A modified shuffled frog leaping algorithm and its application to short-term hydrothermal scheduling

Hydrothermal system scheduling is a large, dynamic, non-convex and nonlinear programming with complicated constraints. In order to solve such complex engineering problems, a new modified shuffled frog leaping algorithm is presented in the paper. In the proposed algorithm, a threshold judging strategy is introduced to the local evolution process to improve the search ability and convergence accuracy of standard shuffled frog leaping algorithm (SFL). Numerical experience proves validity of the modified algorithm and results show it outperforms standard SFL and PSO. Finally, the proposed algorithm is applied to a short-term hydrothermal system scheduling problem. The operation results show its feasibility and high efficiency.