基于Lbest PSO和NNs的电液伺服系统输出力PID控制研究

RESEARCH ON PID CONTROL FOR OUTPUT FORCE OF ELECTRO-HYDRAULIC SERVO SYSTEM BASED ON LBEST PSO AND NNS

为了实现电液伺服系统输出力的稳定控制,结合局部最优粒子群优化算法和神经网络模型,提出一种PID控制器设计方法。该方法将神经网络模型(NNS)与PID控制器耦合,得到基于神经网络的PID控制器参数整定结构;再采用局部最优粒子群优化算法(Lbest PSO)确定神经网络的权重,从而得到基于局部最优粒子群优化算法和神经网络的PID控制算法;最后将提出的PID控制算法用于控制虚拟的电液伺服加载系统,以进行仿真实验。仿真结果表明,由该PID控制器控制的电液伺服系统的输出力平稳地收敛于给定力,从而提高了系统的稳定性。

In order to realise stable control on the output force of electro-hydraulic servo system,by combining the local best particle swarm optimisation（ Lbest PSO） with neural network model,we proposed a design method for PID controller. This method couples the neural networks（ NNs） model and PID controller and gets the NNs-based parameters tuning structure of PID controller. Then it uses the Lbest PSO to determine the weight of NNs,so that obtains the best PID control algorithm which is based on both Lbest PSO and NNs. Finally,we applied the proposed PID control algorithm in controlling the virtual electro-hydraulic servo loading system and made simulation experiment as well. Simulation results showed that the output force of electro-hydraulic servo system controlled by the proposed PID controller smoothly converged to the given force,so that the controller did improve the stability of the system.