Quantum‐Behaved Particle Swarm Optimization Based on Concentration Selection Probability Assignment Weights for Power System Economic Dispatch
Miao Wang,Zhiao Cao,Yuchun Li
DOI: https://doi.org/10.1002/tee.24015
IF: 0.923
2024-02-14
IEEJ Transactions on Electrical and Electronic Engineering
Abstract:The economic dispatch (ED) problem is very important in the economics of power systems, and a suitable dispatch algorithm can help power plants save huge amounts of money and use energy efficiently. To ensure the economics of the power system dispatch strategy, this paper proposes a quantum‐behaved particle swarm improvement algorithm for solving the ED problem. First, the algorithm addresses the problem that the classical quantum‐behaved particle swarm optimization (QPSO) tends to be premature and falls into local optimality. We propose a method to calculate the mean best position by assigning weights to different particles, by adopting the idea of concentration selection in the immune concentration regulation mechanism, and designing a Weight‐Concentration selection probability function, so that those particles located outside the local optimal solution region but with smaller fitness value and better evolutionary trend can be retained. Second, to make the algorithm focus on global search at the beginning of the iteration and more focused search at the end to enhance the neighborhood search capability at the later stage, a decreasing function of the slope K in the Weight‐Concentration selection probability function is designed so that the influence of the weighting term on the calculation of the mean best position decreases gradually. The algorithm is evaluated with four power systems, and the results show that the overall performance of the improved algorithm is better than other algorithms in solving the ED problem in terms of minimum solution, convergence, and stability. © 2024 Institute of Electrical Engineer of Japan and Wiley Periodicals LLC.
engineering, electrical & electronic