Overall Improved Genetic Algorithm Applied in Optimal Generation Dispatching of Multi-reservoir System

Y. F. Li†* & K. A. Tutu‡

†Sichuan University, Chengdu, 610065, China
‡Mendel University, Faculty of Regional and Territorial Studies, Brno, Czech Republics

ABSTRACT: This research first introduces the basic principles of genetic algorithms, and improves the algorithm by combining with self-adaptive optimization theory concerning on its shortage of low global searching ability and slow convergent rate, so that crossover operator and mutation operator can automatically change over the fitness, thus forming adaptive genetic algorithm (AGA) and maintaining the diversity of the population effectively; secondly, this paper introduces the concept of niche to form the ultimate niche adaptive genetic algorithm to improve the global search ability, to avoid falling into local optimal solution, and to accelerate the convergent rate. The results shows that the improved genetic algorithm can obtain better results, and provide an effective and viable new way in optimization of hydropower station.

Keywords: Hydropower generation of multi-reservoir; Optimal generation dispatching; Genetic algorithm; Niche.