Boiler Steam Temperature Control Based On Neural Network Modeling And Particle Swarm Optimization Algorithm

Superheated steam temperature (SST) plays an important role on the boiler unit’s security and economy. It is a main control parameter of a large coal-fired boiler unit. If SST is over high, it will affect the safety of the boiler. If SST is too low, it will reduce the plant thermal efficiency and increase the exhausting humidity of the turbine, even will affect the safety of steam turbine. Thus SST should be maintained at rated setting point, the deviation is generally controlled between the range of-5℃-+5℃. Water-Spray desuperheaters are often used to control the SST and cascade PID control scheme with several PID controllers are widely adopted to get better temperature control effect. As the boiler is a complex and large system, the superheater has nonlinear, large delay, large inertia and strong coupling characteristics. When the load changes at wide loading range, the previously-determined PID parameters often can not fit the new condition and seriously affect the quality of steam temperature control. The optimization of PID parameters is time-consuming and labor-intensive. The intelligent control is very suitable for nonlinear, strong coupling, multi-variable system modeling and optimization, so it has been got more and more attention.This paper studied the superheated steam temperature predictive optimal control method based on neural network modeling and Particle Swarm Optimization algorithm. To overcome the existed limitations of the method, this paper separately used BP neural network and Elman neural network to establish the performance model. A new simplified high-efficiency PSO algorithm discarding the concept of velocity is adopted to search the optimal controls with faster convergence rate to meet the real-time control demand. By analyzing the characteristics of the boiler superheater system and the affecting factors of SST, we use MATLAB to establish the neural network model for the600MW supercritical boiler’s superheater system, and to develop the predictive optimal control program. Detailed control simulation tests were made with the power plant simulator. By comparing with the original cascade PID control, it is shown that the proposed intelligent predictive optimal control scheme can greatly improve the superheated steam temperature control with good application prospect.