The intelligent control system of the Dongfanghong intelligent tractor primarily operates based on sensor fusion, intelligent decision-making, and automatic control technologies.
While specific configurations and functions may vary between tractor models, the following are some common operating methods:
Automatic navigation control based on RTK-DGPS: Taking the Dongfanghong X-804 tractor as an example, its automatic navigation system consists of an RTK-DGPS navigation sensor, a navigation controller, an electronically controlled hydraulic steering circuit, a steering control, a steering wheel angle sensor, and the tractor itself. The navigation sensor uses a Trimble 5700 GPS receiver system, achieving a horizontal positioning accuracy of 1cm + 1ppm RMS and a vertical positioning accuracy of 2cm + 1ppm RMS.
The navigation controller uses an ATMEL AT91SAM9261 microprocessor to collect and process DGPS positioning data and make steering control decisions.
The steering control controller uses a C8051F040 microcontroller to collect steering wheel angle data and perform steering control.
The system uses RTK-DGPS to obtain the tractor's precise position.
The navigation controller calculates the deviation between the preset working path and the current position, then controls the steering wheel angle via the electronically controlled hydraulic steering circuit for automatic navigation.
A steering wheel angle detection sensor provides real-time feedback on the steering angle, forming a closed-loop control system to ensure steering accuracy.
Multimodal sensor fusion obstacle avoidance system: For example, the Dongfanghong intelligent tractor used in northern Xinjiang boasts an obstacle avoidance system that utilizes a multimodal sensor fusion architecture, including lidar, millimeter-wave radar, and a visual recognition system.
The lidar constructs a 3D point cloud map in real time to identify obstacles in the field, achieving an outline accuracy of ±2cm and a minimum detection size of 5cm.
A 77GHz forward-facing long-range millimeter-wave radar detects moving obstacles within a range of 200m, while a 24GHz side-facing short-range radar covers a 50m range on either side of the vehicle, preventing scratches during inter-row work in cotton fields.
Binocular stereo vision, based on deep learning, distinguishes crops from weeds with 99.3% accuracy.
The thermal imaging module detects heat sources in living organisms at night or in dusty environments.
Sensor data is processed by Huawei's MDC 610 automotive-grade computing unit.
Using an improved A* algorithm, it generates an obstacle avoidance path within 1 second. It automatically adjusts the detour radius based on obstacle type (living organisms > hard obstacles > soft obstacles) to minimize soil compaction.
Precision Agriculture Platform Based on the Beidou Satellite System: The Dongfanghong Precision Agriculture Platform, powered by the Beidou satellite system and location-based broadcasting (LBS) base stations, enables agricultural machinery positioning, scientific operation, and trajectory recording.
The platform collects and manages Beidou agricultural machinery automatic equipment information, providing real-time visibility into the operation location, quality of work, alarms, and maintenance status of agricultural machinery, enabling centralized management and scientific scheduling.
The vehicle-mounted part of the Beidou agricultural machinery automatic assisted driving system consists of a satellite receiving antenna, an on-board display screen, a Beidou high-precision on-board positioning terminal, a driving controller, a hydraulic valve (or steering wheel), an angle sensor, etc.
Using the high-precision Beidou satellite positioning system, the driving controller controls the hydraulic system of the agricultural machinery, allowing the agricultural machinery to automatically travel according to the set route and display relevant graphical information on the on-board display.
