Study On Multi-objective Optimal Control Of SCR Denitrification System

With the implementation of ultra-low emission policy,coal-fired power plants pay more and more attention to NOx emission index.The large-scale grid connection of new energy makes the coal-fired units face the double pressure of flexible operation and ultra-low emission.The rapid change of load of the units puts forward higher requirements for the denitration system.The spraying amount of ammonia in denitration system not only affects the denitration efficiency of flue gas,but also increases the escaping amount of ammonia by excessive spraying,which leads to the increase of economic cost of denitration system and pollution to the environment.In view of the complex reaction mechanism of SCR(Selective Catalytic Reduction)denitration system,which is greatly affected by the operating conditions of the unit and poor economy,it is difficult to achieve accurate control of ammonia injection volume through traditional control methods.Therefore,in order to ensure the NOx emission standard of SCR export and the economic operation of SCR denitration system,it is urgent to study a multi-objective optimization control strategy with certain load adaptability to solve the above problems.In this paper,the mechanism of SCR denitration reaction was studied,and the mechanism model of denitration system was established.Based on the historical data collected from the power plant,the undetermined parameters in the mechanism model of the denitration system were identified by Particle Swarm Optimization(PSO),and the accuracy of the model was verified,which laid a foundation for the research of control algorithm.Secondly,the mechanism model of the denitration system built is taken as the controlled object.In view of the problem that the output of the system fluctuates greatly under PID control when the working condition changes,the DMC predictive control method is proposed to control the denitration system.Finally,in order to ensure SCR export NOx emission standards and economic operation of SCR denitration system,economic indicators were added into the objective function,genetic algorithm and neural network were combined to improve the predictive control strategy of DMC,and a multi-objective optimization control method was established,which considered both ultra-low emission and economic cost.The results show that the method can meet the emission standards and reduce the cost of denitration,and can adapt to the large range of operation conditions,and the control quality is better than the conventional control.