Dynamic multi-objective differential evolution algorithm based on the information of evolution progress

The multi-objective differential evolution (MODE) algorithm is an effective method to solve multi-objective optimization problems. However, in the absence of any information of evolution progress, the optimization strategy of the MODE algorithm still appears as an open problem. In this paper, a dyna...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Science China. Technological sciences Ročník 64; číslo 8; s. 1676 - 1689
Hlavní autoři: Hou, Ying, Wu, YiLin, Liu, Zheng, Han, HongGui, Wang, Pu
Médium: Journal Article
Jazyk:angličtina
Vydáno: Beijing Science China Press 01.08.2021
Springer Nature B.V
Témata:
Algorithms
Engineering
Evolutionary algorithms
Evolutionary computation
Multiple objective analysis
Mutation
Optimization
Parameters
Probability theory
Wastewater treatment
information of evolution progress
optimization effect
convergence
multi-objective differential evolution algorithm
optimization speed
ISSN:1674-7321, 1869-1900
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:The multi-objective differential evolution (MODE) algorithm is an effective method to solve multi-objective optimization problems. However, in the absence of any information of evolution progress, the optimization strategy of the MODE algorithm still appears as an open problem. In this paper, a dynamic multi-objective differential evolution algorithm, based on the information of evolution progress (DMODE-IEP), is developed to improve the optimization performance. The main contributions of DMODE-IEP are as follows. First, the information of evolution progress, using the fitness values, is proposed to describe the evolution progress of MODE. Second, the dynamic adjustment mechanisms of evolution parameter values, mutation strategies and selection parameter value based on the information of evolution progress, are designed to balance the global exploration ability and the local exploitation ability. Third, the convergence of DMODE-IEP is proved using the probability theory. Finally, the testing results on the standard multi-objective optimization problem and the wastewater treatment process verify that the optimization effect of DMODE-IEP algorithm is superior to the other compared state-of-the-art multi-objective optimization algorithms, including the quality of the solutions, and the optimization speed of the algorithm.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1674-7321
1869-1900
DOI:10.1007/s11431-020-1789-9