Research on emergency scheduling based on improved genetic algorithm in harvester failure scenarios.

In response to the issue of harvesting machine failures affecting crop harvesting timing, this study develops an emergency scheduling model and proposes a hybrid optimization algorithm that combines a genetic algorithm and an ant colony algorithm. By enhancing the genetic algorithm's crossover and mutation methods and incorporating the ant colony algorithm, the proposed algorithm can prevent local optima, thus minimizing disruptions to the overall scheduling plan. Field data from Deyang, Sichuan Province, were utilized, and simulations on various harvesting machines experiencing random faults were conducted. Results indicated that the improved genetic algorithm reduced the optimal comprehensive scheduling cost during random fault occurrences by 47.49%, 19.60%, and 32.45% compared to the basic genetic algorithm and by 34.70%, 14.80%, and 24.40% compared to the ant colony algorithm. The improved algorithm showcases robust global optimization capabilities, high stability, and rapid convergence, offering effective emergency scheduling solutions in case of harvesting machine failures. Furthermore, a visual management system for agricultural machinery scheduling was developed to provide software support for optimizing agricultural machinery scheduling.