Collective information-based particle swarm optimization for multi-fuel CHP economic dispatch problem

• Published in: Knowledge-based systems Vol. 248; p. 108902
• Main Authors: Chen, Xu, Li, Kangji
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier B.V 19.07.2022; Elsevier Science Ltd
• Subjects: Algorithms; Cogeneration; Collective intelligence technology; Combined heat and power economic dispatch; Convergence; Evolutionary computation; Fuels; Intelligence (information); Multi-fuel options; Optimization algorithms; Particle swarm optimization; Power dispatch; Search methods; Swarm intelligence; Multi-fuel options; Collective intelligence technology; Combined heat and power economic dispatch; Particle swarm optimization
• ISSN: 0950-7051; 1872-7409
• DOI: 10.1016/j.knosys.2022.108902

Multi-fuel combined heat and power economic dispatch (MF-CHPED) is a highly non-convex and challenging optimization problem in power system operation. The traditional particle swarm optimization algorithms often suffer from premature convergence and low efficiency when solving the MF-CHPED problem. Collective intelligence is a cutting-edge technology in the evolutionary computation. In this paper, using the concept from collective intelligence, a novel collective information-based particle swarm optimization (CIBPSO) algorithm is proposed. In CIBPSO, two new collective information (CI)-based strategies namely CI-based particle search and CI-based elite fine-tuning are developed. First, in the CI-based particle search strategy, the global best position in traditional particle update equation is replaced by a newly-defined CI-based best solutions, which helps to enhance swarm diversity and alleviate premature convergence. Second, in the CI-based elite fine-tuning strategy, more computing resources are assigned to the elite solutions by using the information of CI-based best solutions, which is beneficial to improve the search efficiency. The proposed CIBPSO algorithm is applied to solve four different MF-CHPED problems considering different operating constraints. By comparing with six well-regarded optimization algorithms, it is found that CIBPSO achieves the overall best results in terms of solution accuracy, stability and convergence. In addition, the effectiveness of the two new CI-based strategies is discussed. •Collective information-based particle swarm optimization (CIBPSO) algorithm is proposed.•CI-based particle search and CI-based elite fine-tuning strategies are designed.•CIBPSO is applied to solve four multi-fuel combined heat and power economic dispatch problems.•Simulation results demonstrate the effectiveness of the proposed CIBPSO algorithm.