A Knowledge-Based Estimation of Distribution Algorithm for Product Assembly Line Balancing

Assembling is a critical process within the manufacturing industries, where the efficiency of assembly lines directly affects the manufacturing productivity. To improve the overall efficiency, this paper proposes a knowledge-based Estimation of Distribution Algorithm (kEDA) specifically designed for the Robotic Assembly Line Balancing Problem. The algorithm utilizes a sampling technique based on a tailored probability model within kEDA to determine the task-workstation assignment. Additionally, it incorporates knowledge-based solution repair and optimization operators to minimize cycle time. The simulation results of various problem scales demonstrate that our algorithm outperforms classic algorithms in terms of the cycle time.