| Coordinated control system is the core to run a thermal power plant stably. It received the orders from AGC and then control the equipment both in the boiler system and turbine system. Its control performance affect the safety and economy of the unit directly, also is important to the power grid. However, as a complex nonlinear MIMO system with large delay, strong coupling and time-varying parameters, is hardly controlled well by traditional control strategy. At the same time, the fire coal is different from the design coal and changed frequently with the situation of the coal market. All above set a higher requirement for the coordinate control system.Predictive control can compute the outputs of the controlled system in the future and adjust inputs ahead of time, which is different from traditional PID control strategies. So predictive control will have more excellent control performance while control a large delay MIMO system. A coal model is set in the prediction model to make up error caused by variation of coal. The multi-model approach is adopted to solve the nonlinear of global system.The major work of this paper is to study an applicable multi-model predictive control method for the coordinated control system. The main contents are as follows:1ã€A coal model is set in the prediction model based on the analysis of coal fuel affect to the unit. After research different conditions of the unit, a criteria for partition coordinate system model is presented. The routine operating data is used to construct the Hankel matrix and then the singular value decomposition (SVD) approach is used to obtain the model order. Then predictive error method is used to identify the model parameters. The identification model is verified by compared with mechanism model.2〠Based on the new prediction model, a dynamic matrix control (DMC) method with coal quality correction for coordinated system is proposed to improve the control effect. And the coal model is corrected online to offset the model error caused by coal change. A Matlab Simulink simulation is carried out to verify the control performances of the new method.3ã€The programme is been developed by using Matlab and C++ after fished the software structure design. And problems such as data communication, system monitor have been solved. Then a method is proposed to verify the parameters of DMC controller. The parameters of weighting matrix is adjusted to avoid triggering the protect system of coordinated system. In addition, a strategy is used to hold the controlled volume while the coal pulverizing system is launched or stopped. |
