An adaptive intelligent Panax notoginseng harvester chassis

Abstract

This invention provides an adaptive intelligent Panax notoginseng harvester chassis, belonging to the field of agricultural machinery equipment technology. The invention comprises two parts: a mechanical structure and a control system. The mechanical structure includes a radiator, motor, oil filler, tracked engine, commutator, guide rail, tracks, digital hydraulic cylinder, solenoid valve, hydraulic gauge, oil pump, battery, and oil tank. The track spacing is adjusted through a digital hydraulic system. The digital hydraulic cylinder integrates a stepper or servo motor, hydraulic valve, displacement sensor, and closed-loop feedback design within the cylinder, connected to a hydraulic oil source. All functions are directly controlled by an internal digital hydraulic cylinder controller. The digital hydraulic system includes components such as a motor, hydraulic gauge, oil pump, oil tank, and solenoid valve. Each track is connected to a tracked engine, with a guide rail below the engine. When the track spacing changes, the engine slides on the guide rail.

Description

Technical Field

[0001] This invention belongs to the field of agricultural machinery and equipment technology, and in particular relates to an adaptive intelligent Panax notoginseng harvester chassis. Technical Background

[0002] Panax notoginseng, a plant belonging to the genus Panax, possesses the effects of dispersing blood stasis, stopping bleeding, reducing swelling, and relieving pain. It is a precious traditional Chinese medicine, growing in high-altitude mountainous and hilly areas, mainly distributed in Yunnan Province and other regions of my country. Due to the terroir-specific nature of Panax notoginseng, it is not cultivated abroad, and there is no equipment available for design reference. Research on domestically developed Panax notoginseng combine harvesters in my country has mainly focused on the experimental and trial production stages. Moreover, existing prototype harvesters are only suitable for specific harvesting scenarios, lacking versatility and failing to truly liberate manual labor.

[0003] With the rapid development of computer vision, control theory, artificial intelligence, sensor technology, and other disciplines, new technologies are being applied more and more widely in the agricultural field. Intelligent agricultural equipment must first solve three problems in its working environment: positioning, identification, and execution.

[0004] In recent years, many companies have invested in the research of agricultural machinery, which has led to the widespread adoption of agricultural machinery in people's production and daily lives. However, the agronomic standards for Panax notoginseng cultivation have not yet been widely adopted, and the existing harvester chassis spacing is fixed, making it unable to adapt to planting ridges with different spacing. This often results in harvester tipping over and the chassis damaging the Panax notoginseng, affecting harvesting efficiency and causing economic losses. Summary of the Invention

[0005] The purpose of this invention is to provide a chassis for a Panax notoginseng harvester in hilly greenhouses based on a vision and digital hydraulic system, which overcomes the problem that the existing technology cannot meet the requirement that the chassis of the Panax notoginseng harvester cannot adjust the track spacing according to the furrow spacing. At the same time, the chassis can achieve accurate positioning and stable operation of the navigation system when the environment inside and outside the greenhouse changes.

[0006] To achieve the above objectives, the technical solution adopted by this invention is as follows: an adaptive intelligent Panax notoginseng harvester chassis, comprising a mechanical structure and a control system. The mechanical structure includes a tracked engine, a commutator, a guide rail, two tracks, and a digital hydraulic cylinder. Each track is connected to a tracked engine, which is mounted on the guide rail. When the track spacing changes, the engine slides on the guide rail. The tracked engine is located at the front of the tracked chassis and is connected to the commutator, which is located on the guide rail. The commutator output shaft is connected to the track drive wheel to drive the track rotation. An oil tank is included, with a radiator on top. A motor is connected to an oil pump, which is connected to a solenoid valve via an oil pipe. The solenoid valve is connected to the digital hydraulic cylinder via an oil pipe. The digital hydraulic cylinder is controlled by a digital hydraulic cylinder controller and is connected to the tracked engine. A filler port is provided on the oil tank. A solenoid valve, a hydraulic gauge, and a hydraulic gauge mounted on the solenoid valve are also included. A battery is connected to the motor to provide power.

[0007] The control system comprises a four-layer structure. The first layer is the environmental perception layer, which includes BeiDou navigation, a two-dimensional single-line LiDAR, and a depth camera to identify the environment around the Panax notoginseng harvester chassis, establish a cloud map, detect target positions, and perform localization. The second layer is the control layer, which uses an NVIDIA Jetson AGX Orin module as the host of the Panax notoginseng harvester chassis to process the acquired images and generate control signals. The third layer is the drive layer, which receives signals from the host and sends them to the execution layer. The fourth layer is the execution layer, which executes the signals sent by the host and adjusts the track spacing.

[0008] The depth camera is a camera installed in front of the Panax notoginseng harvester chassis, and the two-dimensional single-line lidar is installed on the top of the Panax notoginseng harvester chassis.

[0009] In the event of a system error, it can be remotely controlled to take necessary emergency measures.

[0010] The beneficial effects of this invention are:

[0011] (1) The adaptive intelligent Panax notoginseng harvester chassis described in this invention is suitable for large-scale Panax notoginseng harvesting operations;

[0012] (2) It is suitable for harvesting Panax notoginseng in hilly and mountainous areas with complex terrain;

[0013] (3) To achieve mechanized production of Panax notoginseng harvesting, improve efficiency, and save human resources;

[0014] (4) It has a monitoring, identification and control system, which is stable in operation, improves the harvesting efficiency of Panax notoginseng, can adapt to different spacing of furrow operation, and can adjust the track spacing in a timely manner according to different furrow spacing, thereby reducing the situation of track pressing on the ridge, avoiding crushing of Panax notoginseng and side roll, improving the harvest rate, reducing property loss, and has a very broad market prospect. Attached Figure Description

[0015] Figure 1 This is a top view of the overall structure of the present invention;

[0016] Figure 2 This is a three-dimensional schematic diagram of the overall structure of the present invention;

[0017] Figure 3 This is a three-dimensional bottom view of the overall structure of the present invention;

[0018] Figure 4 This is a schematic diagram of the hydraulic system of the present invention;

[0019] Figure 5 It is a system control flowchart;

[0020] Attached reference numerals: 1-Radiator, 2-Motor, 3-NVIDIA Jetson AGX Orin module, 4-Fuel filler port, 5-Crawler engine, 6-Commutator, 7-Guide rail, 8-Crawler, 9-Digital hydraulic cylinder, 10-Solenoid valve, 11-Hydraulic gauge, 12-Oil pump, 13-Battery, 14-Oil tank, 15-BeiDou navigation, 16-2D single-line LiDAR, 17-Depth camera, 18-Ubuntu system, 19-Robot Operating System. Detailed Implementation

[0021] The technical solution of the present invention will now be clearly and completely described in conjunction with the accompanying drawings and embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments are within the scope of protection of the present invention.

[0022] like Figure 1 As shown, this invention provides an adaptive intelligent Panax notoginseng harvester chassis, comprising a mechanical structure and a control system. The mechanical structure includes a tracked engine 5, a commutator 6, a guide rail 7, two tracks 8, and a digital hydraulic cylinder 9. Each track 8 is connected to a tracked engine 5, which is mounted on the guide rail 7. When the track spacing changes, the engine slides on the guide rail 7. The tracked engine 5 is located at the front of the tracked chassis and is connected to the commutator 6, which is located on the guide rail. The output shaft of the commutator 6 is connected to... Track 8 is connected to the drive wheel, which drives track 8 to rotate. There is an oil tank 14, and a radiator 1 is placed on the top of the oil tank 14. The motor 2 is connected to the oil pump 12. The oil pump 12 is connected to the solenoid valve 10 through an oil pipe. The solenoid valve 10 is connected to the digital hydraulic cylinder 9 through an oil pipe. The digital hydraulic cylinder is controlled by a digital hydraulic cylinder controller. The digital hydraulic cylinder 9 is connected to the track engine 5. The oil tank 14 is equipped with a filler port 4. The solenoid valve 10 and the hydraulic gauge 11 are installed on the solenoid valve 10. The battery 13 is connected to the motor 2 to provide power to the motor 2.

[0023] The control system comprises a four-layer structure. The first layer is the environmental perception layer, which includes BeiDou navigation, a two-dimensional single-line LiDAR, and a depth camera to identify the environment around the Panax notoginseng harvester chassis, establish a cloud map, detect target positions, and perform localization. The second layer is the control layer, which uses an NVIDIA Jetson AGX Orin module as the host for the Panax notoginseng harvester chassis to process the acquired images and generate control signals. The third layer is the drive layer, which receives signals from the host and sends them to the execution layer. The fourth layer is the execution layer, which executes the signals sent by the host and adjusts the track spacing.

[0024] An operating method for a chassis of an intelligent Panax notoginseng harvester includes the following steps:

[0025] (1) System initialization: enable Beidou navigation, 2D single-line lidar, and depth camera;

[0026] (2) A depth camera and a two-dimensional single-line lidar are used to monitor the environment of Panax notoginseng planting fields, obtain environmental data, and establish a cloud map;

[0027] (3) The NVIDIA Jetson AGX Orin module performs recognition processing on the Panax notoginseng planting field image returned by the depth camera to determine the furrow location and calculate the distance between adjacent furrows. Specifically, the yolov7 is used to process the Panax notoginseng planting field image acquired by the depth camera, the furrows are identified by the improved segmented centroid B-spline curve fitting algorithm, and the GAM embedding global attention mechanism is used for calibration to obtain the average crossover ratio and pixel accuracy, the distance between two adjacent furrows is determined and returned to the NVIDIA Jetson AGX Orin module.

[0028] The formula is as follows:

[0029]

[0030] In the formula, For the average crossover ratio, For pixel accuracy, p represents the average pixel accuracy. ij p represents the number of pixels where the i-th type of target is predicted to be the j-th type. ii This represents the number of correctly predicted pixels, where k represents the number of categories other than background, and p... i,3 (u) is the i-th segment of the 4th order 3rd degree B-spline curve. For spline curve parameters, The curve controls the vertex. 3 represents the basis functions of the second-order B-spline curve. Let i be the endpoint position and first derivative of the i-th cubic B-spline curve. Let represent the endpoint position and second derivative of the i-th cubic B-spline curve.

[0031] (4) The NVIDIA Jetson AGX Orin module generates control signals and sends them to the digital hydraulic cylinder controller;

[0032] (5) The digital hydraulic cylinder controller receives the control signal to drive the digital hydraulic cylinder 9 to achieve precise control of the track spacing 8.

[0033] In step (3), YOLOv7 is a target detection model, which includes OpenCV, PyTorch, YOLO, etc.

[0034] If a system crash or identification error occurs during this period, resulting in abnormal adjustment of track spacing 8, it will be handled remotely by a human to improve system robustness.

Claims

1. An operating method for an adaptive intelligent Panax notoginseng harvester chassis, comprising the following steps: (1) System initialization: enable Beidou navigation, 2D single-line lidar, and depth camera; (2) A depth camera and a two-dimensional single-line lidar are used to monitor the environment of Panax notoginseng planting fields, obtain environmental data, and establish a cloud map; (3) The NVIDIA Jetson AGX Orin module performs recognition processing on the Panax notoginseng planting field image returned by the depth camera to determine the furrow location and calculate the distance between adjacent furrows. Specifically, the yolov7 is used to process the Panax notoginseng planting field image acquired by the depth camera, the furrows are identified by the improved segmented centroid B-spline curve fitting algorithm, and the GAM embedding global attention mechanism is used for calibration to obtain the average crossover ratio and pixel accuracy, the distance between two adjacent furrows is determined and returned to the NVIDIA Jetson AGX Orin module. The formula is as follows: In the formula, For the average crossover ratio, For pixel accuracy, p represents the average pixel accuracy. ij p represents the number of pixels where the i-th type of target is predicted to be the j-th type. ii This represents the number of correctly predicted pixels, where k represents the number of categories other than background, and p... i,3 (u) is the i-th segment of the 4th order 3rd degree B-spline curve. For spline curve parameters, The curve controls the vertex. The basis functions are for a cubic B-spline curve. Let i be the endpoint position and first derivative of the i-th cubic B-spline curve. Let be the endpoint position and second derivative of the i-th cubic B-spline curve; (4) The NVIDIA Jetson AGX Orin module generates control signals and sends them to the digital hydraulic cylinder controller; (5) The digital hydraulic cylinder controller receives control signals to drive the digital hydraulic cylinder (9) to achieve precise control of the track (8) spacing; The adaptive intelligent Panax notoginseng harvester chassis comprises two parts: a mechanical structure and a control system. The mechanical structure includes a tracked engine (5), a commutator (6), a guide rail (7), two tracks (8), and a digital hydraulic cylinder (9). Each track (8) is connected to a tracked engine (5), which is mounted on the guide rail (7). When the track (8) spacing changes, the engine slides on the guide rail (7). The tracked engine (5) is located in the front of the tracked chassis and is connected to the commutator (6), which is located on the guide rail. The output shaft of the commutator (6) is connected to the drive wheel of the track (8). The track (8) is driven to rotate. The oil tank (14) has a radiator (1) on top. The motor (2) is connected to the oil pump (12). The oil pump (12) is connected to the solenoid valve (10) through the oil pipe. The solenoid valve (10) is connected to the digital hydraulic cylinder (9) through the oil pipe. The digital hydraulic cylinder is controlled by the digital hydraulic cylinder controller. The digital hydraulic cylinder (9) is connected to the track engine (5). The oil tank (14) is equipped with a filler port (4). The solenoid valve (10) and the hydraulic gauge (11) are installed on the solenoid valve (10). The battery (13) is connected to the motor (2) to provide power to the motor (2). The control system comprises a four-layer structure. The first layer is the environmental perception layer, which includes BeiDou navigation, a two-dimensional single-line LiDAR, and a depth camera, used to identify the environment around the Panax notoginseng harvester chassis, establish a cloud map, detect target positions, and perform localization. The second layer is the control layer, which uses an NVIDIA Jetson AGX Orin module as the host for the Panax notoginseng harvester chassis, responsible for processing the acquired images and generating control signals. The third layer is the drive layer, which receives host signals and sends them to the execution layer. The fourth layer is the execution layer, which executes the signals sent by the host and adjusts the track (8) spacing.

2. The operating method of the adaptive intelligent Panax notoginseng harvester chassis according to claim 1, characterized in that, In step (3), YOLOv7 includes OpenCV, PyTorch, or YOLO.

Images