| 基于BP_Adaboost的农机行驶轨迹状态识别 |
| |
| 引用本文： | 李亚硕,赵博.基于BP_Adaboost的农机行驶轨迹状态识别[J].农业工程,2022,12(4):48-53. |
| |
| 作者姓名： | 李亚硕 赵博 |
| |
| 作者单位： | 1.中国农业机械化科学研究院集团有限公司，北京 100083 |
| |
| 摘 要： | 农业机械运动轨迹不仅包括田间作业轨迹，还包括道路行驶轨迹。有效区分农机行驶过程中田间作业和道路行驶等作业轨迹，可精确划分有效作业地块和精准评估农机作业效率，从而实现农机的远程智能管理。通过对农机轨迹点属性的分析，提取典型特征数据，利用BP_Adaboost方法建立的训练模型对农机轨迹点进行识别，将道路与田间交界处易错轨迹点重新标记后再次训练，轨迹识别正确率达96.89%。该方法既避免了传统聚类算法对阈值和参数依赖的问题，也有效解决了将道路行驶轨迹误识别为田间作业轨迹的难题。
|
| 关 键 词： | 农机作业轨迹 BP_Adaboost 聚类算法 |
| 收稿时间： | 2022/3/22 0:00:00 |
| 修稿时间： | 2022/4/1 0:00:00 |
Recognition of agricultural machinery operation trajectory Based on BP_ AdaBoost |
| |
| Li Yashuo and Zhao Bo.Recognition of agricultural machinery operation trajectory Based on BP_ AdaBoost[J].Agricultural Engineering,2022,12(4):48-53. |
| |
| Authors: | Li Yashuo and Zhao Bo |
| |
| Institution: | 1.Chinese Academy of Agricultural Mechanization Sciences Group Co.，Ltd.， Beijing 100083，China2.State Key Laboratory of Soil Plant Machinery System Technology，Beijing 100083，China3.China Machinery Construction Engineering Co.，Ltd.，Beijing 100083，China4.Center for Agriculture Mechanization Technology Extension of Hubei Province，Wuhan Hubei 430017， China |
| |
| Abstract: | The movement track of agricultural machinery includes not only the field operation track, but also the road driving track. Effectively distinguish the operation tracks of field operation and road driving during the driving of agricultural machinery, accurately divide the effective operation plots and accurately evaluate the operation efficiency of agricultural machinery, so as to realize the remote intelligent management of agricultural machinery. This paper extracts typical characteristic data by analyzing the attributes of agricultural machinery trajectory points, and uses the training model established by the method of BP_ AdaBoost to recognize the track points of agricultural machinery. After re marking the error prone track points at the junction of road and field, it is trained again. The correct rate of track recognition is 96.89%. It not only avoids the dependence of traditional clustering algorithm on threshold and parameters, but also effectively solves the problem of mistaking road driving trajectory into field operation trajectory. |
| |
| Keywords: | Agricultural machinery operation track BP_Adaboost clustering algorithm |
|
| 点击此处可从《农业工程》浏览原始摘要信息 |
| 点击此处可从《农业工程》下载免费的PDF全文 |
