Abstract:

Traditional optimization methods, when used to solve the mathematically complex and high-dimensional problem in optimal operation of a cascade hydropower station system, often encounter dimension disaster and trapping into a local optimum. This paper presents an immune-shuffled frog-leaping algorithm (ISFLA) that adopts a clone-selecting algorithm in population space in the SFLA framework. ISFLA constructs subgroup shuffling and executes operations of cloning and selecting, meanwhile it adopts an advanced step-updating policy to improve local search ability. It has been applied to optimized operation of cascade hydropower stations for short-term power generation. Results show that ISFLA has better efficiency and superiority to standard SFLA, particle swam optimization, or progressive optimization algorithm.