粒子群优化技术用于故障诊断中的测点优化配置研究  

Research on Measuring Point Configuration Based on Particle Swarm Optimization Technology in Fault Diagnosis

作　　者：潘宏侠[1] 黄晋英[1] 毛鸿伟[1] 魏秀业[1] 

机构地区：[1]中北大学机械工程与自动化学院,山西太原030051

出　　处：《火炮发射与控制学报》2008年第2期58-62,共5页Journal of Gun Launch & Control

基　　金：国家自然科学基金资助(50575214);西省自然科学基金资助(2007011070)

摘　　要：针对车辆齿轮传动箱这样一个复杂的动力学系统,建立结构在不同点激励和不同点响应的传递特性模型。考虑其内部各种零部件的振动信号经不同的传递路径到达箱体表面并相互叠加在一起的实际情况,提出用基于粒子群优化算法解决故障检测中的传感器优选与测点配置问题的基本方法。通过各测点的测试分析与传递特性参数的优化,不但给出了测点布置的基本优选原则,也证明了基于粒子群优化技术的传感器与测点优化配置是可行的。Aimed at a complex dynamic system such as gear box, the transfer characteristic model of different excitation and responsive points was constructed. In consideration of the actual condition that vibration signals from different parts in the inner gear box arrived at the box body surface through different paths and mutually mixed, the basic method of sensor optimization and measuring point configuration based on PSO was put forward to solve fault detection problem. By means of testing analysis of each point and optimization of transfer characteristic parameters, the basic optimization principle of measuring point configuration was not only given, but the sensor and measuring point optimal configuration technology based on PSO were proved to be feasible.

关 键 词：人工智能 小波神经网络 粒子群优化 故障诊断 测点优化 

分 类 号：TJ810.32[兵器科学与技术—武器系统与运用工程]