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

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 wellregarded 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. (C) 2022 Elsevier B.V. All rights reserved.