粒子群优化三相桥式整流电路故障诊断
|
摘要
针对三相桥式整流电路中晶闸管出现断路故障的问题,提出采用粒子群优化BP网络的故障诊断方法和蚁群粒子群混合算法优化BP网络的故障诊断方法。首先采用Matlab软件对电路建立仿真模型,通过仿真得到电路故障信息;然后将故障信息输入到BP网络中,对比网络输出与期望输出,得出故障诊断率;最后分别采用粒子群算法和蚁群粒子群混合算法优化BP网络,采用相同的诊断方法,分别得出两种优化网络的故障诊断率。通过对比三种网络的输出信息可以看出,未优化的、PSO优化的和混合算法优化的BP网络,收敛时迭代次数分别为6 000、2 750和715,网络故障诊断率分别为68.2、81.8%和95.5%。采用神经网络对整流电路中的晶闸管进行断路故障诊断时,通过对网络的优化,能减少诊断时间和提高诊断率。
Aiming at the problem of thyristor breaking fault in three-phase bridge rectifier circuit, a fault diagnosis method using BP network optimized by Particle Swarm Optimization(PSO) and a fault diagnosis method using BP network optimized by the hybrid algorithm of Ant Colony Optimization(ACO) and PSO were proposed. Firstly, Matlab software was used to establish a simulation model for the circuit, and the circuit fault information was obtained through the simulation. Then, the fault information was input into BP network, and the fault diagnosis rate was obtained by comparing the network output with the expected output. Finally, PSO algorithm and the hybrid algorithm of ACO and PSO were used to optimize BP network respectively. By using the same diagnosis method, the fault diagnosis rates of the two optimized networks were obtained. By comparing the output information of the three networks, it can be seen that the iteration times of the non-optimized BP network, the optimized BP netowrk by PSO, and the optimized BP network by the hybrid algorithm is 6 000, 2 750 and 715 respectively, and the network fault diagnosis rate is 68.2, 81.8% and 95.5%. When the neural network is used to diagnose the thyristor breaking fault in the rectifier circuit. By optimizing the network, the diagnosis time is reduced and the diagnosis rate is improved.
Aiming at the problem of thyristor breaking fault in three-phase bridge rectifier circuit, a fault diagnosis method using BP network optimized by Particle Swarm Optimization(PSO) and a fault diagnosis method using BP network optimized by the hybrid algorithm of Ant Colony Optimization(ACO) and PSO were proposed. Firstly, Matlab software was used to establish a simulation model for the circuit, and the circuit fault information was obtained through the simulation. Then, the fault information was input into BP network, and the fault diagnosis rate was obtained by comparing the network output with the expected output. Finally, PSO algorithm and the hybrid algorithm of ACO and PSO were used to optimize BP network respectively. By using the same diagnosis method, the fault diagnosis rates of the two optimized networks were obtained. By comparing the output information of the three networks, it can be seen that the iteration times of the non-optimized BP network, the optimized BP netowrk by PSO, and the optimized BP network by the hybrid algorithm is 6 000, 2 750 and 715 respectively, and the network fault diagnosis rate is 68.2, 81.8% and 95.5%. When the neural network is used to diagnose the thyristor breaking fault in the rectifier circuit. By optimizing the network, the diagnosis time is reduced and the diagnosis rate is improved.
引文
[1] 周海峰,王荣杰,崔博文.一种用于三相整流装置故障诊断的方法[J].太阳能学报,2013,34(1):172-176.
[2] 张松兰.模拟电路故障诊断方法解析[J].南京工程学院学报(自然科学版),2013,11(1):70-74.
[3] 胡梅,胡列峰,明德.模拟电路统一软故障诊断的研究[J].电子测量与仪器学报,2013,27(11):1060-1066.
[4] PANCERZ K,SCHUMANN A.Rough set models of Physarum machines[J].International Journal of General Systems,2015,44(3):314-325.
[5] 王晓峰,毛德强,冯尚聪.现代故障诊断技术研究综述[J].中国测试,2013,39(6):93-98.
[6] 孙必伟,潘强.模拟电路故障诊断的BP神经网络方法研究[J].现代电子技术,2011,34(14):148-150,153.
[7] 赵德赞,邢军,王智森.基于嵌入式系统的BP神经网络及其在模拟电路故障诊断中的应用[J].大连工业大学学报,2016,35(1):57-61.
[8] 张凯,杜凯,巨永锋.基于BP神经网络的提速道岔故障诊断算法[J].武汉理工大学学报,2014,36 (11) :77-81.
[9] 孙丽欣,胡文亮.神经网络在故障诊断中的研究[J].化学工程与装备,2014(2):165-168.
[10] 贾嵘,李宏斌,康会西,等.改进粒子群优化神经网络在变压器故障诊断中的应用[J].高压电器,2010,46(5):14-17.
[11] 王博,邓明,王猛,等.基于Simulink的三相桥式全控整流器仿真研究[J].实验技术与管理,2016,33(1):61-66.
[12] SOMAN R R,DAVIDSON E M,MCARTHUR S D J,et al.Model-based methodology using modified sneak circuit analysis for power electronic converter fault diagnosis[J].IET Power Electronics,2012,5(6):813-826.
[13] 李猛,王友仁.电力电子电路软故障诊断方法对比分析[J].电力测量技术,2015,38(7):110-114.
[14] 杜京义,张明哲.基于粒子群优化RBF神经网络的液压钻机故障诊断[J].煤矿机械,2012,33(5):251-253.
[15] 谢昊.电力电子电路故障诊断方法仿真[J].计算机仿真,2013,12(30):372-374.
[16] 冯金芝,陈兴,郑松林.一种改进的多目标粒子群优化算法及其应用[J].计算机应用研究,2014,31(3):675-678.
[17] 程加堂,熊伟,徐绍坤,等.基于改进粒子群优化神经网络的电力电子变压器故障诊断[J].高压电器,2012,48(2):42-45.
[18] 黄璇,郭立红,李姜,等.改进粒子群优化BP神经网络的目标威胁估计[J].吉林大学学报(工学版),2017,47(3):996-1002.
[19] 孙凯,吴红星,王浩,等.蚁群与粒子群混合算法求解TSP问题[J].计算机工程与应用,2013,48(34):60-63.
[2] 张松兰.模拟电路故障诊断方法解析[J].南京工程学院学报(自然科学版),2013,11(1):70-74.
[3] 胡梅,胡列峰,明德.模拟电路统一软故障诊断的研究[J].电子测量与仪器学报,2013,27(11):1060-1066.
[4] PANCERZ K,SCHUMANN A.Rough set models of Physarum machines[J].International Journal of General Systems,2015,44(3):314-325.
[5] 王晓峰,毛德强,冯尚聪.现代故障诊断技术研究综述[J].中国测试,2013,39(6):93-98.
[6] 孙必伟,潘强.模拟电路故障诊断的BP神经网络方法研究[J].现代电子技术,2011,34(14):148-150,153.
[7] 赵德赞,邢军,王智森.基于嵌入式系统的BP神经网络及其在模拟电路故障诊断中的应用[J].大连工业大学学报,2016,35(1):57-61.
[8] 张凯,杜凯,巨永锋.基于BP神经网络的提速道岔故障诊断算法[J].武汉理工大学学报,2014,36 (11) :77-81.
[9] 孙丽欣,胡文亮.神经网络在故障诊断中的研究[J].化学工程与装备,2014(2):165-168.
[10] 贾嵘,李宏斌,康会西,等.改进粒子群优化神经网络在变压器故障诊断中的应用[J].高压电器,2010,46(5):14-17.
[11] 王博,邓明,王猛,等.基于Simulink的三相桥式全控整流器仿真研究[J].实验技术与管理,2016,33(1):61-66.
[12] SOMAN R R,DAVIDSON E M,MCARTHUR S D J,et al.Model-based methodology using modified sneak circuit analysis for power electronic converter fault diagnosis[J].IET Power Electronics,2012,5(6):813-826.
[13] 李猛,王友仁.电力电子电路软故障诊断方法对比分析[J].电力测量技术,2015,38(7):110-114.
[14] 杜京义,张明哲.基于粒子群优化RBF神经网络的液压钻机故障诊断[J].煤矿机械,2012,33(5):251-253.
[15] 谢昊.电力电子电路故障诊断方法仿真[J].计算机仿真,2013,12(30):372-374.
[16] 冯金芝,陈兴,郑松林.一种改进的多目标粒子群优化算法及其应用[J].计算机应用研究,2014,31(3):675-678.
[17] 程加堂,熊伟,徐绍坤,等.基于改进粒子群优化神经网络的电力电子变压器故障诊断[J].高压电器,2012,48(2):42-45.
[18] 黄璇,郭立红,李姜,等.改进粒子群优化BP神经网络的目标威胁估计[J].吉林大学学报(工学版),2017,47(3):996-1002.
[19] 孙凯,吴红星,王浩,等.蚁群与粒子群混合算法求解TSP问题[J].计算机工程与应用,2013,48(34):60-63.