上海交通大学学报

上海交通大学学报 >

2021 , Vol. 55 >Issue 6: 774 - 780

DOI: https://doi.org/10.16183/j.cnki.jsjtu.2020.086

一种基于可解释神经网络模型的压缩机功率软测量方法

展开
  • 上海交通大学 机械与动力工程学院, 上海 200240
王煜林(1997-),男,辽宁省大连市人,硕士生,主要研究方向为动力系统数据驱动建模

收稿日期: 2020-04-03

  网络出版日期: 2021-06-08

基金资助

国家自然科学基金资助项目(51706132)

收起

A Compressor Power Soft-Sensing Method Based on Interpretable Neural Network Model

Expand
  • School of Mechanical Engineering, Shanghai Jiao Tong University, Shanghai 200240, China

Received date: 2020-04-03

  Online published: 2021-06-08

Fold

摘要

为在保证测量的准确性和高效性的同时,降低软测量方法对数据集的依赖性,提出一种基于可解释神经网络的压缩机功率软测量方法.实验中,在使用泛化性良好的数据集进行训练时,可解释神经网络模型在测试集上的均方根误差为 0.0094,相比反向传播(BP)神经网络模型降低了1.1%.在使用泛化性较差的数据集进行训练时,可解释神经网络模型在测试集上的均方根误差为 0.0128,相比BP神经网络模型降低了79.8%.实验结果表明,基于可解释神经网络的压缩机功率软测量方法不但具有较高的准确率,且在使用泛化性较差的数据集进行训练时,依然能够保持较高的测量性能.

关键词： 可解释性; 神经网络; 软测量方法; 压缩机

本文引用格式

王煜林, 周登极, 郝佳瑞, 黄大文 . 一种基于可解释神经网络模型的压缩机功率软测量方法[J]. 上海交通大学学报, 2021 , 55(6) : 774 -780 . DOI: 10.16183/j.cnki.jsjtu.2020.086

Abstract

In order to ensure the accuracy and efficiency of measurement, and reduce the dependence of the soft sensing on dataset, a soft-sensing method of compressor power based on interpretable neural network is proposed. When training on a dataset with good generalization in the experiment, the root mean squared error(RMSE) of the interpretable neural network model on the test set is 0.0094, which is 1.1% lower than that of the back propagation(BP) neural network model. When training on a dataset with poor generalization, the RMSE of the interpretable neural network model on the test set is 0.0128, which is 79.8% lower than that of the BP neural network model. The experimental results show that the soft-sensing method based on interpretable neural network not only has a high accuracy rate, but also can maintain a good measurement performance when training on a dataset with poor generalization.

Key words： interpretability; neural network; soft-sensing method; compressor

参考文献

[1] 于静江, 周春晖. 过程控制中的软测量技术[J]. 控制理论与应用, 1996, 13(2):137-144.
[1] YU Jingjiang, ZHOU Chunhui. Soft-sensing techniques in process control[J]. Control Theory & Applications, 1996, 13(2):137-144.
[2] 孔祥志, 杨会林, 孙铜生. 滑片式压缩机功率计算模型的研究[J]. 流体机械, 2005, 33(1):25-27.
[2] KONG Xiangzhi, YANG Huilin, SUN Tongsheng. Study on the calculating model of power of vane compressor[J]. Fluid Machinery, 2005, 33(1):25-27.
[3] 王晓燕, 焦卫东, 朱利民. 变工况运行的制冷压缩机功率消耗的定量关系式研究[J]. 压缩机技术, 2019(1):6-10.
[3] WANG Xiaoyan, JIAO Weidong, ZHU Limin. Study on quantitative relationship formula about refrigeration compressor power consumption based on variable condition running[J]. Compressor Technology, 2019(1):6-10.
[4] 吉文鹏, 杨慧中. 基于自适应等距映射算法的软测量建模[J]. 南京理工大学学报, 2019, 43(3):269-274.
[4] JI Wenpeng, YANG Huizhong. Soft sensor modeling based on adaptive Isomap algorithm[J]. Journal of Nanjing University of Science and Technology, 2019, 43(3):269-274.
[5] 廉小亲, 王俐伟, 安飒, 等. 基于SOM-RBF神经网络的COD软测量方法[J]. 化工学报, 2019, 70(9):3465-3472.
[5] LIAN Xiaoqin, WANG Liwei, AN Sa, et al. On soft sensor of chemical oxygen demand by SOM-RBF neural network[J]. CIESC Journal, 2019, 70(9):3465-3472.
[6] 徐敏, 俞金寿. 软测量技术[J]. 石油化工自动化, 1998(2):1-3.
[6] XU Min, YU Jinshou. Soft measurement technology[J]. =Automation in Petro-Chemical Industry, 1998(2):1-3.
[7] 王美琪, 陈恩利, 刘鹏飞, 等. 融合机理与数据的篦冷机温度软测量模型[J]. 仪器仪表学报, 2018, 39(6):182-188.
[7] WANG Meiqi, CHEN Enli, LIU Pengfei, et al. Soft-testing model for grate cooler temperature measurement with mechanism and data fusion[J]. Chinese Journal of Scientific Instrument, 2018, 39(6):182-188.
[8] RUMELHART D E, HINTON G E, WILLIAMS R J. Learning representations by back-propagating errors[J]. Nature, 1986, 323(6088):533-536.
[9] 倪志伟. BP网络中激活函数的深入研究[J]. 安徽大学学报(自然科学版), 1997, 21(3):48-51.
[9] NI Zhiwei. Deep study on activation function in BP network[J]. Journal of Anhui University (Natural Sciences) , 1997, 21(3):48-51.
[10] 张娜, 林汝谋, 蔡睿贤. 压气机特性通用数学表达式[J]. 工程热物理学报, 1996, 17(1):21-24.
[10] ZHANG Na, LIN Rumou, CAI Ruixian. General formulas for axial compressor performance estimation[J]. Journal of Engineering Thermophysics, 1996, 17(1):21-24.
[11] KINGMA D P, BA J. Adam: A method for stochastic optimization[EB/OL]. (2017-01-30) [2019-11-21]. https:∥arxiv.org/abs/1412.6980 .
[12] 张田, 潘尔顺. 基于时间序列分析的电容器退化模型[J]. 上海交通大学学报, 2019, 53(11):1316-1325.
[12] ZHANG Tian, PAN Ershun. Degradation modeling of capacitors based on time series analysis[J]. Journal of Shanghai Jiao Tong University, 2019, 53(11):1316-1325.
[13] 李军朋, 华长春, 关新平. 基于机理、数据和知识的大型高炉冶炼过程建模研究[J]. 上海交通大学学报, 2018, 52(10):1142-1154.
[13] LI Junpeng, HUA Changchun, GUAN Xinping. Modeling research for blast furnace smelting process based on smelting mechanism, operation data and expert knowledge[J]. Journal of Shanghai Jiao Tong University, 2018, 52(10):1142-1154.
[14] 胡晨星, 丁杰, 竺晓程, 等. 离心压气机蜗壳内非定常流场的动态模态分解[J]. 上海交通大学学报, 2018, 52(9):1044-1049.
[14] HU Chenxing, DING Jie, ZHU Xiaocheng, et al. Dynamic mode decomposition of unsteady flow filed in a volute of centrifugal compressor[J]. Journal of Shanghai Jiao Tong University, 2018, 52(9):1044-1049.
[15] 吴军, 黎国强, 吴超勇, 等. 数据驱动的滚动轴承性能衰退状态监测方法[J]. 上海交通大学学报, 2018, 52(5):538-544.
[15] WU Jun, LI Guoqiang, WU Chaoyong, et al. Data-driven performance degradation condition monitoring for rolling bearings[J]. Journal of Shanghai Jiao Tong University, 2018, 52(5):538-544.
[16] 陈进平, 张树生, 何卫平, 等. 基于驱动参数建模的可行更改路径搜索和优选方法[J]. 上海交通大学学报, 2017, 51(10):1220-1227.
[16] CHEN Jinping, ZHANG Shusheng, HE Weiping, et al. Feasible change path search and optimization method based on driving parameter modeling[J]. Journal of Shanghai Jiao Tong University, 2017, 51(10):1220-1227.
Options
文章导航

/

〈 〉