Research On Predictive Control Of Flue Gas Oxygen Content In Thermal Power Plant Based On Model

In recent years,as the country’s efforts for energy conservation have increased.,research on technologies for the economical and efficient operation of thermal power plants has always been the focus of research in related fields.The oxygen content of the flue gas is critical parameter reflecting the real-time combustion status of the boiler,and optimal control of it is the main way to improve the combustion efficiency of the boiler..At the same time,Realizing the online evaluation of the operating economic performance of thermal power plant units has an evaluation and guiding significance for the control of flue gas oxygen content.The research object in this paper is a 660MW coal-fired boiler in a thermal power plant in Zhejiang.The research content is as follows:Firstly,research on the modeling of the oxygen content of the flue gas is studied.Aiming at the problem of strong nonlinearity in the boiler combustion system of thermal power plants,in this paper,a data-driven PSO-GA-Elman network modeling strategy is used to construct a prediction model of flue gas oxygen content.Elman neural network is used to model the oxygen content of boiler system flue gas,and then PSO(particle swarm optimization)and GA(genetic algorithm)are combined to optimize the iterative process of Elman network weight and threshold,and finally establish the optimal structure of the prediction model.Using historical operating data of thermal power plants,a comparative simulation experiment with PSO-Elman model and PSO-LSSVM model is carried out,which proves that the model has high prediction accuracy and generalization ability.Then,the optimization control problem of flue gas oxygen content is studied.In order to solve the tracking control problem of the boiler system to the optimal flue gas oxygen content setting value,this paper studies a predictive control method based on the PSO-GA-Elman multi-step recursive predictive model.The predicted output error is used as an optimization index,a quadratic performance index function is set up,and the optimal control sequence is obtained by the Newton iteration method.Finally,a feedback correction method for online adjustment of weights is designed,which uses real-time forecast error correction model to predict output.The control simulation experiment based on historical operating data shows that under the action of the predictive control method adopted in this paper,no matter the working conditions are stable or rapid,the error between the output of flue gas oxygen content and the optimal oxygen content setting curve is maintained as a whole.The control effect is good.Finally,after the boiler system is under predictive control,the economic performance evaluation of the unit operation is studied.In view of the traditional methods that rely on subjective factors such as expert scoring,this paper adopts a comprehensive evaluation method to evaluate the economic performance of thermal power plant units.In this study,the boiler thermal efficiency,air preheater leakage rate and auxiliary power consumption rate were selected as the economic performance indicators for unit evaluation,and then the online monitoring of each economic indicator was completed through modeling and thermodynamic calculation methods.Then,referring to the degree of influence of different economic indicators on coal consumption,the subjective weight of each indicator is figured.And use the entropy weight method to calculate the objective weight,and finally apply the minimum discriminant information theory to comprehensively obtain the overall evaluation score of the unit’s operating economic performance.After verifying the rationality of the proposed evaluation method,this method is used to verify the economic performance of the units with predictive control effects.Simulation experiment results show that the economic performance of thermal power generating units under the predictive control proposed in this paper has improved significantly.