Study On Ash Fouling Monitoring And Control For Heat Exchanger Of Coal-fired Boiler

In the new century, with the development of economy, energy is becoming more and more important. In the coming years, coal-fired plant is still the main form of electricity supply. Since the boiler is the key equipment of coal-fired power plant, the energy conservation and safe operation are particularly important. The heat transfer efficiency of heat exchanger is very important in the boiler efficiency. This paper mainly investigates the heat transfer surface of coal-fired boiler. The main content of this paper is given as follows:Firstly, the ash fouling of convective exchanger is studied. The clean coefficient is defined to characterize the ash fouling level. The LSSVM is then used to build the clean heat absorption model. Based on the thermodynamic equation,the real heat absorption of the heat transfer surface is calculated. Thereafter, the heat coefficient is obtained to realize the ash fouling.The proposed model is verified through the operation data from a 300 MW coal-fired boiler in China. The simulation results show the effectiveness and high predictive accuracy of the proposed model.Secondly, the ash fouling of radiation exchanger is studied. Through analysis, the furnace exit gas temperature(FEGT) is found to be the key in the ash fouling. A novel FCM-LSSVM-PLS method is proposed to predict FEGT in this paper. In the process of FCM-LSSVM-PLS method, fuzzy c-means clustering algorithm is used to partition the training data into several different subsets by considering the characteristics of operational data. Submodels are subsequently developed in the individual subsets based on the LSSVM method. Partial least squares algorithm is applied as the combination strategy. Since the system is dynamic changing in the industrial process, the model should be updated online. The online updating algorithm is then applied to the FCM-LSSVM-PLS model. The proposed online model is verified through operational data of a 300 MW generating unit. The simulation results show that the proposed online updating model is effective for online furnace exit gas temperature forecasting.Finally, the control problem of furnace exit gas temperature is studied. Based on the establishing FCM-LSSVM-PLS model, predictive control method is applied to control furnace exit gas temperature. In the control scheme, PSO is used as the receding horizon algorithm to optimize the optimal control inputs. The simulation results show the effectiveness of our proposed control scheme.