进化规划在多振动台随机振动控制中的应用  

An Improved Method Based on Evolutionary Programming for Multi-Shaker Random Vibration Control

作　　者：韩军[1] 鲍明[1] 倪宏伟[2] 

机构地区：[1]南京航空航天大学振动工程研究所 [2]总装工程兵科研一所

出　　处：《中国机械工程》2003年第13期1100-1103,共4页China Mechanical Engineering

摘　　要：针对多振动台随机振动控制中求解补偿矩阵的广义逆算法 ,提出利用进化规划对补偿矩阵的初值进行优化的方法 ,解决了实时控制中的非线性、宽区间内的极值求解问题。双振动台随机振动控制的仿真实验研究表明 ,将基于进化规划的广义逆求解补偿矩阵方法引入随机振动控制过程 ,在同等计算精度条件下 ,计算时间大大缩短 ,并改善了控制中共振峰处的奇异性 ,为广义逆算法在多振动台随机振动控制中的在线应用提供了有力支持。Aiming at the problem of the solution to the generalized inverse algorithm for the compensation matrix in multi-shaker random vibration control,in this paper a method for optimizing the initial values of the compensation matrix is proposed by the Evolutionary Programming (EP), which resolved the problem with nonlinear and extremum searching within a wide interval. It is shown through a simulation experiment for the dual-shaker random vibration control system that, the computational time is decreased under the same precision by introducing the generalized inverse methods to the random vibration control process to solve the compensation matrix with EP, and the singular property near the resonance peaks has been improved. It makes on-line applications of the generalized inverse algorithm to the multi-shaker random vibration control be possible.

关 键 词：多振动台 随机振动控制 广义逆 进化规划 

分 类 号：O324[理学—一般力学与力学基础] TH113[理学—力学]