Intelligent Energy Management Algorithms for EV-charging Scheduling with Consideration of Multiple EV Charging Modes
Intelligent Energy Management Algorithms for EV-charging Scheduling with Consideration of Multiple EV Charging Modes
复制标题
DOI：
10.3390/en12020265
复制
发表时间：
2019-01-02
期刊：
影响因子：
3.2
通讯作者：
Zhou, Baorong
中科院分区：
文献类型：
--
作者：
Mao, Tian;Zhang, Xin;Zhou, Baorong
电动汽车(EVS)作为一种环保、节能的出行方式，越来越受到业界和国家的关注。因此，电动汽车的充放电问题已成为近年来电力系统中的一个重要挑战和研究课题。然而，先进而经济的电动汽车充电过程应该采用智能调度，这依赖于有效和健壮的算法。为此，设计了一种在基本分散搜索框架内同时考虑单向和双向充电的综合智能分散搜索(ISS)算法。ISS结构还支持灵活和恒定的充电功率速率，分别使用Filter-SQP(序列二次规划)和混合整数SQP作为模块控制的局部求解器。给出了ISS的详细设计，并对平滑日负荷分布和最小化充电成本的目标进行了测试。与基于全局搜索(GS)、遗传算法(GA)和粒子群优化(PSO)的方法相比，经过结果验证的ISS能够在显著缩短的计算时间内得到有吸引力的结果。此外，为了处理大规模的电动汽车充电场景，进一步开发了由遗传算法和ISS方法组成的混合方法。仿真结果也验证了其卓越的性能，以及极低的计算时间。
Electric vehicles (EVs) are now attracting increasing interest from both industries and countries as an environmentally friendly and energy efficient mode of travel. Therefore, the EV charging and/or discharging issue has become an important challenge and research topic in power systems in recent years. An advanced and economic EV charging process, however, should employ smart scheduling, which depends on effective and robust algorithms. To that end, a comprehensive intelligent scatter search (ISS) algorithm within the frame of a basic scatter search has been designed with both unidirectional and bidirectional charging considered. The ISS structure also supports both a flexible and constant charging power rate by respectively employing filter-SQP (sequential quadratic programming) and mixed-integer SQP as local solvers with module control. The detailed design of ISS is presented and the objectives of smoothing the daily load profile and minimizing the charging cost have been tested. Compared with methods based on GS (global search), GA (genetic algorithm), and PSO (particle swarm optimization), the outcome-verified ISS can produce attractive results with a significantly short computational time. Moreover, to handle a large scale EV charging scenario, a hybrid method comprised of a GA and ISS approach has been further developed. Simulation results also verified its prominent performance, plus superbly low computational time.