Abstract: Reservoir Flood Control Operation (RFCO) is a complex Multi-objective Problems (MOPs),which has many complex constraints,interdependent decision variables,and conflicting optimization objectives.Traditional research focuses on transforming multi-objective problem into single objective problem,which has some limitations in practical application.A Multi-objective Culture Whale Optimization Algorithm (MOCWOA) was presented for reservoir flood control operation.To improve the diversity and convergence accuracy of the results,the Cultural Algorithm (CA) was taken as MOCWOAs framework,the whale optimization algorithm was adopted in the population space,and three knowledge structures in the belief space were defined.MOCWOA was first tested on benchmark problem.Then it was further applied to the actual reservoir flood control operation problem,and compared with several well-known multi-objective optimization algorithms.The results showed that MOCWOA algorithm had a certain competitive advantage.

Key words: reservoir flood control operation, multi-objective optimization, whale optimization algorithm, culture algorithm, Pareto optimality

摘要： 水库防洪调度问题(RFCO)是复杂的多目标问题(MOPs),具有众多复杂的约束条件,相互依存的决策变量,以及相互冲突的优化目标,传统研究多停留在将多目标问题转换为单目标问题解决,在实际应用中存在一定限制。鉴于此,提出一种针对水库防洪调度的多目标优化方法——文化鲸鱼算法(MOCWOA)。MOCWOA以文化算法(CA)为框架,在种群空间采用鲸鱼优化算法(WOA),在信度空间定义了3种知识结构以提高算法所得结果的多样性和收敛精度。MOCWOA先应用于典型测试函数的优化,之后进一步应用于实际的水库防洪调度问题,并与几种优秀的多目标优化算法进行对比,结果表明,无论是在典型测试函数上,还是在实际RFCO问题上,MOCWOA都具有一定的优势。

关键词: 水库防洪调度, 多目标优化, 鲸鱼算法, 文化算法, 帕累托最优