A Small-population Based Binary PSO Approach for Unit Commitment Problem

摘要

Unit commitment plays one of the most important roles in the operation planning of power systems. This paper proposes a Small-population based binary PSO (SBPSO) approach for Unit Commitment Problem (UCP). The proposed approach employs three operations, i.e., mutation, DE-acceleration and migration to achieve the more-near optimal solution faster. Moreover, a repair strategy that consists of repair to minimum online/offline time, spinning reserve and de-commitment of excessive units is proposed to obtain more feasible particles. The method is implemented and test using C++ programming. The test is carried out using a well-known 10 units system to verify the feasibility and effectiveness of the proposed method as well as the performance of the algorithm. The comparison results with other evolutionary methods clearly show the competitiveness of the proposed SBPSO approach.