Device and method for analyzing slope-ground-trafficability

Abstract

The invention discloses a device and a method for analyzing slope-ground-trafficability. The device comprises a sensor module, a data processing and transmitting module and a radar, the data processing and transmission module comprises a wireless sensing unit sending end, a wireless sensing unit receiving end and an upper computer, and the sensor module is arranged on the wheel step type working platform; a wireless sensing unit sending end and a radar are further arranged on the wheel step type working platform; the wireless sensing unit transmitting end is in wireless communication connection with the wireless sensing unit receiving end, and the wireless sensing unit receiving end is connected with the upper computer; according to the method, multiple groups of sensors and radars are integrated through the wheel step type working platform, and the 3D point cloud modeling and the lightweight semantic segmentation network are combined, so that on-site rapid quantitative detection of a slope-ground-trafficability coupling relationship of the complex terrain is realized, and the limitation of a traditional horizontal ground methodology and the pain points of subjective and large deviation of existing evaluation are broken.

Description

Technical Field

[0001] This invention relates to the field of ground-machine system mechanics, specifically to a device and method for analyzing slope-ground-passability. Background Technology

[0002] Traditional vehicle, agricultural machinery, and engineering equipment designs are based on a "level ground" methodology, and their passability evaluation systems are only applicable to flat, firm, conventional terrain. When facing complex terrains such as hilly areas, slopes, and soft ground, existing technologies lack dedicated devices capable of rapidly and quantitatively detecting the coupling relationship between "slope, soil bearing capacity, and slip limit." Current passability evaluations mainly rely on operator experience during trial runs or simulated testing using indoor soil trench experiments. Experience-based trial runs are heavily influenced by human factors, resulting in subjective and inaccurate evaluations. Furthermore, they cannot predict the equipment's passability in complex terrain, easily leading to safety accidents and operational losses such as equipment failure to navigate obstacles, rollovers, and low traction efficiency. Indoor soil trench experiments cannot realistically reproduce the complex terrain conditions of hilly areas and mountain slopes, resulting in significant discrepancies between experimental results and actual operating scenarios, failing to provide reliable technical support for equipment design and on-site operations.

[0003] Currently, the promotion of agricultural mechanization in hilly and mountainous areas and the construction of ultra-high voltage towers in mountainous areas have created an urgent need for assessment of the passability of complex terrain. There is an urgent need for a technical solution that can quantitatively detect the coupling relationship between "slope-ground-passability" in one go on the spot, so as to solve the technical pain points of existing evaluation methods, such as strong subjectivity, low accuracy and poor adaptability to different scenarios, and provide a scientific basis for equipment design optimization and on-site operation safety assurance. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a device and method for analyzing slope-ground-passability, aiming to solve the problems in the background technology.

[0005] To achieve the above objectives, the present invention provides the following technical solution: a device for analyzing slope-ground-passability, comprising a sensor module, a data processing and transmission module, and a radar; the data processing and transmission module includes a wireless sensing unit transmitter, a wireless sensing unit receiver, and a host computer; the sensor module is mounted on a wheeled walking platform; the wheeled walking platform is also equipped with a wireless sensing unit transmitter and a radar; the wireless sensing unit transmitter and the wireless sensing unit receiver are wirelessly connected, and the wireless sensing unit receiver is connected to the host computer; The host computer executes: The system receives radar data to create 3D point cloud models, and combines these with a trained lightweight semantic segmentation network to identify the slope of the corresponding area. Simultaneously, it analyzes the attitude data of the wheeled walking platform collected by the sensor module to determine the real-time attitude of the platform. Finally, by fusing and analyzing the slope information output by the trained lightweight semantic segmentation network with the real-time attitude of the wheeled walking platform, the system completes the slope-ground-passability analysis.

[0006] Furthermore, the sensor module includes an inertial measurement unit, a tilt sensor, and a multi-dimensional torque sensor; the inertial measurement unit, tilt sensor, and multi-dimensional torque sensor are mounted on the leg axle of the wheel-walking work platform, and the outer end of the leg axle is connected to the wheel section.

[0007] Furthermore, the inertial measurement unit, tilt sensor, and multi-dimensional torque sensor are wirelessly connected to the host computer.

[0008] Furthermore, it also includes a power module, which provides power to the inertial measurement unit, tilt sensor, multi-dimensional torque sensor, wireless sensing unit transmitter, wireless sensing unit receiver, and host computer through a voltage regulator chip LDO.

[0009] Furthermore, four sets of inertial measurement units and tilt sensors are respectively installed on the four leg axles of the wheel-walking work platform.

[0010] A method for analyzing slope-ground-accessibility includes the following steps: Step S1: First, start the wheel-walking work platform, reset the leg axle and wheel to the initial position, and the host computer obtains the initial drift value of the sensor module and performs zero-point calibration. Step S2: Turn on the power module to initialize the sensor module and radar, so that the sensor module and radar are ready to work. Step S3: After the data from the sensor module and radar stabilizes, the surrounding environment is identified by the radar, and the height of the wheel-type work platform is adjusted by controlling the leg axle, thereby sending the radar data collected in the entire three-dimensional plane to the host computer. Step S4: The host computer communicates with the inertial measurement unit and tilt sensor on the leg axis and parses the attitude data; the attitude data is assembled into a data frame; the attitude data includes pitch angle, yaw angle, and roll angle; Step S5: Perform 3D point cloud modeling on the host computer using the radar data uploaded in Step S3; combine the trained lightweight semantic segmentation network to identify the slope of the area corresponding to the model; simultaneously call the attitude data of the four leg axes obtained in Step S4 and the data collected by the multi-dimensional torque sensor to determine the real-time attitude of the wheeled walking platform; finally, by fusing and analyzing the slope information output by the trained lightweight semantic segmentation network with the real-time attitude of the wheeled walking platform, complete the slope-ground-passability analysis.

[0011] Furthermore, the attitude data is parsed and represented as: ; ; ; In the formula, , , These represent pitch angle, roll angle, and yaw angle, respectively. , , , The four components of the quaternion are calculated from data collected by the inertial measurement unit and the tilt sensor.

[0012] Furthermore, the point cloud modeling process can be represented as follows: ; ; ; ; In the formula, This refers to the raw point cloud data, i.e., the radar data uploaded in step S3; This describes a statistical outlier removal algorithm. This represents the clean point cloud data after denoising. The transformation matrix representing point cloud registration; , This represents the points in the two sets of point cloud data to be registered; Represents the optimal registration transformation matrix; This represents the Poisson surface reconstruction algorithm; The parameters represent the Poisson reconstruction. The initial surface mesh model is the mesh structure model corresponding to the point cloud obtained through Poisson reconstruction. This represents a simplified algorithm for quadratic edge folding. Parameters indicating mesh simplification; This represents a multi-level detail mesh model, which is the final mesh model after lightweight simplification.

[0013] An electronic device includes a processor, a memory, and a bus, wherein the processor and the memory are connected via the bus, the memory is used to store a set of program code, and the processor is used to call the program code stored in the memory to execute a method for analyzing slope-ground-accessibility.

[0014] A non-volatile computer storage medium storing computer-executable instructions that perform a method for analyzing slope-ground-accessibility.

[0015] Compared with existing technologies, the present invention has the following advantages: (1) This invention integrates multiple sensors and radars through a wheeled walking platform, combined with 3D point cloud modeling and a lightweight semantic segmentation network, to achieve rapid quantitative detection of the coupling relationship between slope, ground, and passability in complex terrains, overcoming the limitations of traditional level ground methodologies and the pain points of subjective and biased evaluations. The use of wireless sensor transmission eliminates the inconvenience of wires, and the modular layout and compact design make the device portable and easy to deploy. Multi-sensor collaborative compensation for data drift and accurate data analysis provide reliable technical support for equipment design optimization and operational safety in scenarios such as agricultural machinery operations in hilly and mountainous areas and UHV tower erection in mountainous terrain.

[0016] (2) This invention effectively solves the problem that the traditional passability evaluation system based on horizontal ground fails in complex terrains such as hilly and mountainous slopes and soft ground, and fills the technical gap of lacking on-site, rapid and quantitative detection of the coupling relationship between slope-soil bearing capacity-slip limit; adopts radar 3D point cloud modeling and multi-sensor (inertial measurement unit, tilt sensor, six-dimensional torque sensor) collaborative perception design, combined with lightweight semantic segmentation network and platform real-time attitude data fusion analysis, significantly improving the accuracy and scientific nature of passability evaluation.

[0017] (3) This invention uses wireless sensing technology to realize data transmission, eliminating the cumbersome wire connection. Combined with the compact structure and modular layout of the wheeled work platform, the device is small in size and highly portable, making it easy to deploy and maintain on site. It can complete the slope-ground-passability coupling relationship detection on site in one go, without relying on subjective experience for test runs or indoor soil trench experiments with large deviations. The detection results are consistent with the actual operation scenario, providing reliable technical support for the design optimization of vehicles, agricultural machinery, UHV mountain tower equipment, etc., and the on-site operation safety guarantee. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of the device structure of the present invention.

[0019] In the diagram, 1. Wheel section; 2. Leg axle; 3. Inertial measurement unit; 4. Wheel-walking work platform; 5. Wireless sensor unit transmitter; 6. Radar; 7. Wireless sensor unit receiver; 8. Tilt sensor; 9. Multi-dimensional torque sensor; 10. Host computer. Detailed Implementation

[0020] like Figure 1 As shown, the present invention provides a technical solution: a device for analyzing slope-ground-passability, comprising: a sensor module, a data processing and transmission module, and a radar 6; the data processing and transmission module includes a wireless sensing unit transmitter 5, a wireless sensing unit receiver 7, and a host computer 10; The sensor module is installed on the wheeled walking platform 4; the wheeled walking platform 4 is also equipped with a wireless sensor unit transmitter 5 and a radar 6; the wireless sensor unit transmitter 5 and the wireless sensor unit receiver 7 are wirelessly connected, and the wireless sensor unit receiver 7 is connected to the host computer 10. The host computer executes: The system receives radar data to create 3D point cloud models, and combines these with a trained lightweight semantic segmentation network to identify the slope of the corresponding area. Simultaneously, it analyzes the attitude data of the wheeled walking platform collected by the sensor module to determine the real-time attitude of the platform. Finally, by fusing and analyzing the slope information output by the trained lightweight semantic segmentation network with the real-time attitude of the wheeled walking platform, the system completes the slope-ground-passability analysis.

[0021] The sensor module includes an inertial measurement unit 3, a tilt sensor 8, and a multi-dimensional torque sensor 9. The inertial measurement unit 3, the tilt sensor 8, and the multi-dimensional torque sensor 9 are mounted on the leg axle 2 of the wheel-walking work platform 4, and the outer end of the leg axle 2 is connected to the wheel section 1.

[0022] Among them, the inertial measurement unit 3, the tilt sensor 8, and the multi-dimensional torque sensor 9 are wirelessly connected to the host computer 10.

[0023] It also includes a power supply module, which supplies power to the inertial measurement unit 3, tilt sensor 8, multi-dimensional torque sensor 9, wireless sensing unit transmitter 5, wireless sensing unit receiver 7 and host computer 10 through a voltage regulator chip LDO.

[0024] Among them, Radar 6 uses the RPLIDAR C1 model, which is used to identify the outlines of surrounding objects and for point cloud modeling.

[0025] The inertial measurement unit 3 and tilt sensor 8 are in four groups, respectively set on the four leg shafts 2 of the wheel-walking work platform 4; each group has four, arranged in a relative manner to compensate for sensor data drift.

[0026] Among them, the multidimensional torque sensor 9 is a six-dimensional force sensor, including three-dimensional force and three-dimensional torque.

[0027] The wireless sensing unit transmitter 5 and the power module are located inside the wheeled walking platform 4; the radar 6 is installed above the wheeled walking platform 4 and can better identify the surrounding environment.

[0028] A method for analyzing slope-ground-accessibility includes the following steps: Step S1: First, start the wheel-type work platform 4, reset the leg axle 2 and wheel part 1 to the initial position, and the host computer 10 obtains the initial drift value of the sensor module and performs zero-point calibration.

[0029] Step S2: Turn on the power module to initialize the sensor module and radar 6, so that the sensor module and radar 6 are in a working state.

[0030] Step S3: After the data from the sensor module and radar 6 stabilizes, the surrounding environment is identified by radar 6. At the same time, the leg shaft 2 is controlled to adjust the height of the wheeled work platform 4, and then the radar data collected in the entire three-dimensional plane is sent to the host computer 10.

[0031] Step S4: The host computer 10 communicates with the inertial measurement unit 3 and tilt sensor 8 on the leg axis 2 and parses the attitude data (pitch angle, yaw angle, roll angle); the attitude data is then combined into a data frame.

[0032] The parsed attitude data (pitch angle, yaw angle, roll angle) are represented as follows: ; ; ; In the formula, , , These represent pitch angle, roll angle, and yaw angle, respectively. , , , The four components of a quaternion are represented. Quaternions are used to describe the attitude of a rigid body and are calculated from the data collected by the inertial measurement unit 3 and the tilt sensor 8.

[0033] Step S5: In the host computer 10, 3D point cloud modeling is performed using the radar data uploaded in step S3; combined with the trained lightweight semantic segmentation network, the slope of the area corresponding to the model is identified; at the same time, the attitude data of the four leg axes 2 obtained in step S4 and the data collected by the multi-dimensional torque sensor 9 are called to determine the real-time attitude of the wheeled work platform 4; finally, by fusing and analyzing the slope information output by the trained lightweight semantic segmentation network and the real-time attitude of the wheeled work platform 4, the slope-ground-passability analysis is completed.

[0034] The point cloud modeling process is represented as follows: ; ; ; ; In the formula, This refers to the raw point cloud data, i.e., the radar data uploaded in step S3; This represents a statistical outlier removal algorithm, a noise reduction method for point cloud preprocessing, used to remove abnormal noise points from the original data; This represents the clean point cloud data after denoising. The transformation matrix representing point cloud registration; , This represents the points in the two sets of point cloud data to be registered; This represents the optimal registration transformation matrix, which is the transformation matrix that minimizes the alignment error between the two sets of point clouds. This represents the Poisson surface reconstruction algorithm; The parameters represent the Poisson reconstruction. The initial surface mesh model is the mesh structure model corresponding to the point cloud obtained through Poisson reconstruction. This represents a simplified algorithm for quadratic edge folding. Parameters indicating mesh simplification; This represents a multi-level of detail (LOD) mesh model, which is the final mesh model after being lightweighted and simplified (to facilitate subsequent calculations and analysis).

[0035] The preparation process of the device of the present invention is as follows: 1. Measure the size of the inertial measurement unit 3 and the tilt sensor 8, and make a suitable mold by 3D printing. Make a silicone shell by mixing 1:1 AB silicone and embedding the inertial measurement unit 3 and the tilt sensor 8 into it. Finally, fix the whole part to the surface of the leg shaft 2 with acrylic glue.

[0036] 2. Fix three copper pillars on the wheeled work platform 4 with hot melt glue, pass the screw parts of the copper pillars through the screw holes of the radar 6, and fix the radar 6 with bolts.

[0037] 3. The sensor module and data processing and transmission module circuit boards are fabricated using flexible circuit board fabrication technology. The circuit boards are fixed in the center inside the wheeled walking platform 4 with acrylic glue.

[0038] An electronic device includes a processor, a memory, and a bus, wherein the processor and the memory are connected via the bus, the memory is used to store a set of program code, and the processor is used to call the program code stored in the memory to execute a method for analyzing slope-ground-accessibility.

[0039] A non-volatile computer storage medium storing computer-executable instructions that perform a method for analyzing slope-ground-accessibility.

[0040] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. An apparatus for analyzing slope-ground-passability, comprising a sensor module, a data processing and transmission module, a radar; the data processing and transmission module comprises a wireless sensor unit sending end, a wireless sensor unit receiving end and an upper computer, characterized in that: The sensor module is arranged on the wheel-step operation platform; The wheel-step operation platform is further provided with a wireless sensor unit sending end and a radar; the wireless sensor unit sending end and the wireless sensor unit receiving end are wirelessly connected, and the wireless sensor unit receiving end is connected with the upper computer; The upper computer executes: Receiving the data collected by the radar to perform 3D point cloud modeling, combining the trained lightweight semantic segmentation network, identifying the slope of the region corresponding to the model; at the same time, the posture data of the wheel-step operation platform collected by the sensor module is analyzed, and the real-time posture of the current wheel-step operation platform is judged; finally, the slope information output by the trained lightweight semantic segmentation network and the real-time posture of the wheel-step operation platform are fused and analyzed to complete the slope-ground-passability analysis.

2. A device for analyzing slope-ground-passability according to claim 1, characterized in that: The sensor module includes an inertial measurement unit, an inclination sensor and a multi-dimensional torque sensor; the inertial measurement unit, the inclination sensor and the multi-dimensional torque sensor are arranged on the leg shaft of the wheel-step operation platform, and the outer side end of the leg shaft is connected with the wheel part.

3. A device for analyzing slope-ground-passability according to claim 2, characterized in that: The inertial measurement unit, the inclination sensor and the multi-dimensional torque sensor are wirelessly connected with the upper computer.

4. A device for analyzing slope-ground-passability according to claim 3, characterized in that: Further comprising a power module, the power module supplies power to the inertial measurement unit, the inclination sensor, the multi-dimensional torque sensor, the wireless sensor unit sending end, the wireless sensor unit receiving end and the upper computer through the voltage stabilizing chip LDO.

5. A device for analyzing slope-ground-passability according to claim 4, characterized in that: There are four groups of inertial measurement units and inclination sensors, which are arranged on the four leg shafts of the wheel-step operation platform.

6. A method for analyzing slope-ground-passability, realized on the basis of a device for analyzing slope-ground-passability according to any one of claims 1-5, characterized in that, The steps include: Step S1: first start the wheel-step operation platform, reset the leg shaft and the wheel part to the initial position, and the upper computer obtains the initial drift value of the sensor module and performs zero correction; Step S2: turn on the power module, initialize the sensor module and the radar, and make the sensor module and the radar enter the working state; Step S3: after the data of the sensor module and the radar is stable, the surrounding environment is identified by the radar, the height of the wheel-step operation platform is adjusted by controlling the leg shaft, and then the radar data collected by the entire three-dimensional plane is sent to the upper computer; Step S4: the upper computer communicates with the inertial measurement unit and the inclination sensor on the leg shaft, and analyzes the posture data; the posture data is composed of a data frame; the posture data includes pitch angle, yaw angle and roll angle; Step S5: in the upper computer, 3D point cloud modeling is performed on the radar data uploaded in step S3; combining the trained lightweight semantic segmentation network, the slope of the region corresponding to the model is identified; at the same time, the posture data of the four leg shafts obtained in step S4 and the data collected by the multi-dimensional torque sensor are called to judge the real-time posture of the current wheel-step operation platform; finally, the slope information output by the trained lightweight semantic segmentation network and the real-time posture of the wheel-step operation platform are fused and analyzed to complete the slope-ground-passability analysis.

7. A method of analyzing grade-ground-passability according to claim 6, characterized in that: The posture data is analyzed and expressed as: ； ； ； In the formula, , , respectively represent the pitch angle, roll angle, yaw angle; , , , represent the 4 components of the quaternion, which are calculated from the data collected by the inertial measurement unit and the tilt sensor.

8. A method of analyzing grade-ground-passability according to claim 6, characterized in that: The point cloud modeling process is expressed as: ； ； ； ； In the formula, represents the original point cloud data, i.e. the radar data uploaded in step S3; represents a statistical outlier removal algorithm; represents the clean point cloud data after denoising; represents a transformation matrix of point cloud registration; , represents a point in the two sets of point cloud data to be registered; represents an optimal registration transformation matrix; represents a Poisson surface reconstruction algorithm; represents a parameter of Poisson reconstruction; represents an initial surface mesh model, which is a mesh structure model corresponding to the point cloud obtained by Poisson reconstruction; represents a secondary edge collapse simplification algorithm; represents a parameter of mesh simplification; represents a multi-detail level mesh model, which is a final mesh model after lightweight simplification.

9. An electronic device, comprising: The device includes a processor, a memory, and a bus, wherein the processor and the memory are connected via the bus, the memory is used to store a set of program code, and the processor is used to call the program code stored in the memory to execute a method for analyzing slope-ground-accessibility as described in any one of claims 6-8.

10. A non-transitory computer storage medium having stored computer- executable instructions, the computer-executable instructions comprising instructions for: The computer can execute instructions to perform a method for analyzing slope-ground-accessibility as described in any one of claims 6-8.

Images