Evaluation and prediction of flexible resource adjustment capability based on big data analysis

It is difficult for traditional methods to accurately evaluate and predict the dynamic characteristics of flexible resources. Therefore, this paper constructs a set of evaluation index system including three dimensions: technical performance, economic cost and environmental impact, and adopts the methods of combining range method, analytic hierarchy process (AHP) and entropy weight method to standardize data and assign weights, and updates the weights through the sliding window mechanism to adapt to the changes of resource performance. In the aspect of forecasting model, quantile regression technology is used to predict the output range, and Monte Carlo simulation is embedded to improve the robustness of power grid dispatching; The LSTM network is used to capture the time series dependence of dynamic response speed, and XGBoost is introduced as an integrated framework to fuse multi-source data to improve the prediction accuracy. The effectiveness of the proposed method is verified by an actual case analysis of a 50MW photovoltaic power station in a provincial power grid. The results show that the photovoltaic plus energy storage system (ESS) is excellent in output stability, response speed and cost-effectiveness, which is obviously superior to traditional gas turbine and conventional thermal power. The evaluation and prediction method of flexible resource regulation ability based on big data analysis proposed in this paper provides strong support for the optimal operation of power system and has important theoretical and practical application value.