Authors

Abstract

In order to rationally plan the amount of tasks and task areas for each agricultural robot in the farm, a cloud-side collaborative task allocation scheme is proposed. The cloud platform divides farm tasks based on field obstacles and extracts the center of gravity prime points for each farm task; plasmas as regional task target points through dynamic genetic algorithms for near-field aggregation, after accelerating the solution process by dynamic crossover and variational operators, the Metropolis criterion is introduced to eliminate the local optimal solution of the algorithm and obtain the globally optimal allocation solution. Simulation experiments show that the optimal allocation reduces 9.21%, 5.66%, and 7.21% in the total cost compared to the random allocation, and the feasibility of the algorithm is proved experimentally. Reasonable task allocation can improve the overall production efficiency of agriculture, which is informative for unmanned farms operating in large areas.

Abstract in Chinese

为合理规划农场中各个农业机器人任务量与任务区域，提出一种云边协同的任务分配方案。云平台依据田间障碍物划分农田任务，提取各个农田任务重心质点；质点作为区域任务目标点通过动态遗传算法进行近场聚合，经动态交叉、变异算子加速求解过程，引入Metropolis准则，剔除算法局部最优解，得到全局最优分配解。仿真实验表明，最优分配较随机分配在总代价中减低了9.21%、5.66%、7.21%，并通过实验证明了算法的可行性。合理的任务分配，可提高农业整体生产效率，对无人农场大区域作业具有参考价值。