基于宽度学习预测的可重构多微电网随机优化框架

STOCHASTIC OPTIMIZATION FRAMEWORK FOR RECONFIGURABLE MULTI-MICROGRID BASED ON WIDTH LEARNING PREDICTION

针对多微电网稳定性差、能量管理复杂等问题,提出一种智能随机优化框架,对可再生能源普及率较高的微电网系统进行优化运行和管理。该模型考虑了电动汽车的高移动充电需求以及可再生能源的随机性。为了减轻车辆对单个微电网的负面影响,采用V2G方案,并与微电网成本函数兼容。作为可再生能源的支持策略,提出一种基于UT框架和宽度学习的深度学习的概率方法,该方法考虑到可再生能源的随机性,将系统运行在安全区域。在此基础上,设计一种基于改进SCE(Shuffled Complex Evolution)的智能优化算法,在全局空间寻找最优解。通过IEEE测试系统检验了该模型的适用性。

Aimed at the problems of poor stability of multi-microgrids and complex energy management,an intelligent stochastic framework is proposed to optimize the operation and management of microgrid systems with a high penetration rate of renewable energy.This model considered the high mobile charging demand of electric vehicles and the randomness of renewable energy.In order to reduce the negative impact of vehicles on a single micro-grid,the V2G solution was adopted and compatible with the cost function of the micro-grid.As a support strategy for renewable energy,a probabilistic method of deep learning based on UT framework and width learning was proposed.This method considered the randomness of renewable energy and ran the system in a safe area.On this basis,an intelligent optimization algorithm based on improved SCE(Shuffled Complex Evolution)was designed to find optimality in the global space.The applicability of the proposed model was verified by the IEEE test system.