Constrained Multiobjective Optimization Algorithm Based on Immune System Model.

Area of Science:

• Computational Intelligence
• Optimization Algorithms
• Bio-inspired Computing

Background:

• Multiobjective optimization problems (MOPs) with multimodal nonlinear constraints present significant computational challenges.
• Existing algorithms often struggle with convergence and maintaining diversity in complex search spaces.

Purpose of the Study:

• To propose a novel immune optimization algorithm inspired by the biological immune system.
• To address the challenges of multiobjective optimization problems featuring multimodal nonlinear constraints.

Main Methods:

• The algorithm divides the population into feasible nondominated and infeasible/dominated groups.
• Feasible individuals undergo clone and hypermutation; infeasible individuals are improved via crossover and mutation.
• A transformation technique and crowded-comparison strategy accelerate convergence and population generation.

Main Results:

• The proposed algorithm was evaluated on benchmark constrained multiobjective optimization problems.
• Performance was compared against state-of-the-art algorithms using inverted generational distance and hypervolume indicators.
• The new method achieved competitive and statistically significant improvements on most benchmark instances.

Conclusions:

• The proposed immune optimization algorithm is effective for solving constrained multiobjective optimization problems.
• The algorithm shows superior performance compared to existing methods, particularly on complex benchmark suites.
• This bio-inspired approach offers a promising direction for advanced optimization techniques.