A Generalized Framework for Multi-objective-based Constraint Handling Technique

This paper introduces a novel approach to handling constraints in multi-objective optimization problems by treating constraints as objectives. Our study consists of three main contributions. First, we present a generalized framework for multi-objective-based constraint handling techniques, which integrates multi-objective-based methods with the constraint dominance principle. This framework systematically incorporates various aspects of constraint violations into the objective function, enhancing optimization strategies. Second, we propose the constraint violation ratio, a metric that uses a constraint weight vector to quantify the severity of violations. Third, we develop the constraint diversity factor, an adaptive version of the constraint weight vector, which adapts as the frequency of constraint violations changes. Extensive experimental evaluations validate the effectiveness of our methods, demonstrating their applicability across a wide range of constrained optimization scenarios.