DSM Research And Application Of Multi-source-based Information

In this paper, present situation and development trend of the domestic and external demand side management research are analyzed comprehensively. Combined with the actual conditions of pilot areas, the DSM platform is developed based on multi-information sources from real-time multi information sources such as the integration of electrical information and weather information to achieve unified management of basic data. Meanwhile, through morphological analysis of massive load data by fuzzy C-means clustering method, load characteristics evaluation index and the correlation analysis method are proposed. Based on the ARIMA model and BP neural network model, it proposes a ultra-short-term load forecasting method which contains prediction and correction, realizing refinement classification of massive data in smart grid, supporting for the development of a top-down hierarchical partitioning optimization decisions electricity demand-side programs and improvements, improving the reliability and economy of the system, and guiding the rational use of electricity. It plays an important role on efforts to promote demand side management information platform and building long-term mechanism that seizing the good opportunity for the current large-scale construction of intelligent power distribution system, taking advantage of the full potential of smart distribution grid, greatly enhancing the power efficiency, playing the management potential of demand side, laying the foundation for further scientific electricity, rational use of electricity, saving electricity and strengthening energy management efforts...