Three-dimensional topographic survey vehicle based on variable-diameter spoked wheels and control system of three-dimensional topographic survey vehicle
By adopting variable diameter spoke wheels and multi-sensor fusion system on the three-dimensional terrain survey vehicle, the existing survey vehicle has limited ability and low perceptual accuracy in complex terrain and high obstacle environments, achieving higher adaptability and survey efficiency.
Patent Information
- Application Number
- CN202510210193.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-25
- Publication Date
- 2025-05-30
AI Technical Summary
Existing surveying vehicles have limited ability to pass in complex terrain and high obstacle environments, and the sensor has low perceptual accuracy in variable environments, making it difficult to effectively survey.
A three-dimensional terrain surveying vehicle based on variable diameter spoke wheels is adopted. The wheel legs are telescopic through a variable diameter motor, and the wheel diameter is dynamically adjusted. It is equipped with a multi-sensor fusion system equipped with a lidar and depth camera to generate a three-dimensional map in real time and optimize the driving path.
It improves the ability and adaptability of surveying vehicles in complex terrain and high obstacle environments, enhances environmental perception accuracy and stability, and improves survey efficiency and application potential.
Smart Images
Figure CN120057144A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of adaptive all-terrain vehicles, and in particular relates to a three-dimensional terrain survey vehicle based on variable-diameter spoke wheels and a control system thereof. Background Art
[0002] Currently, there are many solutions based on mobile robots or unmanned survey vehicles for automatic surveying in unknown environments. These solutions are widely used in planetary exploration, underground mining detection and other hard-to-reach areas. However, existing technologies still have many limitations in environmental adaptability, survey accuracy and terrain crossing capabilities.
[0003] Deficiencies of existing technology:
[0004] Survey vehicles with fixed wheel diameters: Many existing survey vehicles use fixed wheel diameters. Although they can cope with flat terrain to a certain extent, their ability to pass through complex terrain (such as areas with many rocks and obstacles) is greatly limited. Fixed wheel diameters cannot be adjusted flexibly, which makes it easy for the vehicle to be blocked when encountering high obstacles, thus affecting survey efficiency.
[0005] Wheel-foot hybrid robot: Although the wheel-foot hybrid robot has strong passing ability, its structure is complex, the manufacturing cost is high, and it is difficult to maintain. In addition, during long-term operation, the wheel-foot switching mechanism may fail, affecting the overall performance.
[0006] Limitations of terrain perception: Many existing survey systems rely on a single sensor, such as lidar or depth camera, to perceive the terrain. Although these sensors can provide accurate distance and depth information under certain conditions, the sensor performance will be significantly reduced in strong or weak light environments and scenes with a lot of occlusion, resulting in inaccurate terrain perception. A single sensor is often unable to cope with changing ambient lighting or terrain complexity. Even with multi-sensor fusion solutions, there are still perception blind spots when dealing with harsh environments such as strong light and dust obstruction. Summary of the invention
[0007] The present invention aims to solve the problem that existing robots have limited adaptability to different terrains, especially when they need to cross larger obstacles or travel on complex terrains, because the wheel-leg structures are mostly fixed, which limits the application efficiency and adaptability of the robots in different environments.
[0008] The present invention provides the following technical solution: a three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel, comprising a vehicle frame and a variable-diameter spoke wheel, wherein the variable-diameter spoke wheel is connected to the vehicle frame, the variable-diameter spoke wheels are arranged in pairs in the transverse direction of the frame, and the variable-diameter spoke wheels are arranged in a straight line in the longitudinal direction of the frame;
[0009] The variable-diameter spoke wheel includes a traveling motor, a diameter-changing motor, a wheel frame, and a wheel axle; the housing of the traveling motor is connected to the vehicle frame; one end of the wheel axle is connected to the rotating shaft of the traveling motor, and the other end is connected to the wheel frame;
[0010] There are n wheel legs arranged in a circular array centered on the wheel axle on the wheel frame, where n is an integer greater than or equal to 3. The wheel legs can slide radially along the wheel axle. The diameter-changing motor is installed on the wheel frame, and the rotating shaft of the diameter-changing motor is connected to the n wheel legs through a transmission mechanism to drive the n wheel legs to expand and contract synchronously.
[0011] Further, the rotating shaft of the diameter-changing motor is connected to an input shaft, and a driving bevel gear is installed on the input shaft. A lead screw that is threadedly engaged with the wheel leg and parallel to its sliding direction is assembled on the wheel frame. The lead screw is axially fixed and circumferentially free, and a driven bevel gear is installed on the lead screw. The driven bevel gear meshes with the driving bevel gear.
[0012] Further, a guide groove is opened along the sliding direction of the wheel leg, and a pin fixed on the wheel frame extends into the guide groove. The pin and the guide groove cooperate to form an anti-detachment limit.
[0013] Further, the diameter-changing motor is installed on the wheel frame through a pan-tilt and a connecting rod.
[0014] A control system is carried on a three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel, including a sensing module and a control module;
[0015] The sensing module is used to generate a global map of the environment around the three-dimensional terrain survey vehicle;
[0016] The control module judges the height of obstacles in the traveling direction of the three-dimensional terrain survey vehicle based on the global map, and controls the variable-diameter spoke wheel to adjust the wheel diameter or optimize the moving path of the three-dimensional terrain survey vehicle accordingly.
[0017] Further, the sensing module includes a dual-sensing system of a lidar and a depth camera. The lidar and the depth camera are installed on the vehicle frame. The lidar runs the laser SLAM algorithm to generate a two-dimensional map of the surrounding environment in real time and provides the position information of the three-dimensional terrain survey vehicle; the depth camera runs the visual SLAM algorithm to generate a three-dimensional point cloud, then filters the three-dimensional point cloud data through a voxel grid filter, and then performs three-dimensional environment modeling through RViz and generates a global map.
[0018] Further, the control module runs the RANSAC algorithm to fit the ground plane, and then analyzes the global map generated by the sensing module to judge the height of obstacles in the traveling direction of the three-dimensional terrain survey vehicle;
[0019] If the height of the obstacle is less than the maximum crossing height of the variable-diameter spoke wheel, control the variable-diameter spoke wheel to change its diameter; convert the height information of the obstacle into the pulse signal required by the diameter-changing motor, and the edge computing box inputs the pulse signal to the diameter-changing motor of each variable-diameter spoke wheel through the slip ring to dynamically adjust the diameter of each variable-diameter spoke wheel;
[0020] If the height of the obstacle is greater than the maximum crossing height of the variable-diameter spoke wheel, control the three-dimensional terrain survey vehicle to turn. According to the new movement trajectory, the driving motors on both sides of the three-dimensional terrain survey vehicle adjust the rotational speed difference to achieve the turning and path tracking of the three-dimensional terrain survey vehicle until an obstacle that can be crossed is found.
[0021] Further, the method for the control module to judge the height of the obstacle is as follows: for the filtered three-dimensional point cloud data, randomly select 3 three-dimensional points for plane fitting, and the plane equation is:
[0022] ax + by + cz + d = 0; where (a, b, c) is the normal vector of the plane, d is the distance from the plane to the origin, and x, y, z are the coordinates of the remaining three-dimensional points relative to the plane; judge whether a point is on the plane by calculating the distance from other points to this plane; the distance formula is: If the value of distance is less than 0.1, the condition is satisfied. If the number of points that meet the condition is greater than the threshold, the plane is considered to be the ground; then measure the height of the three-dimensional points at the same depth to the ground. If the height is less than the maximum crossing height of the variable-diameter spoke wheel, execute the crossing instruction, otherwise execute the turning instruction.
[0023] Further, the lidar measures the distance by continuously emitting laser beams and calculating the reflection time of the laser beams with surrounding objects to construct the environmental information on the two-dimensional plane;
[0024] The basic process of the laser SLAM algorithm is to generate scan data through the lidar and infer the position change of the three-dimensional terrain survey vehicle based on the maximum likelihood estimation; calculate using the following formula:
[0025] p(x t |z 1:t ,u 1:t ) = ηp(z t |x t )∫p(x t |u t ,x t-1 )p(x t-1 |z 1:t-1 ,u 1:t-1 )dx t-1 ;
[0026] In the formula, x t represents the pose of the three-dimensional terrain survey vehicle; z tRepresents lidar data; z 1:t Represents lidar data from the initial time to the current time; u t Is the control input of the 3D terrain survey vehicle; p(x t |z 1：t , u 1：t ) is the posterior probability distribution, indicating the pose x of the 3D terrain survey vehicle at time t under the condition of given all observation data z 1:t and control condition u 1：t ; η is the normalization constant; p(z t |x t ) is the likelihood function, indicating the probability distribution of the observation data z t under the condition of given pose x t ; p(x t |u t , x t ) is the state transition probability, indicating the probability distribution of the current pose x t-1 under the condition of given the current control input u t and the previous pose x t-1 ; p(x t |z t-1 , u 1：t-1 ) is the posterior probability distribution of the previous moment, indicating the probability distribution of the previous pose x 1：t-1 under the condition of given all previous observation data z 1：t-1 and control input u 1：t-1 ; dx t-1 is the integration variable, indicating integrating over the previous pose x t-1 t-1 i .
[0027] Furthermore, the process of filtering by the visual SLAM algorithm is to divide the 3D space where the 3D point cloud is located into voxel grids, and all points within each voxel grid are replaced by their centroids, thus significantly reducing the number of points; its formula is:
[0028]
[0029] In the formula, N represents the number of points in the voxel; x, y, z are the coordinates of the 3D points;
[0030] Estimate the relative motion of the depth camera by matching feature points in adjacent frames. This process is to solve the rotation matrix R and displacement vector t of the depth camera given the 3D point p i and the projection P i in the image, satisfying: p i = K * [R|t] * P i, where K is the internal parameter matrix of the depth camera. Through feature matching of multiple frames, the visual SLAM algorithm gradually constructs a global three-dimensional point cloud map.
[0031] Compared with the prior art, the advantages of the present invention are as follows:
[0032] A three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel provided by the present invention. The variable-diameter spoke wheel can drive the rotation of the driving bevel gear and the driven bevel gear through a built-in motor, thereby realizing the radial expansion and contraction of the spokes and changing the wheel diameter. With this design, the three-dimensional terrain survey vehicle can automatically adjust the wheel diameter according to different terrain conditions to improve its passability and adaptability, enabling the three-dimensional terrain survey vehicle to better adapt to various terrains, facilitating crossing obstacles, and enhancing its application potential and efficiency in complex environments. In addition, this design can also enhance the stability and energy efficiency of the three-dimensional terrain survey vehicle at different speeds, further expanding its usage scenarios and effectiveness.
[0033] The control system of the three-dimensional terrain survey vehicle based on the variable-diameter spoke wheel adopts multi-sensor fusion, including lidar and depth camera. By using three-dimensional SLAM technology (such as RTAB-Map), it realizes real-time detection of unknown terrain and three-dimensional map construction. The multi-sensor fusion scheme improves the accuracy and stability of environmental perception and can effectively cope with complex and changing environments. BRIEF DESCRIPTION OF THE DRAWINGS
[0034] Figure 1 is a schematic diagram of a three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel;
[0035] Figure 2 is a schematic diagram of the inside of the wheel frame;
[0036] Figure 3 is a schematic diagram of the connection between the variable-diameter motor and the wheel frame.
[0037] In the figure: 1 - vehicle frame; 2 - traveling motor; 3 - variable-diameter motor; 4 - wheel frame; 5 - input shaft; 6 - wheel leg; 7 - driving bevel gear; 8 - lead screw; 9 - driven bevel gear; 10 - pan-tilt; 11 - connecting rod; 12 - lidar; 13 - depth camera; 14 - slip ring; 15 - edge computing box. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.
[0039] Embodiment 1
[0040] As Figure 1 , Figure 2 , Figure 3 shown: A three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel, including a vehicle frame 1 and variable-diameter spoke wheels. The variable-diameter spoke wheels are connected to the vehicle frame 1. The vehicle frame 1 is horizontal, and the variable-diameter spoke wheels are paired in twos. The vehicle frame 1 is longitudinal, and the variable-diameter spoke wheels are arranged in a straight line.
[0041] The variable-diameter spoke wheel includes a traveling motor 2, a diameter-changing motor 3, a wheel frame 4, and a wheel axle. The housing of the traveling motor 2 is connected to the vehicle frame 1. One end of the wheel axle is connected to the rotating shaft of the traveling motor 2, and the other end is connected to the wheel frame 4. There are n wheel legs 6 arranged in a circular array around the wheel axle on the wheel frame 4, where n is an integer greater than or equal to 3. In this embodiment, 3 wheel legs 6 are used. The wheel legs 6 can slide radially along the wheel axle. The diameter-changing motor 3 is installed on the wheel frame 4, and the rotating shaft of the diameter-changing motor 3 is connected to the n wheel legs 6 through a transmission mechanism to drive the n wheel legs 6 to synchronously extend and retract.
[0042] Through the design of the variable-diameter spoke wheel, the adaptability of the three-dimensional terrain survey vehicle on different terrains is significantly improved. Through the radial movement of the wheel legs 6, the three-dimensional terrain survey vehicle can adjust the wheel diameter according to the terrain characteristics, so that it can maintain good stability and mobility on soft or uneven ground such as soft soil and sand.
[0043] The design of the telescopic wheel legs 6 enables the three-dimensional terrain survey vehicle to increase the wheel diameter to improve the passing ability when encountering large obstacles such as stones or steps. This design greatly improves the practicality and flexibility of the robot in complex environments.
[0044] The rotating shaft of the diameter-changing motor 3 is connected to the input shaft 5. A driving bevel gear 7 is installed on the input shaft 5. A lead screw 8 that is threadedly engaged with the wheel leg 6 and is parallel to its sliding direction is assembled on the wheel frame 4. The lead screw 8 is axially fixed and circumferentially free. A driven bevel gear 9 is installed on the lead screw 8, and the driven bevel gear 9 meshes with the driving bevel gear 7. The driving bevel gear 7 drives the three driven bevel gears 9 arranged around it. Through the transmission of the driving bevel gear 7 and the driven bevel gear 9, the function of one driving three is realized. At the same time, through the rotation of the driven bevel gear 9, the lead screw 8 is driven to rotate. By restricting two degrees of freedom of the wheel leg 6 by the wheel frame 4, it can only perform the movement of one degree of freedom of forward and backward telescoping. Finally, the three wheel legs 6 are synchronously extended and retracted, realizing the change of the radial length of the overall spoke wheel.
[0045] Through the design of the built-in driving bevel gear 7 and driven bevel gear 9, the compactness and high-efficiency transmission of the wheel body structure are realized. The transmission efficiency of the driving bevel gear 7 and the driven bevel gear 9 is high, and the structure is more compact. At the same time, by restricting the movement degrees of freedom of the wheel leg 6, the stability and durability of the overall mechanism are improved.
[0046] A guide groove is formed along the sliding direction on the wheel leg 6, and a pin fixed on the wheel frame 4 extends into the guide groove. The pin and the guide groove cooperate to form anti - detachment limit, preventing the wheel leg 6 from falling off.
[0047] The variable - diameter motor 3 is installed on the wheel frame 4 through the cloud platform 10 and the connecting rod 11. The variable - diameter motor 3 and the traveling motor 2 are located on both sides of the wheel frame 4.
[0048] The ground - touching end of the wheel leg 6 is a circumferential surface, which improves the smoothness when the variable - diameter spoke wheel rotates.
[0049] Embodiment 2
[0050] A control system is carried on a three - dimensional terrain survey vehicle based on a variable - diameter spoke wheel in Embodiment 1, and includes a perception module and a control module;
[0051] The perception module is used to generate a global map of the environment around the three - dimensional terrain survey vehicle;
[0052] The control module judges the height of obstacles in the traveling direction of the three - dimensional terrain survey vehicle according to the global map, and thus controls the variable - diameter spoke wheel to adjust the wheel diameter or optimize the moving path of the three - dimensional terrain survey vehicle.
[0053] The perception module includes a dual - line perception system of a lidar 12 and a depth camera 13. The lidar 12 and the depth camera 13 are installed on the vehicle frame 1. The lidar 12 runs the laser SLAM algorithm to generate a two - dimensional map of the surrounding environment in real time and provides the position information of the three - dimensional terrain survey vehicle; the depth camera 13 runs the visual SLAM algorithm to generate a three - dimensional point cloud. To reduce the computational load, the three - dimensional point cloud data is filtered through a voxel grid filter, and then a three - dimensional environment modeling is carried out through RViz to generate a global map.
[0054] The control module runs the RANSAC algorithm to fit the ground plane, and then analyzes the global map generated by the perception module to judge the height of obstacles in the traveling direction of the three - dimensional terrain survey vehicle;
[0055] If the height of the obstacle is less than the maximum crossing height of the variable - diameter spoke wheel, then control the variable - diameter spoke wheel to change its diameter; convert the height information of the obstacle into the pulse signal required by the variable - diameter motor 3, and the edge computing box 15 inputs the pulse signal to the variable - diameter motor 3 of each variable - diameter spoke wheel through the slip ring 14 to dynamically adjust the diameter of each variable - diameter spoke wheel;
[0056] If the height of the obstacle is greater than the maximum crossing height of the variable-diameter spoke wheel, the three-dimensional terrain survey vehicle is controlled to turn. According to the new motion trajectory, the walking motors 2 on both sides of the three-dimensional terrain survey vehicle adjust the rotational speed difference to achieve the turning and path tracking of the three-dimensional terrain survey vehicle until an obstacle that can be crossed is found. The accuracy of differential control directly affects the maneuverability of the three-dimensional terrain survey vehicle in narrow or complex terrains, ensuring its efficient navigation and movement during free exploration.
[0057] The lidar 12 measures distances by continuously emitting laser beams and calculating the reflection time of the laser beams with surrounding objects, and constructs environmental information on a two-dimensional plane;
[0058] The basic process of the laser SLAM algorithm is to generate scan data through the lidar 12 and infer the position change of the three-dimensional terrain survey vehicle based on the maximum likelihood estimation; it is calculated using the following formula:
[0059] p(x t |z 1:t ,u 1:t ) = ηp(z t |x t )∫p(x t |u t ,x t-1 )p(x t-1 |z 1:t-1 ,u 1:t-1 )dx t-1 ;
[0060] In the formula, x t represents the pose of the three-dimensional terrain survey vehicle; z t represents the lidar data; z 1:t represents the lidar data from the initial moment to the current moment; u t is the control input of the three-dimensional terrain survey vehicle; p(x t |z 1：t ,u 1：t ) is the posterior probability distribution, indicating the probability distribution of the pose x 1:t of the three-dimensional terrain survey vehicle at time t under the condition of given all observation data z 1：t and control condition u t ; η is the normalization constant; p(z t |x t ) is the likelihood function, indicating the probability distribution of the observation data z t under the condition of given pose x t ; p(x t |u t ,x t-1 ) is the state transition probability, indicating the probability distribution of the pose x at the previous moment given the current control input u t and the pose x at the previous momentt-1 Under the condition of, the pose x at the current moment t Probability distribution; p(x t-1 |z 1：t-1 , u 1：t-1 ) is the posterior probability distribution at the previous moment, indicating that under the condition of given all previous observation data z 1：t-1 and control input u 1：t-1 the probability distribution of the pose x at the previous moment; dx t-1 is the integration variable, indicating the integration of the pose x at the previous moment t-1 t-1 i .
[0061] The formula estimates the motion trajectory and current state of the three-dimensional terrain survey vehicle by fusing sensor data and prior information.
[0062] The process of visual SLAM algorithm filtering is to divide the three-dimensional space where the three-dimensional point cloud is located into voxel grids, and all points within each voxel grid are replaced by their centroids, thus significantly reducing the number of points; its formula is:
[0063] In the formula, N represents the number of points in the voxel; x, y, z are the coordinates of the three-dimensional points;
[0064] In this way, the geometric characteristics of the original point cloud can be maintained while reducing the computational complexity.
[0065] Next, the relative motion of the depth camera is estimated by matching the feature points in adjacent frames. This process is to solve the rotation matrix R and displacement vector t of the depth camera given the three-dimensional point p i and the projection P in the image, satisfying: p i = K * [R|t] * P i , where K is the internal parameter matrix of the depth camera. Through multi-frame feature matching, the visual SLAM algorithm gradually constructs a global three-dimensional point cloud map.
[0066] In addition, the perception module also integrates the data of IMU (spatial attitude sensor) and odometer to perform multi-sensor data fusion through the Extended Kalman Filter (EKF) to further improve the accuracy and robustness of the system. The acceleration and angular velocity data provided by the IMU can smooth the motion trajectory of the three-dimensional terrain survey vehicle through the prediction-update mechanism and reduce the influence of sensor noise. The prediction equation of the EKF is: The update equation is: where K t is the Kalman gain, and z t is the sensor measurement value.
[0067] The method for the control module to judge the height of the obstacle is as follows: for the filtered three-dimensional point cloud data, randomly select 3 three-dimensional points for plane fitting, and the plane equation is:
[0068] ax + by + cz + d = 0; where (a, b, c) is the normal vector of the plane, d is the distance from the plane to the origin, and x, y, z are the coordinates of the remaining three-dimensional points relative to the plane. Determine whether a point is on the plane by calculating the distance from other points to this plane; the distance formula is: If the value of distance is less than 0.1, the condition is met. If the number of points that meet the condition is greater than the threshold, the plane is considered to be the ground; then measure the height from the three-dimensional points at the same depth to the ground. If the height is less than the maximum crossing height of the variable-diameter spoke wheel, execute the crossing instruction, otherwise execute the turning instruction.
[0069] The crossing instruction specifically first adjusts the length of the wheel leg 6 through the variable-diameter motor 3 so that R 辐条轮 =(h obstacle + 20) mm, where the speed of the wheel leg 6 can be obtained from the real-time rotation speed of the variable-diameter motor 3. where z 1 is the number of teeth of the driving bevel gear, z 2 is the number of teeth of the driven bevel gear, n 1 is the rotation speed of the driving bevel gear, and L is the lead of the lead screw. Secondly, adjust the pose of the three-dimensional terrain survey vehicle, and perform self-rotation through the differential rotation function of the walking motor 2 to make the three-dimensional terrain survey vehicle move forward in the direction with the lowest obstacle height, and finally cross the obstacle and move forward.
[0070] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present invention. Various modifications to these embodiments will be obvious to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein, but will be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. A three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel, characterized in that: It comprises a vehicle frame (1) and a variable-diameter spoke wheel, wherein the variable-diameter spoke wheel is connected to the vehicle frame (1), the variable-diameter spoke wheels are arranged in pairs in the horizontal direction of the vehicle frame (1), and the variable-diameter spoke wheels are arranged along a straight line in the vertical direction of the vehicle frame (1); The variable-diameter spoke wheel comprises a travel motor (2), a variable-diameter motor (3), a wheel frame (4) and a wheel axle; the housing of the travel motor (2) is connected to the vehicle frame (1); one end of the wheel axle is connected to the rotating shaft of the travel motor (2), and the other end is connected to the wheel frame (4); The wheel frame (4) is provided with n wheel legs (6) arranged in a circular array with the wheel axle as the center, where n is an integer ≥ 3, and the wheel legs (6) can slide radially along the wheel axle. The variable diameter motor (3) is mounted on the wheel frame (4), and the rotating shaft of the variable diameter motor (3) is connected to the n wheel legs (6) through a transmission mechanism, so as to drive the n wheel legs (6) to extend and retract synchronously.
2. A three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel according to claim 1, characterized in that: The rotating shaft of the variable diameter motor (3) is connected to the input shaft (5), a driving bevel gear (7) is installed on the input shaft (5), a screw rod (8) parallel to the sliding direction of the wheel leg (6) is threadedly engaged, the screw rod (8) is assembled on the wheel frame (4), the screw rod (8) is axially fixed and circumferentially free, a driven bevel gear (9) is installed on the screw rod (8), and the driven bevel gear (9) is engaged with the driving bevel gear (7).
3. A three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel according to claim 2, characterized in that: A guide groove is formed on the wheel leg (6) along its sliding direction, and a pin fixed on the wheel frame (4) extends into the guide groove, and the pin cooperates with the guide groove to form an anti-slip limiter.
4. A three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel according to claim 1, characterized in that: The variable diameter motor (3) is installed on the wheel frame (4) via a pan head (10) and a connecting rod (11).
5. A control system, mounted on a three-dimensional terrain survey vehicle based on a variable-diameter spoke wheel according to claim 1, characterized in that: Includes a perception module and a control module; The perception module is used to generate a global map of the environment around the 3D terrain survey vehicle; The control module determines the height of obstacles in the direction of travel of the three-dimensional terrain survey vehicle according to the global map, and controls the variable-diameter spoke wheel to adjust the wheel diameter or optimize the moving path of the three-dimensional terrain survey vehicle.
6. A control system according to claim 5, characterized in that: The perception module comprises a dual-line perception system of a laser radar (12) and a depth camera (13). The laser radar (12) and the depth camera (13) are mounted on a vehicle frame (1). The laser radar (12) runs a laser SLAM algorithm to generate a two-dimensional map of the surrounding environment in real time and provide position information of a three-dimensional terrain survey vehicle. The depth camera (13) runs a visual SLAM algorithm to generate a three-dimensional point cloud, and then filters the three-dimensional point cloud data through a voxel grid filter, and then performs three-dimensional environment modeling through RViz and generates a global map.
7. A control system according to claim 6, characterized in that: The control module runs the RANSAC algorithm to fit the ground plane, and then analyzes the global map generated by the perception module to determine the height of obstacles in the direction of travel of the three-dimensional terrain survey vehicle; If the height of the obstacle is less than the maximum span height of the variable-diameter spoke wheel, the variable-diameter spoke wheel is controlled to change diameter; the height information of the obstacle is converted into a pulse signal required by the variable-diameter motor (3), and the edge computing box (15) inputs the pulse signal to the variable-diameter motor (3) of each variable-diameter spoke wheel through the slip ring (14), so as to dynamically adjust the diameter of each variable-diameter spoke wheel; If the height of the obstacle is greater than the maximum crossing height of the variable-diameter spoke wheel, the three-dimensional terrain survey vehicle is controlled to turn, and the travel motors (2) on both sides of the three-dimensional terrain survey vehicle adjust the speed difference according to the new motion trajectory to achieve the turning and path tracking of the three-dimensional terrain survey vehicle until a traversable obstacle is found.
8. A control system according to claim 7, characterized in that: The control module determines the height of the obstacle by randomly finding three three-dimensional points for plane fitting based on the filtered three-dimensional point cloud data. The plane equation is: ax+by+cz+d=0; (a, b, c) is the normal vector of the plane, d is the distance from the plane to the origin, x, y, z are the coordinates of the remaining three-dimensional points relative to the plane; the distance from other points to the plane is calculated to determine whether it is a point on the plane; the distance formula is: If the value of distance is less than 0.1, the condition is met. If the number of points that meet the condition is greater than the threshold, the plane is considered to be the ground. Then the height from the three-dimensional point at the same depth to the ground is measured. If the height is less than the maximum span height of the variable-diameter spoke wheel, the span instruction is executed, otherwise the steering instruction is executed.
9. A control system according to claim 6, characterized in that: The laser radar (12) measures the distance by continuously emitting laser beams and calculating the reflection time between the laser beams and surrounding objects, thereby constructing environmental information on a two-dimensional plane; The basic process of the laser SLAM algorithm is to generate scanning data through a laser radar (12) and infer the position change of the three-dimensional terrain survey vehicle based on maximum likelihood estimation; the following formula is used for calculation: p(x t |z 1:t ,u 1:t )=ηp(z t |x t )∫p(x t |u t ,x t-1 )p(x t-1 |z 1:t-1 ,u 1:t-1 )dx t-1 ; In the formula, x t represents the position and posture of the 3D terrain survey vehicle; z t Represents lidar data; z 1:t Represents the laser radar data from the initial moment to the current moment; u t is the control input of the 3D terrain survey vehicle; p(x t |z 1：t ,u 1：t ) is the posterior probability distribution, which means that given all the observed data z 1:t and control condition u 1：t Under the condition of t The probability distribution of t |x t ) is the likelihood function, which means that given the position x t Under the condition of t The probability distribution of p(x t |u t , x t-1 ) is the state transition probability, indicating that given the current control input u t and the previous moment's position x t-1 Under the condition of t The probability distribution of p(x t-1 |z 1：t ,u 1：t-1 ) is the posterior probability distribution of the previous moment, indicating that given all previous observation data z 1：t-1 and control input u 1：t-1 Under the condition of t-1 The probability distribution of dx t-1 is the integral variable, which represents the position x at the previous moment t-1 Perform integration.
10. A control system according to claim 6, characterized in that: The filtering process of the visual SLAM algorithm is to divide the three-dimensional space where the three-dimensional point cloud is located into voxel grids, and all points in each voxel grid are replaced by their centroids, thereby significantly reducing the number of points; The formula is: Where N is the number of points in the voxel; x, y, z are the coordinates of the three-dimensional point; The relative motion of the depth camera is estimated by matching feature points in adjacent frames. i And the projection P in the image i , solve the rotation matrix R and displacement vector t of the depth camera, satisfying: p i =K*[R|t]*P i , where K is the intrinsic parameter matrix of the depth camera. Through multi-frame feature matching, the visual SLAM algorithm gradually constructs a global three-dimensional point cloud map.