Research and application of power consumption index based on power big data analysis

As the "thermometer" and "barometer" of the national economy, electricity consumption is closely related to the economic operation. The change of electricity consumption refects the activity of economic operation and profoundly refects the local economic and social operation. In order to deeply analyze the dynamic change degree and trend of power consumption, based on the power related data of Zhejiang Province in recent three years, this paper proposes and constructs the power consumption index model by introducing the power consumption status factor and expectation factor, which realizes the accurate analysis of the power consumption level of various regions, industries and industries. In addition, the deep learning long short term memory（LSTM） neural network is used to analyze the power consumption level of various regions, industries and industries, This paper forecasts the power consumption index, analyzes the social and economic development trend, and provides important reference for the government to formulate policies.