一种基于BP神经网络PID的液冷系统设计

针对温度控制系统的时滞、非线性等特点,提出一种基于BP神经网络PID的液冷控制方法.由于空中环境复杂,液冷系统中散热效率与飞行高度、速度等关系很大而且具有不可预测性,因此也无法给控制系统建立准确的数学模型.利用BP神经网络的自学习性和函数逼近能力,结合PID控制器,可以动态调节PID参数,使系统达到优秀的控制效果.实验...

A Liquid Cooling System Design Based on BP Neural Network PID

In view of temperature control system＇s delay and nonlinear characteristics, this paper proposes a kind liquid cooling control method based on BP neural network PID. Because the airplane environment is complex in air, and the liquid cooling system have close relations with flight altitude and flight speed so it unpredictable. Therefore is also unable to give the control system establishment accurate mathematical model. Using the BP neural network self-learning and function approximation capabilities unifies the PID controller to be possible the dynamic regulation PID parameter to enable the system to achieve the outstanding control effect. The experiment proved： liquid cooling system has the very good stability and robustness based on BP the neural network PID, it solved the key problem of temperature control.