基于数据回归的热力站耗热量时间遗传期研究
Time genetic period of heat consumption in heat supply station based on data regression
摘要：
通过分析热力站耗热量的主要影响因素，建立了基于历史数据的耗热量预测模型，并采用MATLAB软件，对未来第1，3，7天的耗热量进行数据回归预测。结果表明：建筑耗热量具有时间遗传特性，耗热量不仅与室外温度有关，还与前7～10 d内的历史耗热量具有很强的相关性。耗热量时间遗传期分为3个阶段，不同热工性能建筑的热力站耗热量的时间遗传特性不同，越是新建建筑，其时间遗传期m值越大。采用预测模型对未来1 d的耗热量进行预测，其精度在5%以内。
Abstract：
Proposes a heat consumption prediction model based on historical data by analysing the main influencing factors of heat consumption in heat supply station. Predicts the data regression of heat consumption in the next first, third and seventh day by MATLAB software. The results show that the building heat consumption has time genetic characteristics, which is not only related to the outdoor temperature, but also to historical heat consumption of 7 to 10 days before. Divides time genetic period of heat consumption into three stages. It varies with different thermodynamic properties of the buildings. The newer the building, the higher the m value of time genetic period. Predicts the heat consumption of the next 1 day using this model, the result shows the less 5% of relative error.
Keywords:data regression, time genetic period, heat supply station, heat consumption, MATLAB, prediction model