Power Marketing System Required The Design And Realization Of The Electricity Demand Forecasting

With the deepening of institutional reform in electricity industry and fierce competition from outside uncertainties, managerial expertise and decision-making capabilities have become increasingly essential for domestic electricity company leaders. Meanwhile, it requires more accurate and timely information. Therefore, a comprehensive and unified information system that can operate on different platforms and transfer real-time data has been a trend of information-based industrial development.In the process of the informational construction of electric power industry, there is a massive volume of historical data lack of adequate utilization to offer substantial support for the decision-making of electric power enterprises.Based on the investigation and practices of many electric power corporations, this dissertation presents a new method of using computational verb theory (CVT) in electric power marketing. Subspace-based Multi-Dimensional classification algorithm is used in data preprocessing to ensure data optimization and reduce CVT computational complexity.With its excellent non-linear character, CVT is competent for non-linear processing, and predicting methods based on CVT are good for non-linear problems. The experiment results indicate that CVT model has good predicting effects in electric power distribution.