基于分解多目标进化的椭圆定日镜场布局  

Elliptical Heliostat Field Layout Optimization Based on MOEA/D

作　　者：邓立宝[1] 吴怡然 郭苏[2] DENG Libao;WU Yiran;GUO Su(School of Information Science and Technology,Harbin Institute of Technology,Weihai,Weihai 264209,China;College of Energy and Electrical Engineering,Hohai University,Nanjing 210098,China)

机构地区：[1]哈尔滨工业大学(威海)信息科学与工程学院,山东威海264209 [2]河海大学能源与电气学院,江苏南京210098

出　　处：《郑州大学学报（工学版）》2020年第5期37-43,共7页Journal of Zhengzhou University（Engineering Science）

基　　金：江苏省自然科学基金项目(BK20181308);广西自动检测技术与仪器重点实验室基金项目(YQ19203)。

摘　　要：针对塔式太阳能热电站中的定日镜场布局多目标优化问题,将基于分解的多目标进化算法(MOEA/D)应用于定日镜场布局领域,提出了基于改进的MOEA/D多目标定日镜场布局优化算法(MOEA/D-HFL)。首先建立了以镜场年均综合光学效率和镜场占地面积为目标的椭圆形定日镜场优化模型,接着将基于佳点集和反向学习的初始种群生成策略、目标函数稳定归一化机制以及动态遗传交叉分布指数引入MOEA/D用于求解该问题,获得了定日镜场布局问题的Pareto前沿,并利用模糊集理论获得了最优折中解。为验证所提算法的性能,将MOEA/D-HFL算法与NSGA-Ⅱ和基本MOEA/D对比,仿真结果证明了MOEA/D-HFL在多目标定日镜场布局问题上的高效性与准确性。To solve the multi-objective heliostat field layout optimization in solar power tower system,multiobjective evolutionary algorithm based on decomposition(MOEA/D)was introduced into the domain of heliostat field layout,and a heliostat field layout optimization algorithm based on an improved MOEA/D(MOEA/D-HFL)was proposed in this paper.In this method,firstly an elliptical heliostat field model was set up aimed at optimizing annual-averaged overall optical efficiency and the land area occupied.Secondly,initial population generation strategy based on good-point set and opposition-based learning,stable normalization of objectives and dynamic genetic crossover distribution index were applied into MOEA/D to solve this problem.Pareto front of heliostat field layout problem was obtained and optimal compromise solution was got through fuzzy set theory.To validate the performance of the proposed algorithm,MOEA/D-HFL was compared with NSGA-Ⅱand original MOEA/D algorithms,and the simulation results confirmed the effectiveness and accuracy of the proposed method.

关 键 词：多目标优化 MOEA/D 塔式太阳能发电 定日镜场布局 佳点集 反向学习 归一化 

分 类 号：TK513.1[动力工程及工程热物理—热能工程]