基于混合粒子群的RBF神经网络PID控制策略在某随动系统测试平台中的应用

Application of RBF Neural Network PID Control Strategy Based on Hybrid Particle Swarm in Test Platform of Servo System

针对某武器随动系统的研制需求,设计了一种基于电动负载模拟器技术的随动系统测试平台。为了解决该平台的模型不确定性、非线性和PID参数难以匹配的问题,提出了一种基于混合粒子群算法的RBF神经网络PID控制策略。实际应用表明该控制策略易于实现PID参数的自整定,控制效果良好,具有快速响应性和较好的鲁棒性、自适应性。

A test platform of servo system based on the technology of electric dynamic load simulators was designed by aimed at the development needs of a weapon servo system. In order to solve the platform model uncertainty,nonlinearity and the problem of the proportion integration and differential( PID) parameters matched hardly,a control strategy of RBF neural network PID based on hybrid particle swarm optimization( HPSO) was proposed. In practical application,it shows that this control strategy is easy to realize selftuning of PID parameters,and has a good control effect with a fast response,strong robustness and self-applicability.