Novel Multiobjective Shuffled Frog Leaping Algorithm with Application to Reservoir Flood Control Operation

Abstract

Reservoir flood control operation (RFCO) is a large scale multiobjective problem with complex constraints that require powerful algorithms to solve it. As a new metaheuristic evolutionary algorithm, shuffled frog leaping algorithm (SFLA) has the potential ability to solve multiobjective optimization problems because of its group evolution characteristic. In this paper, we present a novel multiobjective shuffled frog leaping algorithm (MOSFLA), which incorporates an archiving strategy based on self-adaptive niche method to maintain the nondominated solutions, and improves the memetic evolution process of SFLA to adapt to the multiobjective optimization problem. The numerical experiments of five Zitzler-Deb-Thiele functions indicate that MOSFLA yields better-spread solutions and converges closer to the true Pareto frontier than non-denominated sorting genetic algorithm (NGSA)-II and SPEA2. Furthermore, MOSFLA is applied to solve RFCO of the Three Gorges Project, and the results demonstrate that this algorithm can generate a solution set with uniform spread and good convergence for the problems with two conflicting objectives, including minimizing the highest reservoir water level and minimizing the peak flood discharge. Additionally, if compared with dynamic programming and NGSA-II, MOSFLA is verified to be more efficient and competitive, and thus can be provided as a new effective alternative for solving the complex reservoir operation problems.