中文说明：
多变量输入、输出、多干扰、非线性和强耦合的复杂系统控制是一个比较困难的问题，常用的控制器可能因为多变量耦合问题难以控制系统。PID神经元网络是一种多层前向神经元网络，它的各层神经元个数、连接方式、连接权值是按照PID控制规律的已有原则和经验确定的，是一种动态的符合控制系统的前向网络。但是由于PID网络初始权值随机取值的原因，每次控制的效果都有所差别，个别情况下控制效果还比较差。本案例研究了基于PID神经元的多变量耦合系统控制，并用PSO算法来优化控制器以取得更好的控制效果。
English Description：
The control of complex systems with multivariable input, output, multi disturbance, nonlinearity and strong coupling is a difficult problem. The commonly used controllers may be difficult to control the system because of the multivariable coupling problem. PID neural network is a kind of multi-layer forward neural network. The number of neurons, connection mode and connection weight of each layer are determined according to the existing principles and experience of PID control law. It is a dynamic forward neural network in line with the control system. However, due to the random value of the initial weights of PID network, the effect of each control is different, and the control effect is relatively poor in individual cases. This case studies the PID neuron based multivariable coupling system control, and uses PSO algorithm to optimize the controller to achieve better control effect