An efficient multi-objective adaptive differential evolution with chaotic neuron network and its application on long-term hydropower operation with considering ecological environment problem

Abstract

The long-term hydropower operation problem plays an important part of power generation system nowadays. For increasing concern about the requirement of reservoir ecological environment, this operation problem has been extended to be a multi-objective optimization problem (MOP). In this paper, a multi-objective adaptive differential evolution with chaotic neuron network (MOADE-CNN) is proposed to solve this problem, and an adaptive crossover rate is developed to adjust the search scale along with the evolution proceeds. Furthermore, the chaotic neuron operation is integrated into the mutation operator to avoid premature convergence problem, it controls the population diversity especially when differential evolution falls into local optima. The efficiency of the proposed MOADE-CNN is verified by the simulation on some benchmark problems, and more desirable results are obtained in comparison to those well-known multi-objective optimization algorithms. On achieving satisfactory performance of these test problems, MOADE-CNN is applied on the cascaded power operation system, the obtained result proves that MOADE-CNN can be a promising alternative and provide optimal trade-offs for multi-objective long-term reservoir operation scheduling with considering ecological environment problem.