Research progress on Autonomous Navigation Technology of Agricultural Robot

Agricultural robot technology is now developing explosively. Autonomous navigation is the key technology of agricultural robots. Compared with that of driverless cars, the environment of the agricultural robot is complex and diverse, and their navigation modes and core technologies are significantly different. Our research team of the agricultural robot, Jiangsu University has proposed the overall technical framework for the autonomous navigation of agricultural robots in fields, orchards, and greenhouses, and has made important research progress: (1) Aiming at the large-scale farms, we have broken through the autonomous forward-turn planning and vehicle-working part cooperative control technology based on Beidou RTK positioning, and pioneered the unmanned working in paddy fields in China. (2) Since pre-built map could not meet the need of a cross-regional operation, and the global planning method based on satellite positioning makes it complex to realize multi-machine parallel operation or master-slave collaboration, the portable base station of autonomous navigation technology based on UWB sensors of agricultural machinery in the field was put forward. It has made great breakthroughs in the realization of cross-regional single-machine operation technology, distributed multi-machine parallel operation technology, and cluster master-slave cooperative operation technology.(3) Aiming at harvest operations of crops, the local-global integrated navigation method based on depth vision and magnetic information was constructed, the harvesting path planning method based on harvesting edge, lodging area, and magnetic information fusion was proposed. It can realize autonomous navigation and self-adaptive harvest of different sites and all kinds of lodge morphology crops. (4) Aiming at the vineyard environment of the slender discreted stems, the navigation method combining laser scanning with an electronic compass was put forward. In this technology, autonomous diagnosis of pose/posture of the vehicle in the environment, fast driving to navigation line, and route planning in real-time. It can realize the robot's autonomous work in the vineyard. (5) Aiming at the environment of the continuous structured corridor in the greenhouse and animals/plants factory environment, the navigation method of sided arc array of photoelectric switches was put forward. It can realize curb-following navigation based on combination judgment from the numbers of triggered photoelectric switches and the center sequence number of triggered photoelectric switches to accomplish heading deviation and lateral deviation synchronous detection. Our research team of the agricultural robots, Jiangsu University is accelerating research on different types of autonomous navigation in technology and equipment, meanwhile, it is hopeful to provide differentiated navigation solutions for the rapid development of agricultural robot technology in China.