基于LSTM网络学习的电动汽车实时能量管理优化策略

Real⁃time energy management optimization strategy of electric vehicle based on LSTM network learning

中文关键词: 电动汽车  分层架构  长短期记忆神经网络  深度学习  有序调度策略

英文关键词: electric vehicle  hierarchical architecture  LSTM  deep learning  ordered scheduling strategy

中文摘要:

      对电动汽车负荷进行有序控制可以改善地区电网的负荷特性，降低充电成本。由于无法预测未来电动汽车的准确接入时间及充电需求，故无法对电动汽车的接入进行全局最优安排。针对该问题，提出基于深度长短期记忆神经网络的电动汽车实时能量管理系统及优化策略。首先构建了包括电网层、区域能量管理系统和充电站能量管理系统的电动汽车 3 层管理架构，对大规模电动汽车进行分层分区管理；然后提出了基于深度长短期记忆神经网络的区-站两级交互策略，利用历史负荷信息求解出的历史最优解训练学习网络，用以指导新的实时优化；提出的策略在保证用户充电需求的前提下，能够进一步降低充电成本，改善区域负荷峰谷特性。最后，通过仿真算例验证了提出的分层架构及管理策略的有效性及优越性。

英文摘要:

      Orderly control of electric vehicle load can improve load characteristics of the regional power grid and reduce the charging cost. Since it is impossible to predict the accurate access time and charging demand of electric vehicles in the future, it is impossible to make a global optimal arrangement for the access of electric vehicles. Aiming at this problem, a real-time energy management system and optimization strategy for electric vehicles based on deep long short-term memory neural networks are proposed. Firstly, a three-tier management architecture for electric vehicles including the grid layer, regional energy management system and charging station energy management system is constructed, and arge-scale electric vehicles are managed hierarchically and partitioned；Then a district-station two-level interaction strategy based on deep long short-term memory neural network is proposed, and the historical optimal solution obtained by historical load information is used to train the learning network to guide new real-time optimization；The proposed strategy can further reduce the charging cost and improve the area under the premise of ensuring the users’charging demand load peak and valley characteristics. Finally, asimulation example verifies the effectiveness and superiority of the proposed layered architecture and management strategy.