摘要
Unit commitment problem with huge number of constraints is considered as an important optimization problem encountered in the electric power systems. This paper proposes a modified binary differential evolution (MBDE) inspired by the idea of estimation of distribution and differential evolution algorithms to solve the multi-constrained unit commitment problem. The updating strategy of the standard differential evolution (DE) is reserved in the proposed MBDE so that the excellent characteristics of DE, such as easy implementation and parameter tuning, are inherited. A new probability estimator operator is applied in MBDE, which efficiently preserves the diversity of population and boost the global search ability. Further, this paper also explores various priority methods based on unit characteristics for unit scheduling and allocating the power to committed units. The effectiveness of proposed method has been investigated on the small scale to large scale power systems ranging from 10 to 100 generating units for 24hours scheduling period. Results are validated by performing comparison with previously published research papers. Simulated results achieved by proposed method are found superior to the previously reported algorithms used to solve the unit commitment problem. A Wilcoxon signed rank test for paired samples also proves predominance of the proposed MBDE algorithm.
学科与关键词
- A synergy of binary differential evolution and binary local search optimizer to solve multi-objective profit based unit commitment problemDhaliwal, Jatinder Singh*; Dhillon, J. S.APPLIED SOFT COMPUTING, 2021, 107: 107387.被引 17
- A binary differential evolution algorithm learning from explored solutionsChen, Yu*; Xie, Weicheng; Zou, XiufenNEUROCOMPUTING, 2015, 149: 1038-1047.被引 38
- Optimizing a unit commitment problem using an evolutionary algorithm and a plurality of priority listsTsalavoutis, Vasilios A.*; Vrionis, Constantinos G.; Tolis, Athanasios I.Operational Research, 2021, 21(1): 1-54.被引 5
