蚁群算法优化BP神经网络在主汽温控制中的应用

摘要

针对火电厂主汽温被控对象的不确定性及大延迟、大惯性及非线性等特点,设计了一种基于蚁群算法、BP神经网络的智能PID串级控制系统.采用蚁群算法对BP神经网络的初始权值进行优化,再利用BP神经网络算法对PID参数进行在线调整,从而实现了对主蒸汽温度的动态控制.仿真结果表明:该系统在控制品质、鲁棒性方面都明显优于常规PID控制系统.%Considering the uncertainty,the large delay and inertia and the nonlinearity of the main steam temperature control in power plant,an intelligent PID cascade control system based on particle ant colony optimization algorithm and BP neural network was designed,in which,the ant colony optimization algorithm can optimize the initial weights of BP neural network,and the BP neural network algorithm can on-line adjust the PID parameters so as to realize dynamic control of main steam temperature.The simulation results show that this system outperforms conventional PID control system in control quality and robustness.