基于改进粒子群算法的PIDNN解耦控制研究（英文）

PID neural network decoupling control based on improved particle swarm optimization

对于具有非线性、强耦合、大迟滞特点的多输入多输出的非线性系统,传统的控制方法无法对其进行精确解耦,导致控制精度较低。提出一种基于免疫机制的改进粒子群算法,并用此方法对PID神经网络权值进行优化,形成新型PID神经网络控制器。利用两个PID神经网络控制器对双输入双输出耦合系统进行控制以减弱系统的耦合影响。通过仿真结果表明:相对于传统PID解耦控制,该算法具有更好的动态和静态特性。可为控制领域中的解耦问题提供一定的参考。

The traditional control method cannot accurately decouple the multi-input and multi-output (MIMO) nonlinear system due to the characteristics of heavy delay,and strong coupling,which leads to lower control precision.An improved particle swarm optimization (PSO)algorithm on basis of the immune mechanism was proposed in this study,and this algorithm was used to optimize PID neural network (PIDNN)weights to form a new PIDNN controller.Two new PIDNN controllers were used to control the 2-in-2out coupling system to reduce the coupling effect of the system.The simulation results showed that this method exhibited better dynamic and static characteristics than traditional PID decoupling control algorithm.It could provide some references for the decoupling problem in the field of control.