Fuzzy Least Squares Support Vector Machine Based Condition Evaluation Method of Microgrid Community

Abstract

Aiming at the demand of microgrid community energy management and coordination control system for adaption to flexible adjustment of inter-microgrid, multi-working condition control strategy, a fuzzy Least Squares Support Vector Machine based low-voltage microgrid community operating state real-time assessment model is proposed. Based on traditional electricity system operating state description method, the model establishes the boundary condition for safe, normal operating state of microgrid community and sub-microgrid, classifying the state of microgrid with multidimensional characteristic vectors such as voltage deviation rate, residual energy storage capacity, charging and discharging time, power generation, power consumption, evaluating re-al-time operating state by using fuzzy Least Squares Support Vector Machine, finally judging the operating state of microgrid community. Case calculation and analysis show that the model can make a real-time evaluation in line with system sampling period, accurately and effectively classify the operating state of sub-microgrid both on-grid or off-grid, providing reliable basis for microgrid community to quick judge the condition of the grid and flexibly adjust control strategy.