基于改进烟花算法的非线性模拟电路测试激励优化  

Optimizing the test stimulus of the nonlinear analog circuits based on improved firework algorithm

作　　者：吴世浩 孟亚峰 王超[3] WU Shihao;MENG Yafeng;WANG Chao(Department of Electronic and Optical Engineer,Army Engineering University,Shijiazhuang 050003,China;No.63850 Unit of PLA,Baicheng 137000,China;No.65735 Unit of PLA,Dandong 118000,China)

机构地区：[1]陆军工程大学电子与光学工程系,河北石家庄050003 [2]中国人民解放军63850部队,吉林白城137000 [3]中国人民解放军65735部队,辽宁丹东118000

出　　处：《中国测试》2019年第6期138-145,共8页China Measurement & Test

基　　金：国家自然科学基金(61372039)

摘　　要：为增强非线性模拟电路故障诊断中故障模式之间的可辨识性,提高故障诊断率,提出一种基于改进烟花算法的非线性模拟电路测试激励优化方法。该方法首先采用基于Volterra频域核和BP神经网络的方法对非线性模拟电路进行建模,进而针对烟花算法存在寻优速度慢、效率低等问题,对其爆炸算子、变异算子、选择策略等进行改进,采用改进后的烟花算法对非线性模拟电路的测试激励进行寻优,通过电路仿真表明,优化后的信号可有效提高故障可分性,从而提高故障诊断率。In order to improve the distinguish ability of fault samples in nonlinear analog circuits and enhance the diagnostic performance of fault diagnosis, a novel test stimulus method of the nonlinear analog circuits based on improved firework algorithm is proposed. First, Volterra frequency-domain and BPNN is used for modeling the nonlinear circuits, then the explosion operator, mutation operator, selection strategy are improved to solve the diagnosis problem of slow speed and low efficiency, the improved firework algorithm is used to optimize the test stimulus of nonlinear analog circuits. The circuit simulation indicates that the optimized signal can improve the fault separability and improve the fault diagnosis rate effectively.

关 键 词：VOLTERRA级数 BP神经网络 非线性模拟电路 烟花算法 

分 类 号：TV641[水利工程—水利水电工程]