Journal of South China University of Technology (Natural Science Edition) ›› 2006, Vol. 34 ›› Issue (10): 94-99.

• Power & Electrical Engineering • Previous Articles     Next Articles

Solving Unit Commitment Problem Based on Fuzzy Genetic Algorithm

Cai Jie-fin  Ma Xiao-qian   

  1. School of Electric Power,South China Univ.of Tech.,Guangzhou 510640,Guangdong,China
  • Received:2005-09-02 Online:2006-10-25 Published:2006-10-25
  • Contact: 蔡杰进(1977-)，男，博士生，主要从事电站系统的优化研究 E-mail:ehiven77@hotmail.com
  • About author:蔡杰进(1977-)，男，博士生，主要从事电站系统的优化研究

PDF

639

Abstract:

In order to solve the unit commitment problem,a new fuzzy genetic algorithm that closely combines the fuzzy optimization and the genetic algorithm is proposed.Then,some fuzzy inference rules are established to control the crossover rate and the mutation rate,thus improving the convergence speed and avoiding the premature conver-gence.The proposed fuzzy genetic algorithm is finally adopted to solve the unit commitment problem in a practical
case.The results show that,as compared with the traditional genetic algorithm,in the same population scale and termination criteria,the convergence iterative times of the proposed algorithm decrease even by 1 22,while the com-putation time cost per population only increases by about 0.01s.Moreover.the electricity-generating cost with the optimal unit commitment decreases,with the largest contraction up to 0.73%of the corresponding total operation cost.

Key words: unit commitment, fuzzy genetic algorithm, fuzzy inference rule