Directing against the features of long time delay and uncertainty in the main steam temperature control of thermal power plants

a RBF-PID cascade control strategy has been put forward.Adopting a hybrid learnging algorithm

which combines the nearest neighbor-clustering method with the gradient descent method

to construct a RBF neural network

the controlled objects can be identified on-line and the parameters of PID master controller can be adjusted on-line.The simulation results show that the PID controller using RBF neural network based on hybrid learning algorithm has the advantages of high precision control and rapid response speed

the dynamic quality of system being more better than that of PID controller using RBF neural network of conventional learning algorithm.