Application of evolutionary computation to decision-making in production processes

In a highly industrialized and data-driven society, ensuring product quality and safety has become a top business priority, especially in critical areas such as medical devices and electronics. This study explores the challenges associated with end-of-life (EOL) sampling decisions in complex assembly processes, aiming to achieve an optimal balance between product quality and cost-effectiveness. By comprehensively analyzing the sampling cost, quality risk, and disassembly feasibility, a novel sampling strategy is proposed for electronics, medical and other industries. The proposed method utilizes the Differential Evolutionary algorithm (DE) and the Q-learning algorithm with dynamic exploration rate (ϵ) to optimize the decision model. And we compare DE with other evolutionary algorithms, such as Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) through experimental analysis. The results show that the proposed method can effectively provide a robust sampling decision-making strategy for enterprises and improve the product qualification rate while reducing the cost.