Research On Heat Load Forecasting Of District Heating Systems

Energy saving of urban central heating is one of the effective ways of building energysaving conservation. The operation of the heating system directly affects the overall energysaving of the system. This paper presents the research on heat load forecasting method ofheating system, the purpose is to provide a prediction methods of heating system about heatload value, and achieve the economic operation of the heating system under the condition ofensuring that meet the actual heating demand of each user (thermal station).The district heating pipe network system is the research object in this paper, based onthe neural network algorithm optimized by genetic algorithm, researching the forecastmethod of the heating load of the heating load. First start with the factors affecting heatingload, analyze the characteristics of the building thermal process and the load model.According to the characteristics of the heating load, the factors affecting heating load can bedivided into meteorological factors, construction factors and human factors. Because of theunpredictability of user behavior, it become accidental factors in load forecasting. On thisbasis, according to the main influence factors in the process of heat load forecast, study theheat load method.The methods that can be used to predict the thermal load present were evaluated.Because heating system has the characteristics of time-varying, time-delay, randomness,contingency and so on, neural network method of heating load forecasting has obviousadvantages, among them, the BP neural network algorithm has the disadvantage ofconvergence slow, fault tolerance, easy to fall into local minimum and so on, and thegenetic algorithm can compensate for the shortcomings of BP algorithm, therefore, thisstudy chose this way. With MATLAB software as the platform in this paper, using geneticalgorithm to optimize the weights of network structure, than combining the BP neuralnetwork training algorithm of historical data, obtain heating load forecast results finally.