An Efficient Q-Learning-Based Multi-Objective Intelligent Hybrid Genetic Algorithm for Mixed-Model Assembly Line Efficiency

In real-life mixed-model assembly lines, multiple problems collectively affect the final production’s performance. In this study, mixed-model assembly lines integrated with balancing and sequencing problems are considered simultaneously solved. A comprehensive mathematical model is formulated to evaluate the current multi-objective problem. An intelligent hybrid genetic algorithm (IHGA) is proposed to solve the integrated mixed-model assembly line balancing and sequencing problem. The performance of the proposed algorithm is triggered by integrating heuristic rules through a generation gap mechanism which helps in reducing search space without succumbing to local optima. Additionally, parametric tuning of the algorithm is performed using Q-learning, enabling adaptive optimization through reinforcement learning. This helps to enhance computational efficiency and achieve robust performance of the proposed algorithm. The performance of the IHGA algorithm is rigorously compared with existing approaches, including a non-dominated sorting genetic algorithm, multi-objective artificial bee colony, multi-objective particle swarm optimization, multi-objective evolutionary algorithm based on Decomposition, and multi-objective grey wolf optimizer. Results demonstrate the superior performance of the proposed algorithm across various metrics, showcasing its efficacy in optimizing mixed-model assembly lines, where symmetry in task allocation and sequencing can significantly enhance operational efficiency in contemporary industrial settings. Additionally, a real-life case study is solved to validate the empirical applicability of the proposed IHGA. The extensive experimental analysis notably shows that the proposed IHGA outperforms the existing methods.