| 智能农机多机协同收获作业控制方法与试验 |
| |
| 引用本文： | 满忠贤,何杰,刘善琪,岳孟东,胡炼,黄培奎,汪沛,罗锡文.智能农机多机协同收获作业控制方法与试验[J].农业工程学报,2024,40(1):17-26. |
| |
| 作者姓名： | 满忠贤 何杰 刘善琪 岳孟东 胡炼 黄培奎 汪沛 罗锡文 |
| |
| 作者单位： | 1. 华南农业大学南方农业机械与装备关键技术教育部重点实验室，广州 510642;2. 广东省农业人工智能重点实验室，广州 510642 |
| |
| 基金项目： | 国家重点研发计划项目（2021YFD2000600） |
| |
| 摘 要： | 为提高多台无人化智能收获机和运粮车协同作业效率，该研究以2台不同型号水稻收获机和1台运粮车为研究对象，开展了智能农机多机协同收获作业控制方法研究。根据协同作业控制决策约束条件，建立协同收获作业中有限个状态过程的改进型连续时间马尔科夫链模型。以减少非作业时间为优化目标，通过模型预测未来一段时间内每台收获机的卸粮时间，动态更新每台收获机的卸粮顺序和时间。仿真试验结果表明：本文控制方法相对于仓满后再召唤运粮车的卸粮方式有效减少了作业时间，协同收获任务的农机平均作业时间减少了13.58%。田间试验结果表明：智能农机多机协同作业控制方法实现了2台水稻收获机和1台运粮车协同自主作业，在场景1中，相对于仓满召唤卸粮模式，收获机1和收获机2非作业时间分别减少了71.25%和42%，收获效率提高了6.65%和5.22%；在场景2中，相对于仓满召唤卸粮模式，收获机1和收获机2非作业时间分别减少了77.64%和37.09%，收获效率提高了12.07%和5.78%。本文提出的控制方法可以实现收获-卸粮转运自主作业，减少了收获机的非作业时间，提高了作业效率，可为无人农场智能收获协同作业提供支撑。
|
| 关 键 词： | 农业机械 无人农场 收获 多机协同策略 运粮 |
| 收稿时间： | 2023/11/1 0:00:00 |
| 修稿时间： | 2023/12/26 0:00:00 |
Method and test for operating multi-machine cooperative harvesting in intelligent agricultural machinery |
| |
| MAN Zhongxian,HE Jie,LIU Shanqi,YUE Mengdong,HU Lian,HUANG Peikui,WANG Pei,LUO Xiwen.Method and test for operating multi-machine cooperative harvesting in intelligent agricultural machinery[J].Transactions of the Chinese Society of Agricultural Engineering,2024,40(1):17-26. |
| |
| Authors: | MAN Zhongxian HE Jie LIU Shanqi YUE Mengdong HU Lian HUANG Peikui WANG Pei LUO Xiwen |
| |
| Institution: | 1. Key Laboratory of Key Technology for South Agricultural Machine and Equipment, Ministry of Education, South China Agricultural University, Guangzhou 510642, China;2. Guangdong Provincial Key Laboratory of Agricultural Artificial Intelligence (GDKL-AAI), Guangzhou 510642, China |
| |
| Abstract: | This study aims to improve the cooperative operation efficiency of multiple unmanned intelligent harvesters and grain trucks. A control system of multi-machine cooperative harvesting was proposed in the intelligent agricultural machinery. Two types of rice harvesters and one-grain truck were also taken as the research objects. An improved model of a continuous-time Markov chain was established for the finite state processes in the collaborative harvesting operation, according to the constraints of control decisions. The unloading conditions of harvesters varied greatly over time. A grain truck only served one harvester at a time, while the service time depended on the total amount of grain harvested. The minimum total time of collaborative harvesting was highly required to configure the harvesting sequence of each harvester. Particularly, there was the tradeoff between harvesting first or unloading first in a collaborative grain unloading task. The time that the grain truck stayed in a certain state was independent of that in the system. The process of transferring from one state to another was also irrelated with the previous and subsequent state. The unloading sequence and time of each harvester were predicted and dynamically updated to reduce the non-operation time using the improved model. The simulation results show that the control system of multi-machine cooperative harvesting effectively reduced the non-operation time for the high working efficiency. The completion time of grain truck, harvesters 1 and 2 was reduced by 10.71%, 10.25%, and 17.28%, respectively, compared with the full warehouse calling for grain unloading. The average completion time of agricultural machinery was reduced by 13.58% for the cooperative harvesting task. The waiting time was also greatly reduced in the grain unloading area. The field experiments show that the control system of multiple intelligent agricultural machineries was realized in the cooperative autonomous operation of two rice harvesters and one-grain truck. In scenario 1, there was a similar time for harvesters 1 and 2 to call the grain truck. The non-operation time for the two harvesters was 3.45 and 6.95 min, respectively. Among them, the non-operation time was gradually reduced for harvester 2 to wait in the field. The non-operation time of harvesters 1 and 2 decreased by 71.25% and 42%, respectively, whereas, the harvest efficiency increased by 6.65% and 5.22%, respectively, compared with the full-call grain unloading. In scenario 2, the intelligent agricultural machinery was able to independently enter the grain unloading area. The average non-operation and the completion time of the two harvesters were 54.67 and 59.66 min, respectively. The non-operation time of harvesters 1 and 2 decreased by 77.64% (1.9 min) and 37.09% (5.85 min), respectively, whereas, the harvest efficiency increased by 12.07% and 5.78%, respectively. The autonomous operation of harvesting/unloading transportation was realized to reduce the non-operation time of harvester for the high efficiency. Two scenarios can be expected to allocate the other ones in the vast majority of fields. The findings can provide support to the intelligent harvesting collaborative operation in unmanned farms. The improved model can also be further optimized for the scenarios when a circle of harvesting operation cannot be completed in a larger field, due to the limited granary capacity of the harvester. |
| |
| Keywords: | agricultural machinery unmanned farm harvest multi-machine cooperation strategy transport grain |
|
| 点击此处可从《农业工程学报》浏览原始摘要信息 |
| 点击此处可从《农业工程学报》下载免费的PDF全文 |
|
