摘要
设计了联合收获机机器视觉导航控制硬件系统,提出了基于直方图的导航路径融合算法,在简化的二轮车运动学模型的基础上,设计了模糊PD控制器。麦田试验结果表明：旋转投影算法配合直方图融合算法能够有效地检测收获与未收获边界线,提高了控制系统的准确性与抗干扰性;在正常小麦收获速度挡位下,割幅变化范围在0.18m以内,在有外界扰动下（人为方向盘扰动）,能在2~5 s内回到正常的跟踪误差范围内。
The vision navigation control hardware system was set up in the combine harvester and a histogram fusion algorithm for multi-frame navigation lines was put forward. On the basis of kinematics of simplified two-wheel vehicle, model one fuzzy PD controller was constructed. The tests in wheat field show that the rotation and projection algorithm with histogram fusion algorithm is valid to recognize the cut/uncut line, and can improve the precision of the system and robustness, the change of harvesting width is in 0. lSm during the working speed, and it could come back to the right path in 2 -5 s when given manual disturbance using steering wheel.
出处
《农业机械学报》
EI
CAS
CSCD
北大核心
2010年第5期137-142,共6页
Transactions of the Chinese Society for Agricultural Machinery
基金
"十一五"国家科技支撑计划资助项目(2007BAD89B04)
关键词
联合收获机
视觉导航
控制系统
直方图融合
Combine harvester, Vision navigation, Control system, Histogram fusion
