摘要
针对火电机组主蒸汽温度被控对象的不确定性和大延迟、大惯性、非线性等特点,设计了一种基于粒子群(PSO)算法、蚁群(ACO)算法、BP神经网络的智能PID串级控制系统。采用PSO算法优化ACO算法的参数、信息启发式因子α、期望启发式因子β、以及改进的ACO算法对BP神经网络初始权值进行优化;采用优化后的BP神经网络算法对PID控制器参数进行在线调整,从而实现对主蒸汽温度的动态控制。以某超临界600MW机组为对象,对ACO-BP和BP神经网络PID串级主蒸汽控制系统进行仿真试验。结果表明,ACO-BP PID串级主蒸汽控制系统较BP神经网络PID串级主蒸汽温度控制系统能更有效地克服主蒸汽温度被控对象的大延迟、时变性、非线性特性,提高了主蒸汽温度的控制品质。
Considering the uncertainty,large delay,large inertia and nonlinear property of the main steam temperature control in thermal power plants,an intelligent PID cascade control system based on several calculation methods such as the particle swarm optimization(PSO)algorithm,the ant colony optimization(ACO)algorithm,and the BP neural network was designed.The PSO algorithm was used to optimize the parameters of the ACO algorithm,information heuristic factorα,and expectations heuristic factorβ.The improved ACO algorithm was employed to optimize the initial weights of the BP neural network.Then the optimized BP neural network algorithm was applied to adjust the PID parameters on-line,thus to realize dynamic control of the main steam temperature.An ultra supercritical 600MW unit was taken as an example to conduct simulation test on the ACO-BP and BP neural network PID cascade control system.The results show that this system outperforms conventional PID control systems in control quality and robustness.
出处
《热力发电》
CAS
北大核心
2013年第11期64-68,85,共6页
Thermal Power Generation
