基于粒子群优化的输入整形器参数自整定算法

The input shaper parameters based on particle swarm optimization tuning algorithm

针对共轴传动中的扭振问题,首先对其进行结构分析,建立数学模型,同时搭建多质转动试验平台对共轴传动中的扭振现象进行研究。其次采用输入指令整形的方法对输入信号进行滤波。传统的输入指令整形器对模型有一定的依赖性、参数调节较为复杂。针对输入整形器存在的问题,提出一种基于粒子群算法的参数自整定方法。通过传函变换,实现了在线信号采集,离线优化处理,避免了在线优化容易激起系统振荡,而离线建模精确度不高的问题,优化过程无需建模、无振动、精确度高。算法在试验平台上验证,结果证明输入整形器与陷波滤波器等传统方法相比较,在大幅抑制了系统扭振的同时较小地牺牲其动态性能,实现了系统的快速无振响应。

Regarding torsional vibrations during coaxial transmission,its structe was analyzed and a mathematical model was establised,meanwhile a multi mass rotating test platform was built to study torsional vibrations in coaxial transmssion.Secondly,input command shaping method was used to filter input signals.Traditional input shapers,to some extent,rely on modes and the parameter adjustment is more complex.In terms of the existence of input shapers,a parameter self-tuning method was proposed based on PSO,and by transfer-function transformation,it can achieve online signal collection and offline optimization,avoiding system oscillation aroused easily by online optimization,while offline optimization does not require modeling with vibration free and high accuracy.Different algorithms were verified on test platform and the result shows that comped to traditional trap filter and other filters,the input shaper greatly restrains system torsional vibrations meanwhile comprom ises its dynam ic functions as little as possible,and it achieves fast vibration-free system response.