Investigation Of The Temporal And Spatial Forecast Of Urban Heat Load

With the dramatic urbanization process and the improvement of people’s living standard, we are facing the prolonged pressure on heating load increase and fuel shortage, which is coming by the increasing of heating scale and energy consumption. Generally, there are two types of heat load forecast: one aims at single heating system, heat load forecast is usually based on historical data like thermal/hydraulic parameters of the system and outdoor meteorological parameters, which can not be used to model heat load on city level. The other is to plan the total heating load on city level, considering the existing heating systems and development of cities. Heat load and annual heat supply for a city are forecasted mainly by the methods of indicator estimation,which always overestimate the heat load. Thus, a reasonable urban planning for heating load is imminent. This paper aims at investigation of novel heat load forecasting methods on city level, taking 67 cities in severe cold and cold zones into consideration.Firstly, taking urban heating system as the research object, the effects of urban heating facilities, economic and social development, building construction, climate, town planning and government’s policy on urban heat load are analysized. 17 indexes are chosen as the member of the heat load forecasting index system, such as heating capacity, heating area, and the length of heating pipes, reflecting the development of urban heating facilities; residential land area, public land area, and industrial land area, reflecting urban construction; total population(year-end) and total household(year-end), reflecting urban population; gross regional product and per capita gross regional product, reflecting urban economy; primary industry gross product, secondary industry gross product, and tertiary industry gross product, reflecting urban industry structure; urban household disposable income and urban household living expenditure for consumption, reflecting residential income and consumption; the floor area of buildings under construction and the floor area of buildings completed, reflecting urban building industry. In view of the different functions characteristics of the 67 cities, they are divided into comprehensive cities, mining cities, mining and industrial cities, industrial cities, tourist cities and transportation cities. Calculate the comprehensive score of the six types of cities separately using the Entropy method. Then, analysize the temporal and spatial characteristics of the heat load among different types of cities.Secondly, a heat load forecasting method on urban level is built based on GM(1,1) and GM(1,N) of Grey System Theory, which is used for forecasting of single city(comparing with the second forecasting method will be introduced later). The rationality of the selected factors and the relationship between each factor and heat load are first studied using grey relational grade analysis. Then, relational analysis and stepwise least squares regression are used to reduce the number of influencing factors and select the final explanatory variables for the heat load model. The grey dynamic models of explanatory and dependent variables are established to form the grey system state equation. The heat loads of different cities are forecasted though solving the grey system state equation. At the same time, a second heat load forecasting method using mainly for different types of cities’ heat load forecasting is also presented. The model is based on the Panel Data of different types of cities. Panel Data model consist of pooled regression model, individual-mean corrected regression model and unrestricted model. Respectively choose proper Panel Data model for different types of cities. The second method need the values of explanatory variables in the planning phase to finish the urban heat load forecasting of different types of cities.Finally, several forecasting cases are presented in this paper to evaluate the rationality and the feasibility of the two novel heat load forecasting methods. It is concluded that the forecasting method based on Grey System Theory(GST) have self-study habits. With the model sample data updating, it can adjust the variables and the variables’ parameters, which will be better for heat load forcasting. the forecasting method based on Panel Data(PD) model reflect better the general characteristics of the cities of the same types of cities, whose development trend of heat load can be known. When a city is lake of statistics or the city is new builted, the cities with the same characteristics can be collected as a new type of cities. The two novel heat load forecasting methods using together can contribute to guiding the urban heating system planning.Two novel heat load forecasting methods on city level are presented in this paper. Heat load forecasting not only for single cities but also for different functions characteristics types of cities can be realized. Meanwhile, the novel methods for forecasting heat load on city level presents a comprehensive analysis on the development of a city, reflecting not only heating construction, but also the important influencing factors of urban planning, which can contribute to guiding the urban planning of heating systems and other energy systems.