计及差异化用户偏好的光储充一体化家庭用能系统多目标优化调度

基于价格型需求响应的家庭能源管理系统优化调度可显著提升家庭用能体验。同时，用户侧光储充一体化用电新模式也给家庭能源管理带来新的挑战。针对供给侧光伏出力问题，提出了一种耦合改进惯性权重混沌粒子群算法和长短期记忆神经网络的ICPSO-LSTM组合预测模型，对光伏发电进行精准化预测；针对用能侧负荷多样性特点，将其划分为不可调度、可中断、可转移三类进行精细化建模，并综合考虑电动汽车短途、中途、长途个性化用能行为及反向供电模式Vehicle to Home(V2H)。在此基础上，根据不同用能偏好，将用户划分为经济型、标准型和舒适型，构建考虑用户用能成本和舒适度的多目标优化模型，并采用ICPSO算法进行求解。最后，对比分析了典型场景下家庭能源管理系统的实施效果。

Multi-objective scheduling optimization of home energy system integrating solar and storage equipment considering different user preferences

The optimal scheduling of household energy management system based on price based demand response can significantly improve household energy use experience. At the same time, the new mode of integration of solar, storage and charging brings new challenges to household energy management. Aiming at output prediction of photovoltaic, this research proposes an ICPSO-LSTM model which is coupled with improved inertia weight chaotic particle swarm optimization algorithm and short-term memory neural network. In view of the diversity of energy load, three categories: non schedulable, interruptible and transferable loads are modeled separately. In addition, short distance, midway and long distance of electric vehicles and vehicle to home (V2H) mode are comprehensively considered. On this basis, according to different energy preferences, users are divided into economic type, standard type and comfortable type, and a multi-objective optimization model considering user energy cost and comfort is constructed, which is solved by ICPSO algorithm. Finally, the implementation effect of household energy management system in typical scenarios is compared and analyzed.