A composite particle swarm optimization algorithm with future information inspired by non-equidistant grey predictive evolution for global optimization problems and engineering problems

Particle swarm optimization (PSO) and its numerous performance-enhancing variants are a kind of stochastic optimization technique based on collaborative sharing of swarm information. Many variants took current particles and historical particles as current and historical information to improve their...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Advances in engineering software (1992) Ročník 202; s. 103868
Hlavní autori: Hao, Rui, Hu, Zhongbo, Xiong, WenTao, Jiang, Shaojie
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Elsevier Ltd 01.04.2025
Predmet:
Future information
Improved particle swarm optimization algorithm
Non-equidistant grey prediction evolutionary algorithm
Particle swarm optimization algorithm
Future information
Improved particle swarm optimization algorithm
Non-equidistant grey prediction evolutionary algorithm
Particle swarm optimization algorithm
ISSN:0965-9978
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:Particle swarm optimization (PSO) and its numerous performance-enhancing variants are a kind of stochastic optimization technique based on collaborative sharing of swarm information. Many variants took current particles and historical particles as current and historical information to improve their performance. If future information after each current swarm can be mined to participate in collaborative search, the algorithmic performance could benefit from the comprehensiveness of the information including historical, current and future information. This paper proposes a composite particle swarm optimization algorithm with future information inspired by non-equidistant grey predictive evolution, namely NeGPPSO. The proposed algorithm firstly employs non-equidistant grey predictive evolution algorithm to predict a future particle as future information for each particle of a current swarm. Secondly, four particles including prediction particle, particle best and swarm best of the current swarm, and a history memory particle are used as guide particles to generate four candidate positions. Finally, the best one in the four positions is greedily selected as an offspring particle. Numerical experiments are conducted on 42 benchmark functions given by the Congress on Evolutionary Computation 2014/2022 and 3 engineering problems. The experimental results demonstrate the overall advantages of the proposed NeGPPSO over several state-of-art algorithms. •Integrate future information for the first time to improve PSO algorithm.•Present a composite particle swarm optimization with future information.•Apply NeGPE’s predictive ability to predict future information for particles.•The proposed algorithm outperforms the state-of-the-art on several test suites.•The proposed algorithm surpasses the comparative algorithms in engineering problems.
ISSN:0965-9978
DOI:10.1016/j.advengsoft.2025.103868