.png){kind=icon}
.png){kind=icon}
Return
A stage-based adaptive penalty method for constrained evolutionary optimization
.png){kind=icon}
PRE
AI
DOI：10.1016/j.eswa.2024.126351.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Selecting an appropriate penalty coefficient is imperative when using penalty functions to address constrained optimization problems. This paper introduces a novel approach, named Stage-based Adaptive Penalty Method (SAPDE), which adjusts the penalty coefficient based on overall trends. Unlike previous methods that rely on decomposition or fuzzy logic, SAPDE develops a stage-based strategy for adjusting the penalty coefficient, ensuring a balance between exploration and exploitation. In the early stage, where no feasible solutions are present in the population, static coefficient values are assigned to each candidate solution. In the later stage, an exponential function is used to determine the penalty coefficients, which gradually decrease as the evolutionary generations increase. During population evolution, the differential evolution (DE) is employed to generate offspring populations, combined with a mutation scheme to further enhance population diversity. Additionally, a repair strategy combining feasibility rules and restart strategies is used to address some of the more challenging problems. Tests on the benchmark suites and six engineering design problems demonstrate that the proposed SAPDE is competitive.
Keywords:
Constraint handling techniques
Penalty function approach
Mutation schemes
Repair strategy
Journal
IF:
7.5
Papers:
2.9W
Citations:
10.2W