【摘要】 过热汽温控制是电厂锅炉控制系统的一个重要环节。针对电厂过热汽温对象具有较大的惯性、时滞、非线性和动态特性随运行工况变化的特点,提出一种模糊径向基函数(RBF)神经网络的自整定PID控制器应用于过热汽温控制中,它结合了传统PID及神经网络和模糊控制的优点,可在线调整得到一组最优的PID控制参数。介绍了所提控制器在超临界机组过热汽温控制中的应用。对负荷为100%、88%、62%、44%的仿真结果表明,所提控制器能获得满意结果,优于PID控制器。更多还原

【Abstract】 The superheated steam temperature control is an important part in the boiler control system of power stations.As the superheated steam temperature has large inertia,time-delay and nonlinearity,and its dynamic characteristics change with the operating conditions,a self-tuning PID controller based on fuzzy-RBF(Radial Basis Function) neural networks is presented for its control,which has the advantages of traditional PID control,neutral networks control and fuzzy control and on-line optimizes PID parameters.Its application in a supercritical unit is introduced.Simulations under 100 %,88 %,62 %,44 % unit load conditions validate its better performance than normal PID controller.更多还原