Short-Term Load Forecasting of Microgrid Based on Grey Correlation Analysis and Neural Network Optimized by Mind Evolutionary Algorithm

Aiming at solving the accuracy problem of short-term load forecasting in microgrid, a short-term load forecasting method based on grey correlation analysis and back-propagation neural network optimized by mind evolutionary algorithm (GCA-MEABP) is proposed. The factors such as meteorological characteristics and date types that affect the load of microgrid are used to select similar day sets with higher similarity by grey correlation analysis. In order to reduce the deviation of training samples and improve the forecasting accuracy, the neural network optimized by mind evolution algorithm for load forecasting is trained by samples of the similar day sets. Taking the historical data of a region in southern China as an example, the short-term load forecasting results based on GCA-MEABP are compared with the forecasting results of conventional back-propagation neural network (BPNN) and radial basis function neural network (RBFNN). The results proves that GCA-MEABP has a better forecasting accuracy.