Resource Loss Detection Method for Fixed Power Virtual Power Plant Considering Spatio Temporal Correlation

Resource shortage detection can avoid the phenomenon of abandoned wind and light in distribution network. In order to fix the performance of resource shortage detection on power demand side, a fixed resource shortage detection method considering temporal and spatial correlation is proposed. Using outlier data mining algorithm, the outlier data of power demand side resources are obtained, and the power load is predicted by normalizing the power load data. The multivariate normal distribution is introduced to analyze the time correlation of wind farm output, and the Copula function is used to analyze the spatial correlation of wind farm output, thus completing the time-space correlation analysis of wind farm output. The fixed power demand side resource tracking filter is used to suppress the eigenvectors of missing resources, and combined with cluster analysis, the fixed power demand side resources are classified and tested, and the constraint conditions are obtained through expectation operation, so as to realize the detection of missing resources on the fixed power demand side. The experimental results show that the proposed method can improve the resource utilization rate of fixed power demand side and the energy storage capacity of energy storage device to above 90% and above 800 kW·h respectively.