Research On Load Forecasting And Control Algorithm Of Heat Exchange Station In Central Heating System

The central heating system is an important infrastructure,which is widely used in northern China.The heat exchange station is the central part of the central heating system.Its operating conditions directly affect the user's heating quality.Studying the heat exchange station in central heating system can not only ensure the quality of heating on demand,but also protect the environment and save energy,which has very important social and economic benefits.In this thesis,the load forecasting and control algorithm of the heat exchange station in central heating system is taken as the research core.Combined with the measured data o f heat exchange station,the heat load forecasting,the analysis and prediction of the secondary backwater temperature of heat exchange station and the control algorithm of heat exchange station are mainly studied.The major contents of this thesis include:Firstly,in this thesis,the purpose and significance of this research are introduced.And the current development situation of central heating system at home and abroad is analyzed.And at the same time,the main contents of this thesis are introduced in general.Secondly,the structure and working principle of the central heating system of heat exchange station are studied,and the system characteristics and control difficulties of heat exchange station are analyzed.The heating regulation mode of heat exchange station is also introduced.Thirdly,in this thesis,a hybrid model that combines similar day selection and Deep Neural Networks to construct SD-DNNs model for short-term load forecasting is presented.An improved Euclidean Norm weighted by eXtreme Gradient Boosting is used to evaluate the similarity between the forecasting day and historical days.In the improved Euclidean Norm,the outdoor temperature,wind force and day-ahead load are simultaneously considered.And eight features are chosen as inputs of the DNNs to predict the heat load.In the verification experiment,the proposed SD-DNNs model is compared with other common heat load prediction models.The experimental results demonstrate that the SD-DNNs model can accurately forecast the heat load.Fourthly,according to the collected data of heat exchange station,the secondary backwater temperature of heat exchange station is analyzed based on the Self-Organizing Feature Maps.Through the feature selection,three prediction feature sets were established to predict the secondary backwater temperature.The results show that DNNs model can accurately predict the secondary backwater temperature and save energy.Fifthly,due to the slow convergence speed of conventional PID control algorithm,based on the collected data,the temperature control system model of heat exchange station is established in this paper.In addition,the temperature control algorithm of BP-PID heat exchange station is used.The simulation results show that BP-PID has fast convergence speed and good control stability.Finally,the main work of this thesis is summarized.And the major innovation of this thesis is clarified.Moreover,the future work of the thesis is prospected.