Application Of Predictive Control In Reheat Steam Temperature Control

The control of reheated steam temperature in thermal power plant has the characteristic of large delay.Under different load conditions,the reheated steam temperature control system shows different dynamic characteristics.Aiming at the characteristics of the system,a multi-model predictive control algorithm based on comprehensive weighting is proposed.Firstly,a local linear model is established for each steady-state case.According to the data of reheated steam temperature and flue gas baffle opening,use Particle Swarm Optimization method for model identification.And the model set is obtained under some steady-state conditions.Then,according to the principle of model predictive control and the local models,the generalized predictive control is used to design the corresponding model predictive controller to solve the problem of large delay.Finally,according to the model output and the actual system output,take the integrated weighting coefficient of each controller to get the control quantity.The simulation results show that the nonlinear predictive controller used in the essay exhibits good control quality and can adapt to the change of parameters of the controlled object,and has satisfactory performance of full load tracking.