.png){kind=icon}
.png){kind=icon}
Return
An accelerated method for solving constrained multi-objective optimization
.png){kind=icon}
PRE
AI
DOI：10.22124/jmm.2025.30721.2753.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
A novel non-parametric algorithm is introduced for solving constrained multi-objective optimization problems. At each iteration, a convex subproblem is solved to determine the search direction, while a non-monotone line search technique is used to determine the step size. An adaptive acceleration term, computed from changes in the search directions, is incorporated to scale the step and dynamically enhance convergence performance. The algorithm's effectiveness relies on a diverse set of initial feasible solutions to accurately approximate the non-dominated boundary. Benchmark tests validate the approach, with Pareto fronts compared to those obtained using the Zoutendijk method. Numerical evaluations demonstrate superior performance in terms of convergence rate and solution quality. The algorithm is also applied to a real-world engineering design problem involving speed reduction, highlighting its computational efficiency and robustness in practical applications.
Keywords:
Constrained multi-objective optimization problems
feasible direction methods
line search techniques
pareto critical point
Journal
J
IF:
0.8
Papers:
55
Citations:
0