基于改进遗传算法的智能柔性机械臂压电片参数优化

Optimization of position and piezoelectric parameter of smart flexible manipulator based on modified genetic algorithm

针对压电智能柔性机械臂的挠度形变振动问题,建立了一种基于改进的遗传算法的压电片参数优化方法。由于传统遗传算法在优化到某些代时会丢失优良遗传基因,因此发展了精英交配策略,以最大限度地保留了优良基因。本文首先建立了机械臂非线性主动控制模型,然后以平均对数衰减率为优化目标函数,采用改进遗传算法的优化策略,对压电片位置、长度、振动控制参数进行了优化。将优化结果与传统遗传算法及数值仿真实验结果进行了比较,结果表明:改进的遗传算法在第18代得到最优值,而传统遗传算法在第25代得到最优值,改进算法比传统遗传算法的收敛速度快,改进遗传算法最优化位置振动控制在1s时使臂2中点振幅降低约为数值仿真实验法的五倍;另外,当优化参数的数量增多时,改进的遗传算法仍然能够快速得到最优结果。

In order to decrease the deflection deformation vibration of smart flexible manipulators, the method of using the average attenuation rate as objective function and a modified Genetic Algorithm(GA) as optimization algorithm is developed by analyzing the nonlinear performance of smart piezoelectric flexible manipulator. The Elite Mating Strategy(EMS) is got to improve the efficiency of GA. A double-linkage flexible manipulators is established in Matlab/Simulink for verification. The position and length of piezoelectric actuator and some control parameters of the control system are optimized. Compared with the traditional GA and the exhausted method, the result of optimization shows that the optimization method in this paper can solve the parameters optimizing problem of smart piezoelectric flexible manipulator much better.