Unrelated Parallel Machine Scheduling Based on an Enhanced Whale Optimization Algorithm

In the context of intelligent manufacturing, the unrelated parallel machine scheduling problem (UPMSP) has become an NP-hard challenge due to machine heterogeneity, task dynamics, and multi-objective conflicts. Traditional algorithms face the bottlenecks of premature convergence and insufficient adaptability. To address this issue, this paper proposes an improved whale optimization algorithm (IWOA), which achieves efficient solution through a four-dimensional strategy optimization: a discrete continuous mapping mechanism is established based on random key encoding to convert job allocation into a machine preference value optimization problem; the initial population diversity is enhanced by combining Logistic chaotic mapping, and adaptive weights are introduced to dynamically balance global exploration and local exploitation; the Levy flight strategy is embedded to avoid local optima, and job swapping mutation is used to maintain population diversity. Theoretically, a hybrid integer programming model integrating machine preference encoding is constructed to break through the performance bottleneck of traditional meta-heuristic algorithms; applicationally, a scheduling optimization system supporting dynamic constraint expansion is developed, providing a core module for manufacturing decisions. Experiments show that IWOA improves the convergence speed by 15% in a 1000 × 100 scale scenario, reduces the maximum completion time by 2.4%, and is significantly superior to the standard algorithm. Future research will focus on distributed parallel acceleration, dynamic rescheduling based on deep reinforcement learning, and collaborative optimization with digital twins, promoting autonomous decision-making throughout the intelligent manufacturing chain.