Automatic driving tractor for orchard and control method of automatic driving tractor

Abstract

The invention discloses an orchard automatic driving tractor, which is a small tractor capable of freely driving between a fruit shed and a fruit tree, and the whole system of the tractor comprises an information acquisition system, an information fusion system, a path planning decision system, an execution control system, a communication system, a remote operation system and a power supply control system, according to the invention, a single-thread laser radar and a ZED binocular camera are combined for 3D mapping on the basis of not borrowing an expensive multi-thread laser radar, so that the system has the characteristics of low cost and high universality; and the tractor can be monitored at a remote terminal, and an automatic driving mode and a remote operation mode can be switched at any time, so that the safety of the whole operation of the tractor is ensured.

Description

[0001] The present application is a divisional application, the original application is named as an orchard automatic driving tractor, the application number is 202210063275.4, and the application date is January 20, 2022. TECHNICAL FIELD

[0002] The present application belongs to the field of intelligent agricultural machinery, and particularly relates to an orchard automatic driving tractor. BACKGROUND

[0003] Agriculture is the foundation of the national economy. In today's era of rapid development of science and technology, the fundamental way out for agricultural development lies in mechanization, automation and intelligence. Promoting unmanned agricultural equipment is the key to synchronously promoting the "four modernizations", the basis for improving productivity and transforming production methods, and also the key to China's participation in international agricultural competition. The development of automatic driving in China is relatively slow. Most of the existing automatic driving tractors have the problems of rigid path planning, inability to maintain a straight line in actual driving, and inability to adapt to day and night environments and different orchard environments, which need to be improved. SUMMARY

[0004] To solve the above problems, the present application discloses an orchard automatic driving tractor, which uses an information fusion system to fuse and process data obtained by laser radar, ZED binocular camera, global satellite positioning GPS and inertial navigation INS combination navigation, ultrasonic range finder, angle sensor and other equipment, to obtain more accurate data, improve the safety and efficiency of the tractor in actual work, so that the tractor can adapt to different orchard environments, the tractor driving path can ensure smooth and straight driving, and uninterrupted work can be ensured day and night, avoiding repeated work and improving work efficiency.

[0005] To achieve the above purpose, the technical scheme of the present application is as follows:

[0006] An orchard automatic driving tractor, comprising a tractor body and an automatic driving tractor system, the automatic driving tractor system comprising an information acquisition system, an information fusion system, a path planning and decision system, an execution control system, a communication system, a remote driving system and a power control system.

[0007] As a preferred embodiment of the present application, the tractor body is a small tractor capable of freely driving between a fruit shed and fruit trees, a different type of implement is mounted on the rear of the tractor body, and a guide rail is installed for the implement to slide left and right.

[0008] Further, the implement is a spraying implement, and a spray head on a machine arm that can be raised and lowered is installed on the spraying implement, so that the tractor can perform comprehensive spraying work on fruit trees of different heights and sizes at designated locations, greatly improving the efficiency of the work.

[0009] As the preferred of the present application, the power control system adopts a stable DC power source independent of the tractor itself battery power source, and is configured with three prompt lights, respectively indicating normal working state, power under-voltage state and charging state, prompting the tractor user to take timely corresponding measures, playing the role of power protection and protection of each power module.

[0010] As the preferred of the present application, the mutual communication between each system adopts CAN bus, greatly improving the real-time performance, stability, anti-electromagnetic interference ability of the overall system of the tractor and reducing the operation cost of the whole tractor.

[0011] The control method of the orchard automatic driving tractor comprises the following steps:

[0012] (1) The information acquisition system collects various data;

[0013] (2) The data is sent to the information fusion system, and the multi-source sensing data is fused and processed;

[0014] (3) The processed data is transmitted to the path planning decision system, the path planning decision system makes corresponding decision instructions, and the execution control system performs corresponding execution operations.

[0015] As the preferred of the present application, the information acquisition system of step (1) comprises a laser radar, a ZED binocular camera, a global satellite positioning GPS and an inertial navigation INS combined navigation, an ultrasonic range finder and an angle sensor; the laser radar and the ZED binocular camera are pre-3D mapped and saved, the specific position information of obstacles is marked in the map, and a desired path is generated on the saved map; the ZED binocular camera detects the features on both sides of the road based on deep learning; the global satellite positioning GPS and the inertial navigation INS combined navigation contact the electric steering wheel of the tractor, the combined navigation system obtains the position information of the tractor in real time, compares the position information with the position information of the desired path, calculates the deviation, corrects the actual driving direction of the tractor, and installs an angle sensor at the right wheel of the front wheel of the tractor, forms a closed loop control by returning the actual turning angle of the tractor to the CAN bus in real time, and ensures that the tractor travels smoothly according to the desired path; the combined navigation system contacts the path decision system according to the actual position information of the tractor to complete the fixed-point operation and fixed-point parking; the ultrasonic range finder is installed in the area that cannot be scanned by the laser radar, the laser radar is placed on the top of the tractor head, the ultrasonic range finder is placed at the front position of the tractor head, and the safety of the whole automatic driving process is improved.

[0016] Consider the actual orchard environment fruit tree planting is specification planting in row leaving agricultural machinery special straight channel, tractor returns to the warehouse after completing the work, if the ZED binocular camera in step (1) detects that there is a long straight line in front, the path planning decision system contacts the execution control system, and the tractor executes the light accelerator instruction to improve the driving speed of the tractor and save the operation time.

[0017] As a preferred embodiment of the present application, the path planning decision system of step (3) can select the automatic driving or remote driving mode of the tractor, and detect in real time through cloud video transmission. In the case of automatic driving, the tractor can switch to remote driving mode in case of failure or large deviation of automatic driving, thereby improving the safety of tractor operation.

[0018] The present application has the following advantages:

[0019] The present application adopts an information fusion system, a single-thread laser radar and a ZED binocular camera to scan the information around the tractor and build a 3D map. The expensive multi-thread laser radar is abandoned, which greatly reduces the assembly cost of the automatic driving of the tractor. The specific position information of the obstacles is marked in the built map, and an expected trajectory route is generated. The ZED binocular camera and the integrated navigation system are used to detect the straight line of the road and obtain the position information, which assists the tractor to travel smoothly along the expected path and perform spot operation and spot parking. The ultrasonic range finder assists the laser radar in ranging, which accurately ensures that the tractor will not have an unexpected collision accident. The above multi-sensor fusion system can ensure that the tractor works day and night and adapts to different orchard environments. The present application adopts an independent DC stabilized power supply, configures different state power supply prompts, so that each module is normally powered and the service life of the power supply is improved. The guide rail that can slide left and right and the mechanical arm that can be adjusted up and down are used for spraying operation on fruit trees, which ensures the accurate and comprehensive operation of the tractor on fruit trees in actual operation and improves the operation efficiency. In the case of automatic driving, the tractor can switch to remote driving mode through cloud sending and receiving commands in case of failure of automatic driving, thereby improving the safety of tractor operation. BRIEF DESCRIPTION OF DRAWINGS

[0020] Figure 1 The automatic driving tractor system structure diagram of the present application is shown in the figure.

[0021] Figure 2 The path tracking deviation principle diagram of the present application is shown in the figure. DETAILED DESCRIPTION

[0022] The present application will be further illustrated in conjunction with the drawings and specific embodiments. It should be understood that the following specific embodiments are only used to illustrate the present application and are not used to limit the scope of the present application.

[0023] AsFigure 1 The tractor overall system structure schematic diagram is shown, the communication between the modules of the tractor is based on CAN bus, has strong real-time, long transmission distance, strong anti-geomagnetic interference ability, can adapt to the stable work of the tractor in the high-noise and high-dither environment, double-line communication, and reduces the transmission cost.

[0024] The tractor tail is provided with a spraying machine capable of sliding left and right and lifting up and down, which ensures the comprehensiveness of the tractor in actual work and improves the work efficiency. When the tractor is started, the default mode is automatic driving mode, the remote detection room obtains the surrounding environment information and working state of the tractor in real time through the cloud, and the remote driving mode is in an open state, so the mode can be switched at any time.

[0025] As shown in Figure 1 The information acquisition system includes a laser radar, a ZED binocular camera, a global satellite positioning GPS, an inertial navigation INS integrated navigation, an ultrasonic range finder, an angle sensor and the like, and vehicle driving information is collected. The tractor slowly drives around the new orchard environment, the laser radar and the ZED binocular camera scan the tractor body information in real time, 3D mapping and saving, marking the specific position information of the obstacles in the established map, generating an expected path, the ZED binocular vision camera detects the features on both sides of the orchard road based on deep learning, the integrated navigation system obtains the tractor driving information in real time, compares the position information of the expected real-time path, calculates the heading angle deviation generated in the actual driving process of the tractor, controls the electric steering wheel to correct, ensures the tractor to drive according to the expected path, and the angle sensor is installed at the right wheel of the front wheel of the tractor to obtain the actual wheel deflection angle in real time, and forms a closed loop with the steering wheel control system to ensure the accurate driving of the tractor. The integrated navigation system contacts the path decision system to complete the fixed-point work and fixed-point parking according to the actual tractor position information, the ultrasonic range finder is installed in the area that cannot be scanned by the laser radar, the laser radar is placed on the top of the tractor head, and the ultrasonic range finder is placed in front of the tractor head to assist the laser radar in obstacle avoidance and improve the safety of the whole automatic driving process.

[0026] Figure 2 is a principle diagram of the deviation between the actual driving path of the tractor and the expected path, d represents the distance from CG to the nearest point M of the expected path, i.e. the orthogonal projection point of the expected path CG, which is defined as the lateral deviation. represents the actual heading angle and the tangent direction of the expected path The heading error between them can be obtained as: where s is the curve coordinate from the initial position to point M, and ρ(s) represents the curvature of the expected path at point M, and the curve coordinate of point M along the path can be obtained as: where vxand vyrepresent the longitudinal and lateral velocities of the vehicle, respectively. The path tracking error model based on Serret-Frenet equations is given as follows:

[0027] Further, there are: Select The state variable is selected as u(t)=δf, and the tractor path tracking model is as follows:

[0028]

[0029] It should be noted that the above content only illustrates the technical idea of the present application, and cannot limit the protection scope of the present application. For ordinary skilled in the art, without departing from the principles of the present application, a number of improvements and refinements can be made, which fall within the scope of the claims of the present application.

Claims

1. An orchard autonomous tractor, characterized by: The tractor body and the automatic driving tractor system are included, The tail of the tractor body is provided with a left-right sliding guide rail, and a spraying agricultural implement is hung on the guide rail, and a mechanical arm of the spraying agricultural implement is provided with a liftable spray head; The automatic driving tractor system includes an information acquisition system, an information fusion system, a path planning and decision system, an execution control system, a communication system, a remote driving system and a power control system; The information acquisition system includes a single-thread laser radar, a ZED binocular camera, a global satellite positioning GPS and an inertial navigation INS combined navigation, an ultrasonic range finder and an angle sensor, the single-thread laser radar is fixed on the central top of the tractor body, the ultrasonic range finder is installed on the front license plate position of the tractor body, and the angle sensor is installed on the right wheel of the front wheel of the tractor body. The power control system adopts an independent direct current power supply, is physically isolated from the main battery of the tractor, and is integrated with a three-state prompt lamp.

2. The tractor of claim 1, wherein, The three-state prompt lamp includes a normal working state prompt lamp, a power under-voltage state prompt lamp and a charging state prompt lamp.

3. The tractor of claim 1, wherein, The systems are communicated through a CAN bus.

4. The control method of an orchard autonomous tractor according to any one of claims 1 to 3, characterized in that, It includes: (1) collecting various data through the information acquisition system; (2) sending the data to the information fusion system and fusing the sensing data; (3) transmitting the processed data to the path planning and decision system, the path planning and decision system making corresponding decision instructions and adopting automatic driving or remote driving mode; (4) the tractor driving process adopts the guide rail which can slide left and right and the mechanical arm which can be adjusted up and down to spray the fruit trees.

5. The method of claim 4, wherein: In the information acquisition system of step (1), the single-thread laser radar and the ZED binocular camera are saved in a 3D map, and the specific position information of the obstacles is marked in the map, and an expected path is generated; the ZED binocular camera detects the features on both sides of the road based on deep learning; The global satellite positioning GPS and the inertial navigation INS combined navigation contact the electric steering wheel of the tractor, the combined navigation system obtains the driving route information of the tractor vehicle in real time, compares with the expected driving path, calculates the deviation, corrects the actual direction of the tractor, and installs an angle sensor on the right wheel of the front wheel of the tractor, returns the actual turning angle of the tractor through the CAN bus to form a closed loop control, and ensures that the tractor can smoothly drive according to the expected path; the combined navigation system contacts the path decision system to complete the fixed-point operation and fixed-point parking according to the actual tractor position information; the ultrasonic range finder assists the single-thread laser radar to avoid obstacles, and improves the safety of the whole automatic driving process.

6. The method of claim 4, wherein: In step (3), the tractor adopts automatic driving or remote driving mode in the path planning and decision system, and detects in real time through cloud video transmission, under the premise of giving priority to automatic driving, when the automatic driving fails or the deviation is too large to be corrected automatically in an emergency, the tractor can be switched to remote driving mode, thereby improving the safety of the whole operation of the tractor.

Images