Fault Diagnosis of Manifold SVM Analog Circuit based on Particle Swarm Optimization
-
摘要: 支持向量机(SVM)一直被广泛应用于分类判别领域,在模拟电路的故障诊断中,电路普遍复杂多样,传统支持向量机只考虑数据类间距离最大化。本文中提出的粒子群优化的流形支持向量机,在保证数据最大类间间隔的同时,使映射在特征空间的数据,能保持原始空间的流形结构。同时将粒子群算法与SVM相结合,对支持向量机中的权重参数优化,使得对故障的诊断率比传统方法提高了2%~6%。通过实验发现,本文方法有效增强了模拟电路故障诊断的精确度。Abstract: The support vector machine (SVM) has been widely used for classification and discrimination, but the traditional SVM only considers the maximum distance between different data classes. To diagnose the fault of a complex and diverse analog circuit, this paper uses the manifold SVM based particle swarm optimization to ensure that the data mapped in the feature space can maintain the manifold structure of the original space while ensuring the maximum interval between different data classes. At the same time, the particle swarm optimization is used to optimize the weight parameters of the SVM, so that the fault diagnosis rate can be improved by 2%~6% compared with the traditional SVM. The experimental results show that this method effectively enhances the accuracy of analog circuit fault diagnosis.
-
表 1 电路中软故障模式
故障类型 标称值 故障值 C1/μF ↑ 0.01 0.015 C2/μF ↑ 0.01 0.014 C3/μF ↑ 0.001 0.001 4 C4/μF ↑ 0.001 0.001 5 R1/Ω ↑ 1 000 1 600 R2/Ω ↓ 1 000 400 R3/Ω ↑ 4 000 6 500 R4/Ω ↓ 4 000 1 800 R5/Ω ↑ 10 000 16 000 R6/Ω ↓ 10 000 4 500 R7/Ω ↑ 60 000 95 000 R8/Ω ↓ 60 000 25 000 注:表中符号↑和↓分别表示偏大和偏小故障。 表 2 电路故障诊断率
故障类型 SVM方法 LS_SVM方法 本文方法 C1 ↑ 0.800 0 0.860 0 0.880 0 C2 ↑ 0.740 0 0.760 0 0.800 0 C3 ↑ 0.700 0 0.660 0 0.760 0 C4 ↑ 1 1 1 R1 ↑ 0.800 0 0.760 0 0.820 0 R2 ↓ 0.820 0 0.800 0 0.860 0 R3 ↑ 0.880 0 0.920 0 0.920 0 R4 ↓ 0.640 0 0.720 0 0.740 0 R5 ↑ 0.940 0 0.980 0 0.980 0 R6 ↓ 0.960 0 0.980 0 0.980 0 R7 ↑ 0.860 0 0.860 0 0.940 0 R8 ↓ 0.860 0 0.900 0 0.920 0 -
[1] 廖剑, 史贤俊, 周绍磊, 等.基于局部图嵌入加权罚SVM的模拟电路故障诊断方法[J].电工技术学报, 2016, 31(4):28-35 doi: 10.3969/j.issn.1000-6753.2016.04.005Liao J, Shi X J, Zhou S L, et al. Analog circuit fault diagnosis based on local graph embedding weighted-penalty SVM[J]. Transactions of China Electrotechnical Society, 2016, 31(4):28-35(in Chinese) doi: 10.3969/j.issn.1000-6753.2016.04.005 [2] 辛健, 马力.小波变换和与神经网络的电路故障诊断[J].现代电子技术, 2017, 40(5):174-177 http://d.old.wanfangdata.com.cn/Periodical/xddzjs201705044Xin J, Ma L. Circuit fault diagnosis based on wavelet transform and neural network[J]. Modern Electronics Technique, 2017, 40(5):174-177(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/xddzjs201705044 [3] 何拥军, 曾文英, 曾文权.信息熵支持向量机算法传感器故障诊断研究[J].计算机仿真, 2011, 28(10):184-186, 382 doi: 10.3969/j.issn.1006-9348.2011.10.045He Y J, Zeng W Y, Zeng W Q. Circuit fault diagnosis based on energy entropy and SVM[J]. Computer Simulation, 2011, 28(10):184-186, 382(in Chinese) doi: 10.3969/j.issn.1006-9348.2011.10.045 [4] 高艳云, 庞敏.基于最小流形类内离散度的支持向量机[J].计算机应用研究, 2015, 32(9):2639-2642 doi: 10.3969/j.issn.1001-3695.2015.09.019Gao Y Y, Pang M. Support vector machine based on minimum manifold-based within-class scatter[J]. Application Research of Computers, 2015, 32(9):2639-2642(in Chinese) doi: 10.3969/j.issn.1001-3695.2015.09.019 [5] Xue B, Zhang M J, Browne W N. Particle swarm optimization for feature selection in classification:a multi-objective approach[J]. IEEE Transactions on Cybernetics, 2013, 43(6):1656-1671 doi: 10.1109/TSMCB.2012.2227469 [6] 刘忠宝, 王召巴, 赵文娟.流形判别分析和支持向量机的恒星光谱数据自动分类方法[J].光谱学与光谱分析, 2014, 34(1):263-266 doi: 10.3964/j.issn.1000-0593(2014)01-0263-04Liu Z B, Wang Z B, Zhao W J. Automatic classification method of star spectra data based on manifold-based discriminant anaysis and support vector machine[J]. Spectroscopy and Spectral Analysis, 2014, 34(1):263-266(in Chinese) doi: 10.3964/j.issn.1000-0593(2014)01-0263-04 [7] 张凯军, 梁循.马氏距离多核支持向量机学习模型[J].计算机工程, 2014, 40(6):219-224 http://d.old.wanfangdata.com.cn/Periodical/jsjgc201406047Zhang K J, Liang X. Learning model of multiple kernel support vector machine with mahalanobis distance[J]. Computer Engineering, 2014, 40(6):219-224(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jsjgc201406047 [8] 陶剑文, 王士同.局部保留最大信息差υ-支持向量机[J].自动化学报, 2012, 38(1):97-108 http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=CAS201303040000635486Tao J W, Wang S T. Locality-preserved maximum information variance v-support vector machine[J]. Acta Automatica Sinica, 2012, 38(1):97-108(in Chinese) http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=CAS201303040000635486 [9] 宋国明, 王厚军, 姜书艳, 等.最小生成树SVM的模拟电路故障诊断方法[J].电子科技大学学报, 2012, 41(3):412-417 doi: 10.3969/j.issn.1001-0548.2012.03.018Song G M, Wang H J, Jiang S Y, et al. Fault diagnosis approach for analog circuits using minimum spanning tree SVM[J]. Journal of University of Electronic Science and Technology of China, 2012, 41(3):412-417(in Chinese) doi: 10.3969/j.issn.1001-0548.2012.03.018 [10] 张进, 丁胜, 李波.改进的基于粒子群优化的支持向量机特征选择和参数联合优化算法[J].计算机应用, 2016, 36(5):1330-1335 http://d.old.wanfangdata.com.cn/Periodical/jsjyy201605029Zhang J, Ding S, Li B. Improved particle swarm optimization algorithm for support vector machine feature selection and optimization of parameters[J]. Journal of Computer Applications, 2016, 36(5):1330-1335(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jsjyy201605029 [11] 赵卫伟, 潘宏侠.基于PSO参数优化的支持向量机齿轮箱故障诊断研究[J].机床与液压, 2014, 42(7):152-154 doi: 10.3969/j.issn.1001-3881.2014.07.041Zhao W W, Pan H X. Research on gearbox fault diagnosis based on PSO parameter optimization and SVM[J]. Machine Tool & Hydraulics, 2014, 42(7):152-154(in Chinese) doi: 10.3969/j.issn.1001-3881.2014.07.041 [12] Ding S W, Jiang H Q, Li J J, et al. Optimization of well placement by combination of a modified particle swarm optimization algorithm and quality map method[J]. Computational Geosciences, 2014, 18(5):747-762 doi: 10.1007/s10596-014-9422-2 [13] 谷文成, 柴宝仁, 滕艳平.基于粒子群优化算法的支持向量机研究[J].北京理工大学学报, 2014, 34(7):705-709 http://d.old.wanfangdata.com.cn/Periodical/kzllyyy200605014Gu W C, Chai B R, Teng Y P. Research on support vector machine based on particle swarm optiminzation[J]. Transactions of Beijing Institute of Technology, 2014, 34(7):705-709(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/kzllyyy200605014 [14] 庄严, 白振林, 许云峰.基于蚁群算法的支持向量机参数选择方法研究[J].计算机仿真, 2011, 28(5):216-219 doi: 10.3969/j.issn.1006-9348.2011.05.053Zhuang Y, Bai Z L, Xu Y F. Research on parameters of support vector machine based on antcolonyalgorithm[J]. Computer Simulation, 2011, 28(5):216-219(in Chinese) doi: 10.3969/j.issn.1006-9348.2011.05.053 [15] 王勤勇, 王月海, 潘国庆, 等.基于小波分析的最优故障特征提取研究[J].计算机测量与控制, 2016, 24(1):295-299 http://d.old.wanfangdata.com.cn/Periodical/jsjzdclykz201601082Wang Q Y, Wang Y H, Pan G Q, et al. Research of optimal fault feature extraction based on wavelet analysis[J]. Computer Measurement & Control, 2016, 24(1):295-299(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jsjzdclykz201601082