Modified Fruit Fly Optimization Algorithm for Robotic Assembly Line Balancing

Aiming at the NP-hard characteristics and complex constraints of the second type of Two-Sided Assembly Line Balancing Problem (RTALBP-II), a Modified Fruit Fly Optimization Algorithm (MFOA) is proposed. It balances global exploration and local exploitation through three improvements: 1) a dynamic hybrid search mechanism that combines an adaptive step-size strategy with an olfactory-visual synergistic search model for Drosophila populations; 2) a multimodal synergistic mutation operator that integrates Levy flights, differential evolution, and Gaussian variation to enhance population diversity; 3) constraints on the feasibility of constraint-aware hierarchical coding with neighborhood search to ensure a solution. Experiments to run standard test cases show that MFOA outperforms the Fruit Fly Optimization Algorithm (FOA), Simulated Annealing Algorithm (SA), Dung Beetle Optimization Algorithm (DBO), and Gray Wolf Optimization Algorithm (GWO) in terms of solution quality, convergence speed, and robustness, and maintains stable optimization capability under complex constraints. The study verifies the significant advantages of MFOA in RTALBP-II and provides an efficient optimization method for the production of intelligent manufacturing systems.