Abstract: Abstract: With the continuous advancement of land transfer policy in China, large-scale production and management of planting has been realized. The simultaneous harvesting of multiple combine harvesters not only improves the efficiency, but also has great significance for completing planting and harvesting ahead of schedule, reducing the risk of natural disasters and realizing grain storage. However, due to improper operation path planning, there are some problems such as operation conflict, lower efficiency and higher cost. Therefore, it is of great theoretical significance and practical value to study how to plan and optimize the conflict-free cooperative operation of multiple combine harvesters. Farmers are more concerned about when to finish their homework, while combine harvester operators are more concerned about the operation efficiency and the cost of combine harvesters. How to effectively organize multiple combine harvesters to cooperate in operation and how to save operation cost and shorten completion time is the key problem in combine harvester operation to meet the needs of farmers and operators to a greater extent. When multiple combine harvesters turn at the same headland and at the same time window, two combine harvesters may have the same or the opposite direction conflict, if the operation path planning is improper. To ensure that multiple combine harvesters could complete the harvest tasks of crop in a specified time with no conflicts and low operation cost in the process of cooperative operation, a conflict-free path optimization algorithm (IGA) for cooperative operation of multiple combine harvesters is proposed in the paper. IGA take the shortest total operation time and completion time as the optimization objective (the weights of total operation time and completion time are 0.4 and 0.6 respectively). The results showed that the total operation time and the completion time of the operation path optimized by traditional genetic algorithm (GA) in rectangular farmland are 33.72% and 34.00% lower than that of the operation path optimized by GA, respectively. When optimizing the operation path of trapezoidal farmland, the total operation time and the completion time of the operation path optimized by IGA are 29.93% and 30.00% lower than that of the operation path optimized by GA, respectively. It proves that the operation performance of the operation path optimized by IGA is better than that optimized by GA. Compared with side-by-side operation, the total operation time of rectangular and the trapezoidal farmland operation paths optimized by IGA decrease by 2.45% and 2.42% on average, and the completion time decrease by 2.29% and 2.02% on average. The results prove that IGA can always find better operation paths than side-by-side operation for different number of combine harvesters on the premise that there is no conflict between operation paths. With the number of combine harvesters increasing from 2 to 6, compared with side-by-side operation, the total operation time and the completion time of rectangular farmland operation path optimized by IGA respectively increase from 0.05% and 0.1% to 4.71% and 5.95%, and the decreasing extent is increasing, the effective field capacity is increasing. Similar to rectangular farmland operation, the total operation time and the completion time of trapezoidal farmland operation path optimized by IGA increase from 0.43% and 0.55% to 5.63% and 3.35% respectively. The study proves that IGA can always find better operation paths with no conflict for different number of combine harvesters than side-by-side operation, and the operation efficiency of combine harvester is higher. The research results provide a more reasonable conflict-free path for the cooperative operation of multiple combine harvesters.