Multi-objective Software Quality Optimization Evaluation Model Based on Genetic Algorithm

Software quality assessment is inherently a multi-objective problem, involving trade-offs among factors such as functionality, reliability, performance, maintainability, and security. Traditional evaluation models often rely on static or subjective weighting schemes, which fail to adapt to dynamic project requirements and cannot effectively balance conflicting objectives. This paper proposes a novel software quality evaluation model based on Genetic Algorithm (GA) optimization. The model integrates real-world quality metrics aligned with ISO/IEC 25010:2023 and utilizes GA to dynamically search for optimal weight distributions across quality dimensions. An adaptive fitness function is designed to penalize low-performing attributes while promoting balanced quality outcomes. Experimental validation using actual software metrics from enterprise and open-source projects shows that the proposed GA-based model significantly outperforms conventional methods, achieving higher evaluation accuracy, better convergence, and more robust handling of underperforming factors. The results demonstrate the model’s effectiveness in providing an intelligent, scalable, and flexible framework for multi-objective software quality assessment.