Application Research Of Multivariable Constrained Predictive Control In Ultra-supercritical Boiler-turbine Coordination Control System

In China, electric power is mainly generated by coal power plant due to the primary energy structure. Ultra-supercritical unit technology has good inheritance and is easy to be implemented in large-scale. Improving Ultra-supercritical unit proportion and optimization of Ultra-supercritical unit control level is an important direction of promoting the coal-fired power generation technology.Coordination control system of Ultra-supercritical unit is developed on the basis of the traditional boiler-turbine control system. It is the core of thermal power unit control system. Improving the level of Ultra-supercritical unit coordination control is very important to improve the efficiency of Ultra-supercritical unit. Ultra-supercritical units have complex and volatile controlled characteristics and the boiler of the Ultra-supercritical unit is once-through boiler. It needs to operate with high parameter and adapt to a wide range of load change, the coordination control of Ultra-supercritical unit has the control difficulty of Coupling between boiler and turbine, large delay and strong nonlinearity.Due to the above control difficulty, the paper adopts the multiple model multivariable contrained model predictive control. Model predictive control can effectively deal with the coupling and time-delay problems and multiple model method can deal with the nonlinearity problem.The main research work is as follows:1. The Ultra-supercritical unit coordination control system is simplified as a multivariable control system with three inputs and three outputs.The three input variables are the fuel flow, the governor valve opening and the water flow; The three output flow are the steam pressure, load and steam temperature.In addition, the interaction relationship of the input variables and output variables and the control difficulty of Ultra-supercritical coordination control is analyzed in detail. Due to the above control difficulty, this paper proposes the multiple modle multivariable contrained model predictive control scheme and gives the structure diagram of the system.In addition, the double-layer structure are also described, which are the steady optimization and dynamic optimization. At last, the multiple models policy are described.2. Based on the above analysis, the multiple model multiple variable contrained model predictive control software is developed.This paper gives the system structure diagram of the software and describle the module division.For the UI module, the relationship of the class, data flow diagram, the choice of data structure and so on are given.For the algorithm module, the main program flow chart are given.In addtion, the core realization of the algorithm are also describled in detail.For the OPC interface, the general opc communication interface is realized.3. In Simulation, the nonlinear model of 1000 MW Ultra-supercritical Unit is controlled by the multiple model multivariable contrained model predictive control software.The result of real-time simulation showed the proposed control method could guarantee good steady and dynamic performance.