Fault Diagnosis of Mechanical Equipment Based on GA-SVR with Missing Data in Small Samples

doi:10.16182/j.issn1004731x.joss.20-0046

Abstract

Abstract: In view of the equipment fault diagnosis with small and missing sample data, a method of missing data filling based on support vector regression optimized by genetic algorithm is proposed to improve the accuracy of equipment fault diagnosis. The support vector regression optimized by genetic algorithm was trained by other data values of missing data, and univariate prediction results were obtained. The training set was reconstructed through correlation analysis, so as to obtain the multivariate prediction results. Dynamic weights were established to combine univariate prediction results and multivariate prediction results to fill in the missing data. The complete data is taken as the input, and the equipment fault is diagnosed by support vector machine. Example analysis shows that the method proposed in this paper has a high fault diagnosis accuracy.

Key words: fault diagnosis, small sample, missing data, support vector machine, dynamic weight, combination prediction

CLC Number:

TP29
TP391

Wei Jingjing, Liu Qinming, Ye Chunming, Li Guanlin. Fault Diagnosis of Mechanical Equipment Based on GA-SVR with Missing Data in Small Samples[J]. Journal of System Simulation, 2021, 33(6): 1342-1349.

References

[1] 张远绪, 程换新, 宋生建. 基于改进的RBF神经网络的滚动轴承故障诊断[J]. 工业仪表与自动化装置, 2018(6): 31-34.
Zhang Yuanxu, Cheng Huanxin, Song Shengjian.Fault Diagnosis of Rolling Bearing based on Improved RBF Neural Network[J]. Industrial Instruments and Automation Equipment, 2018(6): 31-34.
[2] Xing C H, Xu F T, Yao Z Y, et al.A Fault Diagnosis Method of Reciprocating Compressor based on Sensitive Feature Evaluation and Artificial Neural Network[J]. High Technology Letters (S1006-6748), 2015, 21(4): 422-428.
[3] 王杉, 苑津莎, 张卫华. 基于BP神经网络的变压器故障诊断[J]. 科学技术创新, 2011(29): 40-41.
Wang Shan, Yuan Jinsha, Zhang Weihua.Transformer Fault Diagnosis based on BP Neural Network[J]. Science and Technology Innovation, 2011(29): 40-41.
[4] 胡杰, 刘博, 颜伏伍, 等. 汽油机失火诊断GA-SVM方法研究[J]. 汽车技术, 2017(1): 38-42.
Hu Jie, Liu Bo, Yan Fuwu, et al.Study on GA-SVM Method for Engine Fire Diagnosis[J]. Automotive Technology, 2017(1): 38-42.
[5] 卢向华, 舒云星. 改进AFSA算法优化SVM的变压器故障诊断[J]. 计算机工程与应用, 2017, 53(17): 173-179.
Lu Xianghua, Shu Yunxing.Improved AFSA Algorithm to Optimize Transformer Fault Diagnosis based on SVM[J]. Computer Engineering and Application, 2017, 53(17): 173-179.
[6] Huang J, Hu X G.Support Vector Machine with Genetic Algorithm for Machinery Fault Diagnosis of High Voltage Circuit Breaker[J]. Measurement (S0263-2241), 2011, 44(6): 1018-1027.
[7] Fei S W, Zhang X B.Fault Diagnosis of Power Transformer based on Support Vector Machine with Genetic Algorithm[J]. Expert Systems with Applications (S0957-4174), 2009, 36(8): 11352-11357.
[8] Yang D L, Liu Y L.Gear Fault Diagnosis based on Support Vector Machine Optimized by Artificial Bee Colony Algorithm[J]. Mechanism and Machine Theory (S0094-114X), 2015(90): 219-229.
[9] 卢亚洲, 王学文, 杨兆建, 等. PSO-SVM在提升机制动系统故障诊断中的应用[J]. 机械设计与制造, 2018(6): 246-249.
Lu Yazhou, Wang Xuewen, Yang Zhaojian, et al.Application of PSO-SVM in Fault Diagnosis of Lifting Mechanism Dynamic System[J]. Machinery Design and Manufacturing, 2018(6): 246-249.
[10] Zhang Y Y, Wei H, Liao R J, et al.A New Support Vector Machine Model Based on Improved Imperialist Competitive Algorithm for Fault Diagnosis of Oil-immersed Transformers[J]. Journal of Electrical Engineering and Technology (S1975-0102), 2017, 12(2): 830-839.
[11] Zhang Z D, Dong F L.Fault Detection and Diagnosis for Missing Data Systems with a Three Time-slice Dynamic Bayesian Network Approach[J]. Chemometrics and Intelligent Laboratory Systems (S0169-7439), 2014, 138.
[12] 陈嘉宁, 杨翾, 叶承晋, 等. 基于缺失数据修复的变压器在线故障诊断方法[J]. 电力系统保护与控制, 2019, 47(15): 86-92.
Chen Jianing, Yang Xuan, Ye Chengjin, et al.Transformer On-line Fault Diagnosis Method based on Missing Data Repair[J]. Power System Protection and Control, 2019, 47(15): 86-92.
[13] 蒋少华, 桂卫华, 阳春华. 基于不完备数据的密闭鼓风炉故障诊断[J]. 华中科技大学学报(自然科学版), 2009, 37(增1): 122-125.
Jiang Shaohua, Gui Weihua, Yang Chunhua.Fault Diagnosis of Closed Blast Furnace based on Incomplete Data[J]. Journal of Huazhong University of Science and Technology (Natural Science Edition), 2009, 37(S1): 122-125.
[14] 吴健飞, 刘勤明, 吕文元, 等. 基于灰色EM-SHSMM的缺失数据下设备健康预测研究[J]. 计算机应用研究, 2018, 35(11): 3255-3258.
Wu Jianfei, Liu Qinming, Lu Wenyuan, et al.Research on Equipment Health Prediction based on Grey Em-shsmm Missing Data[J]. Computer Application Research, 2008, 35(11): 3255-3258.
[15] Liu Z K, Liu Y H, Zhang D W, et al.Fault Diagnosis for a Solar Assisted Heat Pump System under Incomplete Data and Expert Knowledge[J]. Energy (S0360-5442), 2015, 87: 41-48.
[16] 陈强强, 陈志平, 李春光, 等. 贝叶斯网络故障诊断的先验概率评估研究[J]. 电子机械工程, 2017, 33(2): 56-60, 64.
Chen Qiangqiang, Chen Zhiping, Li Chunguang, et al.Study on Prior Probability Assessment of Bayesian Network Fault Diagnosis[J]. Electro-Mechanical Engineering, 2017, 33(2): 56-60, 64.
[17] Chen D M, Yang S, Zhou F N.Transfer Learning Based Fault Diagnosis with Missing Data Due to Multi-Rate Sampling[J]. Sensors (Basel, Switzerland)(S1424-8220) 2019, 19: 1826.
[18] 杨金芳, 翟永杰, 王东风, 等. 基于支持向量回归的时间序列预测[J]. 中国电机工程学报, 2005, 25(17): 110-114.
Yang Jinfang, Zhai Yongjie, Wang Dongfeng, et al.Time Series Prediction based on Support Vector Regression[J]. Chinese Journal of Electrical Engineering, 2005, 25(17): 110-114.
[19] 董辉, 傅鹤林, 冷伍明. 支持向量机的时间序列回归与预测[J]. 系统仿真学报, 2006, 18(7): 1785-1788.
Dong Hui, Fu Helin, Leng Wuming.Time Series Regression and Prediction of Support Vector Machine[J]. Journal of System Simulation, 2006, 18(7): 1785-1788.
[20] 石怀涛, 赵纪宗, 宋文丽, 等. 基于人工蜂群优化核主元分析故障检测方法[J]. 控制工程, 2018, 25(9): 1686-1691.
Shi Huaitao, Zhao Jizong, Song Wenli, et al.Fault Detection Method based on Core Principal Component Analysis of Artificial Bee Colony Optimization[J]. Control Engineering, 2008, 25(9): 1686-1691.
[21] 郭洋. 深度学习在时间序列模式识别中的研究与应用[D]. 北京: 北京邮电大学, 2018.
Guo Yang.Research and Application of Deep Learning in Time Series Pattern Recognition[D]. Beijing: Beijing University of Posts and Telecommunications, 2018.