Research On Predictive Control Of Ammonia Injection Denitration System In Thermal Power Plant

Thermal power generation is currently China’s main power generation method,and national environmental protection departments have formulated strict emission standards for thermal power plant pollutants.Selective catalytic reduction(SCR)system is an effective way to remove nitrogen oxides(NOx)in flue gas,which is commonly used in coal-fired thermal power plants around the world.The reducing agent used is ammonia(NH3).This article first divides the SCR system catalyst into two parts,the outer surface and the inner surface.The outer surface mainly plays the role of rapid response,and the inner surface mainly plays the role of storage and release of NH3,which is a buffer.And considering the effects of sulfur oxides(SOx)and oxygen(O2)on the denitration reaction,improved SCR system dynamics equations are proposed,finite difference equations are established,and a new SCR system mechanism model can be established consecutively.The simulation simulates two exit indicators:ammonia escape(NH3 concentration at the outlet)and NO concentration at the outlet.The simulated annealing algorithm(SAA)is used to identify the mechanism model parameters in two steps based on whether the inlet molar ratio NH3/NO is greater than 1.Experiments prove that the mechanism model is suitable for a wide range of static and dynamic conditions and has good steady-state and transient characteristics.SCR system is a non-linear,large delay,constrained and multi-objective complex industrial system.Considering the control quality and robustness of the traditional export NOx fixed value cascade control strategy cannot meet the requirements,this paper designs the simulation mechanism model.A multi-objective and multi-model predictive control(MPC)algorithm is used to control the opening degree of the ammonia injection valve.The cascade double circuit is adopted.The inner loop is a proportional integral(PI)controller to quickly suppress secondary interference(internal interference),and the outer loop is a predictive controller.Linear transfer functions identified from the step response is use as the prediction models.To simulate the dynamic and static characteristics of the SCR system in a wide range of operating conditions,linearization is performed at multiple equilibrium points.The method of automatic switching of multiple models after linearization is analyzed,and the causes and elimination methods of the vibration during the switching process are analyzed briefly.To simplify the online calculation process of predictive control,the control variable under actual constraint conditions is calculated by reducing dimension and iterating.The control scheme considers two indicators of ammonia escape and outlet NO concentration and can automatically increase the ammonia escape weight factor when the ammonia escape is too large.Simulation results show that the multi-model predictive control has fast adjustment rate,small dynamic error,and strong robustness,which effectively impro ves the control quality.