Nonconvex/nonsmooth economic load dispatch using modified time‐varying particle swarm optimization. (22nd April 2019)

• Record Type: Journal Article
• Authors: Parouha, Raghav Prasad
• Is Part Of: Computational intelligence. Volume 35:Number 4(2019)
• Page Start: 717
• Page End: 744
• Publication Date: 2019-04-22
• Subjects: economic load dispatch (ELD) -- inertia weight -- optimization -- particle swarm optimization (PSO) -- time‐varying acceleration coefficients; Artificial intelligence -- Periodicals; Computational linguistics -- Periodicals; 006.3
• DOI: 10.1111/coin.12210
• Languages: English
• ISSNs: 0824-7935

Abstract

Abstract: In this paper, a modified time‐varying particle swarm optimization (MTVPSO) is proposed for solving nonconvex economic load dispatch problems. It is a variant of the traditional particle swarm optimization (PSO) algorithm. In an MTVPSO, novel acceleration coefficients for cognitive and social components are presented as linear time‐varying parameters in the velocity update equation of the PSO algorithm. In the early stages of the optimization process, it improves the global search capability of particles and directs the global optima at the end stage. Additionally, a linearly decreased inertia weight is introduced in an MTVPSO, instead of a fixed constant value, which helps improve the diversity of the population. Through this modification mechanism in PSO, the proposed algorithm has a higher probability of avoiding local optima, and it is likely to find global optima more quickly. Six complex benchmark functions have been used to validate the effectiveness of the proposed algorithm. Furthermore, to demonstrate its efficiency, feasibility, and fastness, six different cases (3‐, 6‐, 13‐, 15‐, and 40‐unit systems and one large‐scale Korean power 140‐unit system) of the economic load dispatch problem are solved by an MTVPSO. The results of the proposed algorithm have been compared with state‐of‐the‐art algorithms. It was found that the proposed MTVPSO can deliver better results in terms of solution quality, convergence characteristics, and robustness.