminum alloy cylindrical part with inner ribs, and inhibit the axial flow of materials, leading to a higher geometric accuracy of the inner ribs. The comparison of the simulation results and the test results shows that the average error of the inner rib filling fullness of simulation result is 11.3%, and the simulation model has a good reliability.
In the multi-variety and small batch manufacturing workshop, digital twin model is mostly established for specific scenarios. Due to its lack of adaptive ability under working conditions, the prediction accuracy of machining quality is insufficient. To solve this problem, an adaptive transferring method of the digital twin model is proposed. By building the transferable digital twin model, the online prediction of machining quality based on the fusion of mechanism and algorithm model is realized. The transferring process and strategy of the digital twin model are proposed. Based on the analysis and calculation of characteristic data, the source model to be transferred is selected. At the same time, in combination with the transfer learning theory, the transfer of digital twin models is realized under simple and complex changing conditions. Taking drilling as an example, the drilling experiment platform is built and the feasibility of digital twin model transfer is verified. The results show that the model can keep the mean absolute error of prediction less than 1.5% under changing working conditions. This method provides a new idea to improve the adaptive ability of digital twin models.
In view of the low efficient of virtual human upper limb simulation in the virtual assembly environment and the inaccurate evaluation of ergonomics, this paper analyzes hand assembly action, defines hand joint structure, arm and hand size, establishes a parametric assembly gesture model, and forms a gesture library. The fuzzy algorithm is used to improve the rapid upper limb assessment (RULA) method. The trapezoidal function is used to optimize the evaluation score when the joint angle is at the critical value. The final evaluation result is obtained through the rule base. Strain index(SI) is used to evaluate the risk of musculoskeletal operations such as hands and wrists, and the comprehensive score is weighted with fuzzy RULA evaluation. The ergonomics of the virtual hand assembly process is continuously evaluated to capture the risk posture in the assembly operation. Finally, the above methods are integrated based on the 3D Experience platform and verified by the assembly of mobile communication components.
To give full play to the advantages of pure electric drive, and aimed at the transmission system of electric loader, the low efficiency hydraulic torque converter and reverse direction clutch of traditional models are cancelled. By analyzing the shift law of the pure electric drive system, the electro-hydraulic shift control system is used to control the wet clutch, and based on the feedback of pressure and speed, combined with the driving motor active working in the speed and torque mode, the matching of torque and speed in the process of clutch engagement and disengagement is realized. For a 50-type pure electric drive loader, a shift control strategy of pure electric drive loader based on composite control of drive motor and clutch pressure is proposed. The result of vehicle test shows that the proposed control strategy can give full play to the advantages of pure electric drive. The shift time is reduced by about 50%, the sliding friction work is greatly reduced, and the maximum shift impact is 14.08 m/s3, which is within the recommended limit value of 17.64 m/s3 for Chinese vehicles.
Aimed at the problems that the proportion and position of different fault characteristics of equipment components under variable working conditions and variable load in the signal are not fixed, and include the multi-scale complexity of the original vibration signal in a large number of different scenarios, a convolutional recurrent neural network (CRNN) rolling bearing fault diagnosis method based on feature pyramid network (FPN) was proposed. Using the convolution neural network (CNN) framework, the convolution layer of CNN and the long and short-term memory (LSTM) layer of recurrent neural network (RNN) were connected in parallel to form a new CRNN, so as to make full use of the learning ability of CNN to spatial domain information and RNN to time domain information. The weights were shared in each layer to reduce network parameters. A novel feature map was constructed using FPN, and one-dimensional signal and two-dimensional signal formed after stacking were input to extract the feature of the signal collected by the sensor, and realize fault diagnosis. The average diagnostic accuracy of this method is 99.20%, which is at least 3.62% higher than that of the basic neural network model, indicating that this method has a high diagnostic accuracy and a strong robustness. Using the bearing data set of Case Western Reserve University, it is proved that the method has a good universality. The t-SNE method was used to visually analyze the feature learning effect of the model. The results show that different fault category features have good clustering effect.
In order to accurately predict the stability of orthogonal cutting of cylindrical workpiece, a nonlinear orthogonal cutting system model is established, which includes the nonlinear stiffness caused by the surface wave of work as well as the deformation of the tool and work. The multi-scale method is used to solve the system. The effect of machining parameters and system parameters on the stability of the primary resonance and 1/2 subresonance is analyzed to gain the overall stability cloud map compared with the lobe diagrams of linear approximation system. The results show that the instability of primary resonance, 1/2, 1/3, and 1/4 subresonance occur in the orthogonal cutting system with the quadratic nonlinearity and cubic nonlinearities stiffness, which makes the system have period-doubling, quasi-periodic, and chaotic operation behavior. The comparison indicates that the dynamics model of nonlinear orthogonal cutting can accurately predict the stability of the system.
This paper proposes an adaptive migrating birds optimization (AMBO) method based on variable neighborhood search to solve the inequal lot streaming hybrid flowshop scheduling problem (ILS-HFSP) for a 2+1+1 hybrid flowshop, which realizes multi-objective optimization of minimizing makespan and minimum average work in process. Compared with the original migrating birds optimization, the AMBO algorithm adopts the variable neighborhood search strategy with an adaptive selection probability of neighborhood operator that is adaptively adjusted with the number of iterations. Besides, a time-window operator is adopted to improve the search performance of exchange operators and convergence rate. Several orders of different scales generated randomly are studied, and the results show that the AMBO algorithm has a higher solution quality and a better convergence performance than the migrating birds optimization algorithm and the genetic algorithm, thereby verifying the effectiveness of the AMBO algorithm.
This paper uses a single fiber bragg grating (FBG) sensor to implement an experiment to measure vibration and temperature signals at the same time, and proposes a MATLAB-based decoupling method to separate vibration and temperature signals. The experimental results show that under the condition of single signal measurement, the static temperature measurement error of the FBG sensor is within ±0.4 ℃ and the relative error of the dynamic measurement of the main frequency of vibration is 0.5%. The FBG sensor measures the composite signal of vibration and temperature. The relative error of the main vibration frequency obtained by the decoupling method proposed in this experiment is 0.65%, the relative error of the vibration amplitude is 7.14%, and the temperature signal error is within ±3.3 ℃.
Traditional similarity-based methods generally ignore the diversity of equipment fault modes, the difference in degradation rates, and the inconsistency among monitoring data lengths. Thus, a similarity-based multi-scale ensemble method in multiple fault modes (MFM-MSEN) is proposed to improve remaining useful life (RUL) prediction accuracy and characterize prediction uncertainty. By training the fault mode classification model, designing the time-series weighted prediction strategy, and recognizing the fault mode of equipment, the test equipment is matched with the training equipment with the same fault mode to reduce matching complexity, based on which, a multi-scale ensemble strategy is proposed to overcome the data utilization limitation caused by single-scale matching methods and enhance the generalization ability of the proposed MFM-MSEN method. This strategy matches the similarities between test equipment and training equipment at multiple time scales, and then multiscale prediction results are integrated to fit accurate RUL probability distribution by employing kernel density estimation. Experimental results demonstrate the superiority of the proposed MFM-MSEN method in dealing with the differences in equipment degradation.
A novel method for conceptual design of a mechatronics system was proposed, and the design framework of “requirement-environment-conflict-original understanding-solution” was established. The contradiction between design requirements and environmental constraints was identified by using the method of environment-based design (EBD), and the requirements and environment were iteratively analyzed to decouple complex and abstract issues. According to the original understanding of the theory of inventive problem solving (TRIZ), the product solution for specific needs was obtained, which ensured the directionality of innovation, the systematic generation of product design solutions, and the standardization of the design process. The friction and wear test machine of the aeroengine variable stator vane (VSV) shaft bushing system shows that the proposed design method is reasonable and effective.
The non-uniform stiffness distribution of irregular large thin-walled structures leads to warping deformation in the process of machining and assembly. The structure modeling and accurate calculation of the large deformation is the key to predict and control the deviation. A novel irregular higher-order composite shell element based on the absolute nodal coordinate formulation (ANCF) was proposed to accurately calculate the deformation of irregular large thin-walled structures with higher-order slope coordinates and improved shape function, considering the continuity of elements at the contact interface. A discrete method of irregular large thin-walled structures was proposed to solve the problem of node connection and deformation coupling between elements with different thicknesses, and to describe the large deformation of irregular large thin-walled structures effectively by using the combination of irregular higher-order elements and composite elements. An assembly deformation calculation model of irregular large thin-walled structures was proposed based on continuum mechanics. The influence of the thickened area on assembly deviation and stiffness of scalloped segment structures of heavy rocket tank bottom was studied with the new elements and calculation model, so as to provide guidance for structure design.
A dynamic model predictive control method for driving robots is proposed to realize accurate steering control of the test vehicle. First, the coupling dynamics model of the driving robot and the controlled vehicle is established, and the controllability of the coupling model is judged. Then, the Kalman filter is used to estimate the state of the coupled model, and a model predictive controller is designed according to the estimated state. The least square method is adapted to fit the nonlinear relationship between path curvature and prediction horizon, and a dynamic model predictive controller with variable prediction horizon is designed. Finally, the simulation and the test of the steering control of the driving robot at different conditions are conducted, and the results verify the effectiveness of the proposed method.
Aimed at improving the prediction accuracy of engine emissions under driving conditions which are not covered by the training set, an extrapolated Gaussian process regression (GPR) method is proposed. First, the training set data is fed to the GPR model for pre-training, and then a wide-area input set is constructed by uniform sampling between plus/minus three standard deviations, and the hyperparameters are optimized for the goal of minimizing the prediction variance of the input set. The test results on a direct injection gasoline engine show that the mean absolute error of emission prediction using the extrapolated GPR is 0.53411, which is 24.27% lower than that of using the traditional GPR and 36.32% lower than that of using the back propagation (BP) neural network, which signifies the effectiveness of the proposed method in terms of reducing test costs and improving the accuracy of emission prediction during real driving.
In order to analyze the flow characteristics of the internal flow field in the working chamber of the oil-free double-warp scroll compressor, a three-dimensional unsteady numerical simulation of the internal flow field in the working chamber of the compressor is conducted based on the computational fluid dynamics (CFD) method. In order to verify the deviation between the numerical simulation and the actual working process, a test platform for scroll compressors using air as the working medium is built, and the actual output performance parameters of the oil-free double-warp scroll compressor are obtained. The results show that the mass exchange between adjacent working chambers has a greater impact on the working chamber and the distribution of temperature and velocity along the tooth height direction, but has a little impact on pressure. The inlet and outlet mass flow and velocity rate change periodically, and as the compressor speed increases, the theoretical volume flow increases.
The blow-off limits of methane and hydrogen-mixed gas on the Sydney inhomogeneous inflow jet burner were calculated and studied by large eddy simulation. The GRI 3.0 detailed reaction mechanism, 28-step and 19-step methane simplified reaction mechanisms were used to calculate the laminar premixed flames under the condition of different hydrogen-mixed fuel gases, which demonstrates the good performance of the 19-step simplified mechanism. A dynamic thickening flame combustion model was combined with the 19-step reaction mechanism to compute the Sydney partially premixed jet flame burner with a hydrogen-mixed gas (the volume ratio of V(H2):V(CH4):V(CO):V(CO2) = 0.2:0.2:0.27:0.33). The blow-off limits in FA and FJ layouts of hydrogen-mixed gas are 90 m/s and 109 m/s, while for pure methane flames are 74 m/s and 128 m/s, respectively. The difference of blow-off limits between the two fuels is related to the inhomogeneity of fuel and air mixing in different layouts. This study shows that the combustion stability can be improved by optimizing the inlet layout and mixing process of fuel gas and air.
Aimed at the buckling problem of deep implantation of flexible neural probe, a grooved cross-section silicon needle is designed as an auxiliary implant tool to provide temporary stiffness for probe implantation. In order to quantitatively evaluate the comprehensive performance of auxiliary tools, combining critical buckling force and cross-section area, and considering mechanical and biological properties, the performance evaluation index of auxiliary tools is proposed. Based on this evaluation index, the optimal groove depth ratio and groove width ratio of the grooved cross-section silicon needle are studied. The best groove depth ratio is the maximum value within the process requirements while the best groove width ratio increases with the thickness of the silicon needle. Moreover, the performance evaluation index is used to quantitatively prove that the grooved cross-section silicon needle has obvious performance advantages over the traditional circular and rectangular cross-section silicon needles. The simulation design of the grooved cross-section silicon needle is conducive to screening the best parameter combination of the cross-section, reducing the number of processing and the cost of experiments.
In multilayer ultrasonic welding, the sonotrode presses on metal sheets and drives the sheets to produce ultrasonic vibration. Then, each contact interface generates heat, produces plastic deformation, and forms solid-state bonding. However, the load distributes unevenly in each sheet, which results in the uneven friction state and inconsistent welding quality. Hence, it is necessary to reveal the mechanism of frictional state and energy dissipation based on the load distribution in each sheet. A finite element model of 5-layer copper sheets is established using Abaqus and considering the Cattaneo-Mindlin contact theory. The clamping force and ultrasonic vibration in each interface is simulated. The slip-stick state of each interface is obtained and the effect of the clamping force are analyzed. The frictional energy dissipation and proportion of each interface are calculated, the influencing factors of frictional energy dissipation are summarized, and the optimization of input clamping force is discovered, which can provide theoretical guidance for the improvement of multilayer ultrasonic welding.
In the process of high speed gas metal arc welding (GMAW), large box girder has many complex working conditions, such as positioning weld, low assembly accuracy, difficulty in strictly ensuring the pose of workpiece through tooling, real-time change of 3D pose of weld, etc. The vision based weld recognition method has a large amount of calculation and is not aimed at the workpiece with positioning weld, making it difficult to obtain the 3D pose of large box girders quickly. Aimed at this problem, a fast 3D pose estimation method for large box girder based on laser displacement sensing and point cloud clustering was proposed. Using this method, the vertical plane and flat plane of the welding seam of large box girders were obtained by fast segmentation of point cloud. Then, the pose information of welding seam was calculated. A pose information estimation test was conducted for welds with different poses. The results show that when the welding speed is up to 1200 mm/min, the pose error of the weld is less than 0.25 mm and 1.8°respectively. The robustness of the automatic welding of large box girders to complex conditions of positioning weld and low assembly accuracy is enhanced, which greatly improves the welding quality.
A 5-degree-of-freedom continuum manipulator is designed and implemented to improve the flexible positioning ability of the continuum mechanism in applications such as interactivity operations, light object grabbing, and human-machine collaboration. In the design process, by introducing a two-segment constant curvature model, the distribution of degrees of freedom on the mechanism is explained geometrically by the method of twist. Coupling two stretchable parallel modules in series, a two-segment structure is formed for curving and each segment has two degrees of freedom in bending and one degree of freedom in stretching, thereby giving 5 degrees of freedom to the end-effector. Concentrating on the dynamic performance of the manipulator, an electromechanical system platform is built as a prototype. The experiments show that the structure realizes the control of the end-effector during large deformation of the manipulator, and is able to achieve extreme pose in 2 s with an approximation positioning error of 2% of the nominal arm length.
Aimed at the problem of generating a smooth, safe, and dynamically feasible continuous-time trajectory for unmanned aerial vehicles (UAV) in complex environments, a trajectory planning algorithm is proposed to minimize a multi-objective function based on safe flight corridors. The safe flight corridor represented by a collection of convex polyhedra is built based on the initial discrete waypoints generated by the improved rapidly-exploring random tree(RRT), namely the RRT* algorithm. The safety objective function is established according to the constraints of limiting the trajectory inside safe flight corridors. In combination with the flight smoothness, dynamic characteristics, and time performance, a multi-objective function is built. The gradient-based convex optimization algorithm is used to derive the continuous-time trajectory expressed as a piece-wise polynomial by optimizing the position, velocity, acceleration of waypoints, and time allocation. The effectiveness and performance of the proposed algorithm is tested and compared under complex environments such as the coal mine. The test results demonstrate that the proposed algorithm has a better comprehensive performance in comparison with existing algorithms.
In order to study the flow and heat transfer performance of wedge-shaped latticework channels in the turbine blade trailing edge, this paper conducted an experimental study by employing the transient liquid crystal (TLC) technique to investigate the local heat transfer characteristics of the upper and lower main surfaces and applying the pressure scanning valve to mesure the pressure loss of the channels at different Reynolds numbers. The experiment shows that there is a significant difference between the upper and lower main surfaces under the turning flow configuration condition at the trailing edge section. The average Nusselt number of the lower main surface is over 30% higher than that of the upper main surface. In heat transfer coefficient, the wedge-shaped latticework channel is over 46% higher than that of the needle rib channel. There is a strong mass exchange at the interface between the upper and lower channels of the latticework channel. The intermittent high heat transfer areas on the upper and lower main surfaces are corresponding to the interface. As the inlet Reynolds number increases, the channel pressure drop increases rapidly. The pressure drop of the wedge-shaped latticework channel is 5 to 7 times that of the needle ribs, but the heat transfer area of latticework channel is 107.4% higher than the needle ribs channel, and the overall thermal performance of the wedge-shaped latticework channel is still approximately 66% higher than that of the needle ribs channel.
Aimed at the shortcomings of existing methods in identifying evidence conflicts, in combination with traditional evidence conflicts, an evidence conflict identification and adjustment method is proposed. First, the weighted average of traditional evidence conflicts is used as the identification index in this method. Then, the credibility and uncertainty are introduced to determine the weight coefficients in two situations based on whether it is a conflict between different evidences. After that, the identified evidence conflicts are adjusted in combination with the traditional evidence conflicts and Jousselme information distance. This method is based on the traditional evidence conflicts, which ensures the representativeness of the identification index and the authenticity of the identification results. In addition, the weight coefficients comprehensively consider the credibility within and between the evidence, which is also more representative. Finally, the reduction degree and the average deviation are selected to analysis and verify the proposed method. The results show that the proposed method has a high recognition accuracy and can effectively adjust evidence conflicts.
The orifice plate has the advantages of convenient construction, high energy dissipation efficiency, and economy, and is widely used in hydropower engineering and discharge engineering. The recovery length of the flow pressure behind the orifice plate is related to the distance arrangement between orifice plates in multi-stage orifice plate design. However, little research has been conducted on the pressure recovery length behind the orifice plate. The numerical simulation method is used to quantitatively study the influencing factors of the pressure recovery length behind the orifice plate, and the results show that the recovery length of the flow pressure behind the orifice plate is mainly dominated by the aperture ratio of orifice plate, the relative thickness of the orifice plate, and the Reynolds number of fluids flowing through orifice plate. As the Reynolds number increases to a fixed value, the pressure recovery length behind the orifice plate does not change. When the relative thickness of the aperture ratio of the orifice plate are within a certain range, the relative thickness of the orifice plate will not change, the aperture ratio of the orifice plate decreases, and the pressure recovery length behind the orifice plate increases. As the diameter ratio of orifice plate remains unchanged, the relative thickness of the orifice plate increases, and the pressure recovery length behind the orifice plate decreases. The empirical formula for the pressure recovery length behind the orifice plate is obtained by multivariate regression analysis and curve fitting, and the indoor test verification is adopted, which provide theoretical basis for distance arrangement between orifice plates in multi-stage orifice plate design.
In order to accurately reflect the operation characteristics of office buildings, a convolutional neural network(CNN)-recurrent neural network(RNN)combined model for energy consumption prediction of office buildings is proposed by using the good feature extraction ability of CNN and the good time series learning ability of RNN. Besides, a two-dimensional matrix data input structure suitable for the deep learning model is designed. The case study results show that compared with the simple recurrent neural network and long short term memory network, both the prediction accuracy and computational efficiency of CNN-RNN combined model are significantly improved, and the generalization of the model is also good.
In the production process of aerospace structural parts, there coexist batch production tasks and research and development (R&D) tasks. Personalized small-batch R&D and production tasks lead to frequent emergency insertion orders. In order to ensure that the task is completed on schedule and to solve the flexible job shop dynamic scheduling problem, this paper takes minimization of equipment average load and total completion time as optimization goals, and proposes a dual-loop deep Q network (DL-DQN) method driven by a perception-cognition dual system. Based on the knowledge graph, the perception system realizes the representation of workshop knowledge and the generation of multi-dimensional information matrix. The cognitive system abstracts the scheduling process into two stages: resource allocation agent and process sequencing agent, corresponding to two optimization goals respectively. The workshop status matrix is designed to describe the problems and constraints. In scheduling decision, action instructions are introduced step by step. Finally, the reward function is designed to realize the evaluation of resource allocation decision and process sequence decision. Application of the proposed method in the aerospace shell processing of an aerospace institute and comparative analysis of different algorithms verify the superiority of the proposed method.
In order to improve operation efficiency of the electric multiple unit when running reliably, a multi-stage opportunistic maintenance decision strategy with bi-objective optimization is proposed based on the traditional opportunistic maintenance strategy. The window of opportunistic maintenance is equidistantly divided into multiple stages, and components located at different stages are maintained by different efforts. Aimed to solve problem of assigning multiple maintenance tasks of components, an assignment algorithm with two repairpersons is proposed. To further improve the multi-attribute optimization performance of this model, the proposed algorithm is introduced into the multi-attribute decision-making. The numerical example analysis verifies the effectiveness of the multi-stage opportunistic maintenance strategy in optimizing availability and average cost rate. Furthermore, the influence of weight factors on optimization tendency is discussed as well.
Due to the large degree of freedom of the human body and the working environment, the generation of virtual human working posture during simulation is very complicated, which requires a lot of time and energy. To solve this problem, according to the structure and motion characteristics of human body, the templates of assembly actions for virtual human are constructed. Through theoretical analysis and calculations, the formulas of the reachable region of different action templates are deduced, and the complete reachable region of the virtual human is established, based on which, feasible assembly actions are screened out through reachability analysis. In combination with the existing multi-objective solution model of virtual human working posture, the automatic generation of virtual human working posture is realized. Based on the above research, this method is verified by examples.
In the case of dynamic systems, the traditional kernel principal component analysis (KPCA) method does not perform well. The moving window kernel principal component analysis method can adapt to the normal parameter drift of dynamic systems, but it needs a longer computation time when processing large number of samples. Therefore, an adaptive process monitoring method for online reduced kernel principal component analysis is proposed. In this method, a small training set is selected as the initial reduced set in a large number of samples for modeling, and the online real-time collected data are analyzed to judge whether the new sample is normal or not. If it is a normal sample, the method judges whether the sample is added to the reduced set, and updates the online KPCA model automatically when adding to the reduced set. The proposed method is applied to a numerical example and the Tennessee-Eastman (TE) process. The simulation results show that the proposed method is effective and feasible.
To optimize the joint decisions of equipment maintenance and spare parts ordering with dual sourcing, a joint policy is proposed for a multi-unit parallel manufacturing system. The degradation trajectories of components are described by the Poisson process. The joint decisions are modeled according to the Markov decision process. Based on components and inventory status, the transition probability of a system is formulated. The value iteration algorithm is applied to obtain the optimal maintenance and spare parts ordering policy to minimize the average total cost of the system. Furthermore, to shorten the computation time, a heuristic strategy is developed according to the principle of sequential optimization. Numerical experiments with a two-component system are conducted to analyze the decision difference between the heuristic strategy and the optimal strategy. The sensitivity analysis shows that compared with the optimal strategy, the heuristic strategy can reduce the computation time when the total cost increases by no more than 5%.
In engineering, working environment and operating state are constantly changing, which decreases the correct rate of equipment fault diagnosis, resulting in the loss of time and cost. The structure of the deep belief network is investigated for the time-varying factors in the mechanical system. In combination with the signal decomposition technology of fixed learning step size, the original characteristics of the sensor data are retained. In addition, the deep key information of the signal is repeatedly extracted layer by layer. The data loss technology is integrated to optimize the network structure to avoid over-fitting problems. Further, considering the domain adaptive method in transfer learning, the memory characteristics of different levels of deep belief networks are solidified. Therefore, a domain adaptive deep belief network with shift-invariant features (SIF-DADBN) is proposed for rolling bearing fault diagnosis. By identifying the characteristic information of similar fault signals with variable working conditions, the accuracy and generalization of bearing intelligent fault diagnosis are both improved. Based on the public data set of rolling bearings, the average correct rate of the fault diagnosis technology is found to be as high as 95.65%. Compared with five other methods, the effectiveness and accuracy of SIF-DADBN under variable working conditions are verified.
