| 基于DEM的无人驾驶拖拉机前馈稳速控制 |
| |
| 引用本文： | 吴才聪,高星宇,吴思贤,梁永浩,徐海松,翟卫欣.基于DEM的无人驾驶拖拉机前馈稳速控制[J].农业工程学报,2024,40(10):24-31. |
| |
| 作者姓名： | 吴才聪 高星宇 吴思贤 梁永浩 徐海松 翟卫欣 |
| |
| 作者单位： | 中国农业大学信息与电气工程学院，北京 100083;农业农村部农机作业监测与大数据应用重点实验室，北京 100083 |
| |
| 基金项目： | 国家重点研发计划项目（2021YFB3901302） |
| |
| 摘 要： | 针对农田坡度变化影响无人驾驶农机行驶速度稳定性，进而降低播种均匀性和肥药施用精度等问题，该研究设计了一种基于农田数字高程模型（digital elevation model，DEM）和前馈控制策略的拖拉机稳速控制方法。首先建立坡地干扰补偿模型，基于拖拉机实时位置从农田DEM中提取前方作业路径的坡度信息，计算拖拉机前方目标速度补偿量，实现拖拉机行驶的稳速控制。以DF1204无级变速拖拉机为试验平台，在中国农业大学烟台研究院开展3组不同目标速度的上坡、平地和下坡行驶对比试验。试验结果表明，拖拉机以目标速度4.0、6.0和8.0 km/h行驶时，上坡行驶的实测速度均值分别为4.03、5.94和7.85 km/h，平地行驶的实测速度均值分别为4.04、6.02和8.03 km/h，下坡行驶的实测速度均值分别为4.00、6.10和8.19 km/h，与对照组相比，在上坡、平地和下坡行驶时的速度均方根误差分别降低了46.63%、21.92%和37.15%，试验组上坡和下坡行驶的实测速度均值更接近目标速度。所提方法可有效提高无人驾驶拖拉机在起伏农田的稳速控制精度，有助于提高农机作业质量。
|
| 关 键 词： | 拖拉机 无人驾驶 DEM 无级变速 前馈控制策略 稳速控制 |
| 收稿时间： | 2024/1/22 0:00:00 |
| 修稿时间： | 2024/2/18 0:00:00 |
Feedforward stabilized speed control of driverless tractor based on DEM |
| |
| WU Caicong,GAO Xingyu,WU Sixian,LIANG Yonghao,XU Haisong,ZHAI Weixin.Feedforward stabilized speed control of driverless tractor based on DEM[J].Transactions of the Chinese Society of Agricultural Engineering,2024,40(10):24-31. |
| |
| Authors: | WU Caicong GAO Xingyu WU Sixian LIANG Yonghao XU Haisong ZHAI Weixin |
| |
| Institution: | College of Information and Electrical Engineering, China Agricultural University, Beijing 100083, China;Key Laboratory of Agricultural Machinery Monitoring and Big Data Applications, Ministry of Agriculture and Rural Affairs, Beijing 100083, China |
| |
| Abstract: | Dynamic speed can often cause the fluctuations of unmanned agricultural machines in farmland slopes. Agricultural operations can be significantly impacted, such as seeding and fertilizer application. The stability of planting and the precision of fertilizer application can depend mainly on the driving speed of driverless tractors in undulating farmland. Particularly, the absence of visual cues from the driver can lead to inevitable pre-control of the throttle when tractors operating. In this study, the robust control system of steady speed was proposed using the Digital Elevation Model (DEM). An advanced feed-forward control was also utilized at the longitudinal slopes. A systematic investigation was then made to assess the influence of farmland slope on the tractor speed. The occurrence of slope acceleration was induced by gravity during both uphill and downhill maneuvers in the tractor driving. The new control of the tractor was also used to predict the terrain slope ahead. Consequently, the speed of the tractor was adjusted for the target speed, as the terrain slope increased. An extended duration was avoided to reach the preset velocity. Alternatively, the speed readily surpassed the target speed on the decreasing slopes, resulting in frequent braking and adverse effects on the dynamics of vehicles. DEM data was utilized to extract elevation information from the terrain. The longitudinal slope feed-forward control was also introduced to dynamically obtain the pre-scanning distance. The pre-scanning time was configured to integrate with the current driving speed. At the same time, the slope value of the subsequent driving path was calculated to determine the compensation of speed on the slope. Ultimately, effective control of the tractor''s speed was achieved to adjust the target speed in real time. The optimal performance and stability were obtained across diverse terrain gradients. DF1204 CVT tractor was taken as a test platform. Three tests were conducted with different target speeds for uphill, flat, and downhill at the Yantai Research Institute of China Agricultural University. The test results showed that when the tractor was traveling at the target speeds of 4.0, 6.0, and 8.0 km/h, the mean values of the measured speeds were 4.03, 5.94, and 7.85 km/h for uphill driving, 4.04, 6.02, and 8.03 km/h for level driving, and 4.00, 6.10, and 8.19 km/h for downhill driving, respectively. The speeds closely approached the target values, indicating the efficacy of the control system to regulate the tractor speeds. The speed accuracy was also improved with the mean speeds of the experimental group notably closer to the target speeds for both uphill and downhill driving scenarios. Moreover, the root mean square errors (RMSE) were reduced under all driving conditions, with decreases of 46.63%, 21.92%, and 37.15% for uphill, level ground, and downhill driving, respectively, compared with the control group. The speed stability was enhanced to reduce the speed fluctuations, thereby improving the overall tractor performance and operational efficiency in agricultural settings. The feedforward control of longitudinal gradient speed also surpassed the original, in terms of the speed stability and path tracking. The steady speeds with high precision were maintained in the autonomous agricultural machinery under undulating farmland environments. The tractor efficiency was then promoted to reduce the operational fluctuations in practical agricultural operations. The optimal speeds can be used to realize smoother and more productive farming in the various tasks of streamlining, such as planting and fertilization. Moreover, the operational disruptions can be minimized to maintain high-performance standards. The finding can greatly contribute to the overall quality and reliability of agricultural machine operations. Technical support can also be offered to improve agricultural productivity and sustainability in sustainable production. |
| |
| Keywords: | tractors autonomous driving DEM continuously variable transmission feedforward control steady speed control |
|
| 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 |
|
