基于分布式神经动力学算法的微电网多目标优化方法  

Multiple objective optimization of a microgrid based on a distributed neural dynamics algorithm

作　　者：刘青[1] 赵洋[1] 李宁[1] 马博翔 尚英强 李文杰 LIU Qing;ZHAO Yang;LI Ning;MA Boxiang;SHANG Yingqiang;LI Wenjie(Powercable Branch,State Grid Beijing Electric Power Company,Beijing 100022,China;Wuhan Branch,China Electric Power Research Institute,Wuhan 430074,China)

机构地区：[1]国网北京市电力公司电缆分公司,北京100022 [2]中国电力科学研究院武汉分院,湖北武汉430074

出　　处：《电力系统保护与控制》2021年第11期105-114,共10页Power System Protection and Control

基　　金：国家电网总部科技项目资助(GY71-18-040)。

摘　　要：针对微电网多目标优化计算量较大问题,提出了一种考虑需求响应的微电网分布式神经动力学优化算法。首先考虑平均效率函数、微电网的排放、需求响应引起的不满意度以及总利润函数等因素建立多目标优化模型。然后应用单目标积公式将多目标优化问题转换为单目标优化问题,并证明了最优解是原始多目标问题的帕累托最优点。进一步使用对数障碍物惩罚因子处理不等式约束,利用Lasalle不变性原理和Lyapunov函数证明所提出的算法可以收敛到最优解。最后,通过仿真验证了所提方法可以在保证优化精度与收敛性条件下大大降低计算成本。To solve the problem of large amounts of calculation in multi-objective optimization of a microgrid,a distributed neural dynamic optimization algorithm considering demand response is proposed.First,the multi-objective optimization model is established considering the average efficiency function,micro grid emissions,demand response-induced dissatisfaction and total profit function.Then,the multi-objective optimization problem is transformed into a single objective optimization problem using a single objective product formula,and it is proved that the optimal solution is the Pareto best of the original multi-objective problem.A logarithmic obstacle penalty factor is used to deal with inequality constraints,and the invariance principle of LaSalle and the Lyapunov function are used to prove that the proposed algorithm can converge to the optimal solution.Finally,the simulation results show that the proposed method can greatly reduce the calculation cost under while ensuring optimization accuracy and convergence.

关 键 词：神经动力学算法 多目标优化 最优解 微电网 需求响应 

分 类 号：TM73[电气工程—电力系统及自动化]