摘要
过热汽温调节过程是一个非线性时变过程。目前广泛使用的ＰＩＤ控制虽然有结构简单、鲁棒性强的优点，但也存在自适应性能差的缺点。作者利用神经元的学习特性，给出了一种神经元网络控制模型，并将其与串级控制相结合，形成了过热汽温智能控制结构。该结构可以实现参数的自整定，具有较强的自学习能力。最后通过仿真验证了该智能控制的有效性。
Abstract Although the PID superheated steam temperature control system of power plant is simple and robust,it has poor adaptive ability and can not meet the actual control which is a nonlinear and time varying process.In this paper,a neuron network control model is presented,and combining it with cascade control a superheated steam temperature intelligent control structure is constructed,which enable the automatic tuning of the controller parameters and has stronger self learning ability.The simulating results indicate that the intelligent control is effective.
出处
《动力工程》
EI
CAS
CSCD
1998年第2期7-10,共4页
Power Engineering
