基于多群组均衡协同搜索的多目标优化发电调度

Multiobjective Optimal Generation Dispatch Using Equilibria-Based Multi-Group Synergistic Searching Algorithm

针对多目标、强约束及大规模电力系统发电优化调度问题,提出一种新型多群组均衡协同搜索算法(EMGSS)。该算法基于随机学习自动机的协同进化搜索以实现合作搜索群组之间的适应度分配和策略交互。此外,EMGSS提出一种分级均衡聚类方法为系统调度员提供一系列多样化的帕累托最优均衡前沿,并引入纳什均衡来抽取最终多目标解集的最优决策解。仿真算例采用标准IEEE 30节点及118节点系统,性能对比与仿真测试验证了所提算法在解决高维多目标节能减排发电调度问题中的优越性。

This paper presents a novel equilibria-based multiple group synergistic searching （EMGSS） algorithm to cope with the highly constrained multi-objective generation dispatch （MOGD） with multiple contradictory objectives. As for the proposed algorithm, a synergistic evolutionary searching mechanism based on stochastic machine learning is developed to achieve the fitness assignment and strategic interaction among cooperative multi-groups. Furthermore, a novel equilibria-based hierarchical clustering is proposed to provide power dispatchers with a set of diversified optimum equilibria Pareto frontier （PF）, and Nash equilibrium is used to extract the best decision solution from the resulting PF. The proposed EMGSS has been applied and tested over the IEEE 30-bus system and IEEE l l8-bus system. Case studies have verified and confirmed the superiority of the algorithm to solve the multiobjective optimization problems with high-dimensional and large-scale objective functions.