Improvement Of Monthly Electricity Sales Forecasting Method Based On Time-Series Method And Regression Method

Accurate forecasting accuracy of monthly electricity consumption is very important for improving performance, controlling balance of profits and doing marhing work in Power Company. Currently, the common electricity consumption forecasting methods at home and aboard are time series metheod and regression analysis method. The two methods are relatively mature, simple, and have achieved good results in many cases. However, there are still some problems to be solved in present study. This paper has made improvement research for the problems in present study and the main results are as follows:(1) Proposed an improced forecasting method of monthly electricity consumption based on multiplication model of X12 method and ARIMA model. When we use ARIMA model or seasonal ARIMA model to predict monthly electricity consumption, it is difficult to get a higher prediction accuracy because of the trend component, the seasonal cycle component and the random component in monthly electricity consumption sequence will interfere with each other. In order to solve the above problems, this paper proposed an improced forecasting method of monthly electricity consumption based on multiplication model of X12 method and ARIMA model. First of all, the multiplication model of X12 method is applied to decompose the electricity series to trend component, seasonal cycle component and random component; ARIMA model is used to forecast the trend component, weighting method and average method are used to forecast the seasonal cycle component and random component respectively; Finally, the multiplication model of X12 method is applied to fuse above three predictive value as the final predictive value of the monthly electricity consumption. Making simulation analysis with the actual data of Tong Liang district in Chongqing, and the results showed that the improved method is effective.(2) Analyzed the main influencing factors of monthly electricity consumption. This paper analyzed the main influencing factors of monthly electricity consumption based on the actual data of Tong Liang district in Chongqing, and focused on the relationship between monthly electricity consumption and economic development, tempertature, holidays, the number of holidays, month and business expansion capacity.(3) Proposed three improvement measures that consider the comfortable temperature range, the sudden variable and the Spring Festival distribution.There are three questions in existing regression model of monthly electricity consumption: the existin model ignored the fact that there is no heating and cooling measures in comfortable temperature range, at the same time, the conventional model ignored the influence of random changes on monthly electricity consumption because of random variation is difficult to quantify, in addition, the conventional model also ignored the effect of different distribution of Spring Festival to electricity consumption. These issues will affect the forecast accuracy of monthly electricity sales to a certain extent. In order to solve the above problems, this paper proposed three corresponding improvement measures: Selected low threshold temperature and high threshold temperature, and it will produce heating or cooling measures only when actual temperature is below the low threshold temperature or above the high threshold temperature, and proposed a method that use �random fluctuation level� to quantify random variation and put the quantization value into the forecast model of monthly electricity consumption as an influence factor, and proposed an improved method that consider the impact of distribution of Spring Festival to monthly electricity consumption in which the monthly electricity consumption sequence is converted to a new sequence that Spring Festical is all distributed in February firstly and the new sequence is used to participate in modeling and forecasting secondly, and the precictive value is converted to final precictive value according to the acutual distribution of Spring Festical. Making simulation analysis with the actual data of Tong Liang district in Chongqing, and the results showed that the above three improcement measures are effective.(4) Explored the forecasting method of monthly electricity consumption when considering the business expansion capacity. This paper explored the forecasting method of monthly electricity consumption when considering the business expansion capacity based on the actual data of Tong Liang district in Chongqing...