Heat Load Planning Guidance Of Heat Exchange Station Based On Thermal Comfort Demand

With the continuous improvement of residents' living standards,people's expectation of heating effect has changed from the original temperature necessary to satisfy indoor life to the pursuit of better indoor thermal comfort.However,at present,most heating companies still conduct extensive regulation of heat loads based on experience.The function of the heating control system is imperfect,and there is no scientific and real-time heating guidance to meet the thermal comfort pursued by thermal users.This paper conducts a guidance study on heat load planning for heat exchange station systems based on thermal comfort requirements.It aims to make the next step of heat supply planning and online guidance for heat load regulation by building a heat load prediction model,with the help of information Internet of Things technology and machine learning algorithms to help regulators make operational decisions,and achieve the pressure based on indoor thermal comfort needs Need heating.First,analyze the heat demand.Based on the hierarchical heat demand theory,it is pointed out that the "demand" of heating on demand should be a reasonable heat demand to satisfy thermal comfort.At present,the evaluation standard for central heating in northern China is a single indoor temperature,but the comfort of the indoor thermal environment should be determined by the coupling of multiple factors.Introduce the Human Thermal Comfort Evaluation Index(PMV)and its calculation method to evaluate the heating effect.At the same time,the influence of the characteristics of the central heating system and the heat load control of the heat exchange station on the indoor thermal comfort is analyzed,and the heat exchange station,which is the closest control unit to the heat user,is determined as the modeling and control object.Next,build a heat load forecasting model and give the T+1 moment planned heat supply.The random forest algorithm was used to analyze the correlation of the model influencing factors,and the support vector regression(SVR)algorithm was used to model the historical data.Considering the impact of penalty factors and kernel parameters on the accuracy and speed of the SVR algorithm,a particle swarm optimization(PSO)algorithm was introduced to optimize the parameters,and the prediction accuracy of the optimized model was significantly improved.Since the actual data in the central heating process is not available at one time,the information contained in the sample may also change over time.Traditional offline models based on historical data have mismatches in applications.Then an online incremental learning model based on the PSO-SVR algorithm is proposed.During the modeling process,new data is continuously incorporated into the training set of the model for incremental learning.However,the data accumulated over time,the model training time increased,and the prediction accuracy became stable,and the meteorological and operating conditions of the heating process changed irreversibly with time.Then,an incremental model of heat load prediction based on time windows was proposed.According to the cost-effectiveness ratio N between training time and prediction accuracy,a suitable sliding time window is selected.When the training set data reaches N,when the new sample enters,the oldest historical data is removed after the time window,so that the model can not only guarantee the prediction accuracy,but also meet the needs of online guidance for model training time.Finally,the incremental heat load prediction model is deployed in the heating platform based on Human-Cyber-Physical Systems(HCPS)to conduct online guidance experiments on the heat load regulation of the heat exchange station.It proves that changing the control mode of the heat exchange station from "coarse" to "fine" through load planning guidance can improve the user's thermal comfort without the auxiliary heat source at the end.Reference and research significance.At the same time,build a heating Internet of Things platform to make the heating system functions more complete and visible,and improve the management level of heating enterprises.