A foot end ground contact information detection method, device and equipment based on a wheel-legged robot and a medium

Abstract

The present application provides a method, device, equipment, and medium for detecting foot-end ground contact information of a wheel-legged robot, which relates to the technical field of wheel-legged robots. The method includes: obtaining tire strain data of the wheel-legged robot through a signal acquisition system provided on the wheel-legged robot; obtaining acceleration data of the contact point when the wheel-legged robot's tire is in a ground contact state; obtaining contact force data of the wheel-legged robot and the spatial position of the tire contact point through the tire strain data and acceleration data based on the established mapping relationship between the input vector and the output vector; and analyzing the ground contact information of the wheel-legged robot's position based on the contact force data and the spatial position. The present application can accurately detect the ground contact information of the wheel-legged robot.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wheel-legged robots, and particularly relates to a wheel-legged robot-based foot end ground contact information detection method, device, equipment and medium. BACKGROUND

[0002] A wheel-legged robot is a robot combining the characteristics of wheeled and legged movement, and has high terrain adaptability. It is usually composed of wheels and legs, and can move freely in complex terrain and different environments, so that the wheel-legged robot can cope with various complex environments when performing tasks, and is widely used in exploration, rescue, military, and special operation fields.

[0003] Wheel-legged robot foot end ground contact information detection is crucial for its stability and adaptability. Since the wheel-legged robot needs to perform tasks on complex and variable terrain, accurate perception of the foot end ground contact state can provide real-time feedback on the ground characteristics, helping the robot to judge the hardness, inclination and obstacles of the current terrain. This information plays an important role in switching movement modes, adjusting movement strategies and optimizing control systems. Especially in complex terrain or extreme environments, detection of foot end ground contact information can effectively prevent the robot from losing stability or failing to pass through obstacles, improving its adaptability and task execution ability in dynamic environments. Therefore, accurate ground contact information detection is not only the basis for efficient and safe operation of the wheel-legged robot, but also an indispensable part of its intelligent and automated control system. SUMMARY

[0004] The present application provides a wheel-legged robot-based foot end ground contact information detection method, device, equipment and medium, which can accurately detect the ground contact information of the wheel-legged robot.

[0005] In a first aspect of the present application, a wheel-legged robot-based foot end ground contact information detection method is provided, which is applied to a server, and the method comprises:

[0006] Obtaining tire strain data of the wheel-legged robot through a signal acquisition system arranged on the wheel-legged robot;

[0007] Obtaining acceleration data of the ground contact point when the tire of the wheel-legged robot is in a ground contact state;

[0008] According to the mapping relationship between the established input vector and output vector, the tire strain data and the acceleration data are used to obtain contact force data of the wheel-legged robot and the spatial position of the tire ground contact point;

[0009] Based on the contact force data and the spatial position, the ground contact information of the position where the wheel-legged robot is located is analyzed.

[0010] Preferably, based on the contact force data and the spatial position, ground contact information of a position where the wheel-legged robot is located is analyzed, specifically including:

[0011] Based on the contact force data, a contact area of a contact area of the wheel-legged robot is calculated, the contact area being an area where a tire of the wheel-legged robot contacts the ground;

[0012] According to the contact force data and the contact area, a change rate of the contact force in the contact area is calculated;

[0013] Based on a change of the spatial position, a ground deformation variable is calculated;

[0014] According to the change rate and the ground deformation data, a ground hardness is calculated;

[0015] A size relationship between the ground hardness and different preset thresholds is judged to determine a ground hardness category of the contact area.

[0016] Preferably, according to the change rate and the ground deformation variable, the ground hardness is calculated, specifically through the following formula:

[0017]

[0018] wherein H is the ground hardness, F z is a normal reaction force contained in the contact force data, Δz is the ground deformation variable, A is the contact area, and ΔF z is a change amount of the contact force.

[0019] Preferably, before the contact force data and the spatial position of the tire contact point of the wheel-legged robot are obtained through the tire strain data and the acceleration data according to the mapping relationship between the established input vector and output vector, the method further includes:

[0020] The input vector is constructed, wherein the input vector is composed of the acceleration data and multiple groups of the tire strain data;

[0021] The output vector is constructed, wherein the output vector includes the contact force data and the spatial position;

[0022] The Gaussian process regression model is trained through training data, wherein the training data includes training acceleration data and multiple groups of training tire strain data as input, and training contact force data and training spatial position as output;

[0023] The Gaussian process regression model is trained by the training data, and the Gaussian process regression model is used to model a mapping relationship between the input vector and the output vector, wherein the mapping relationship is represented as follows:

[0024] Y j = k * (K-σ n I) -1 y j

[0025] wherein Y j is the output vector, Y j = [F x , F y , F z ] T , y j is the input vector, σ n is a standard deviation, the input vector is represented as x = [s1, s2] T , k * is a variable irrelevant to the input, and K is a covariance matrix.

[0026] Based on the above technical solutions, preferably, according to the established mapping relationship between the input vector and the output vector, the contact force data of the wheel-legged robot and the spatial position of the tire contact point are obtained through the tire strain data and the acceleration data, and specifically include:

[0027] The tire strain data and the acceleration data are input to the trained Gaussian process regression model, and the contact force data and the spatial position of the tire contact point output by the trained Gaussian process regression model are obtained, wherein the contact force data includes x-axis direction shear force, y-axis direction shear force and normal reaction force, and the spatial position includes the spatial coordinates of the contact point.

[0028] Based on the above technical solutions, preferably, the signal acquisition system includes a plurality of sensor units, wherein:

[0029] The plurality of sensor units are embedded in the solid tire of the wheel-legged robot, and are distributed in the tire in a ring shape at equal angles with the tire rotation axis as the center. Each sensor unit includes a polytetrafluoroethylene hose and a resistance strain gauge, and the resistance strain gauge includes a resin cover sheet, a metal foil and a resin substrate.

[0030] The resistance strain gauge is arranged in the central region of the polytetrafluoroethylene hose, and a two-component acrylic adhesive is injected into the polytetrafluoroethylene hose to solidify and package the resistance strain gauge.

[0031] The metal foil is disposed directly above the resin base, and the resin cover sheet is disposed directly above the metal foil.

[0032] On the basis of the above technical solutions, preferably, the sensor unit further comprises a strain gauge lead wire, and the server is communicatively connected with the sensor unit based on the strain gauge lead wire.

[0033] The tire strain data of the wheel-legged robot is obtained through the signal acquisition system disposed on the wheel-legged robot, and specifically includes:

[0034] A plurality of resistance data transmitted by the sensor unit is obtained, wherein when external pressure acts on the solid tire, the external pressure is transmitted to the polytetrafluoroethylene hose through the solid tire and to the resistance strain gauge, and the resistance value of the resistance strain gauge changes linearly and proportionally with the external pressure.

[0035] According to the plurality of resistance data, a resistance change value is calculated.

[0036] According to the resistance change value, a strain value is calculated to obtain the tire strain data, and specifically the strain value is calculated through the following formula:

[0037]

[0038] wherein ΔR is the resistance change value, R is an initial resistance value, K s is a strain sensitivity, and ε is the strain value.

[0039] In a second aspect of the present application, a foot end ground contact information detection device based on a wheel-legged robot is provided, the device is a server, and includes an acquisition module, a processing module, and an output module, wherein:

[0040] The acquisition module is configured to obtain tire strain data of the wheel-legged robot through a signal acquisition system disposed on the wheel-legged robot.

[0041] The acquisition module is configured to obtain acceleration data of a ground contact point when a tire of the wheel-legged robot is in a ground contact state.

[0042] The processing module is configured to obtain contact force data of the wheel-legged robot and a spatial position of a tire ground contact point according to a mapping relationship between an established input vector and an output vector, through the tire strain data and the acceleration data.

[0043] The output module is configured to analyze ground contact information of a position where the wheel-legged robot is located based on the contact force data and the spatial position.

[0044] On the basis of the above technical solutions, preferably, the processing module is configured to calculate a contact area of a contact region of the wheel-legged robot based on the contact force data, the contact region being a region where a tire of the wheel-legged robot contacts the ground.

[0045] The processing module is configured to calculate a rate of change of the contact force in the contact region according to the contact force data and the contact area.

[0046] The processing module is configured to calculate a ground deformation variable based on the change in the spatial position.

[0047] The processing module is configured to calculate a ground hardness according to the rate of change and the ground deformation data.

[0048] The processing module is configured to determine a ground hardness category of the contact region by judging a size relationship between the ground hardness and different preset threshold values.

[0049] On the basis of the above technical solutions, preferably, the processing module is configured to calculate the ground hardness according to the rate of change and the ground deformation variable, specifically by using the following formula:

[0050]

[0051] wherein H is the ground hardness, F z is a normal reaction force contained in the contact force data, Δz is the ground deformation variable, A is the contact area, and ΔF z is a change amount of the contact force.

[0052] On the basis of the above technical solutions, preferably, the processing module is configured to construct the input vector, wherein the input vector is composed of the acceleration data and a plurality of groups of the tire strain data.

[0053] The processing module is configured to construct an output vector, wherein the output vector includes the contact force data and the spatial position.

[0054] The processing module is configured to train a Gaussian process regression model by using training data, wherein the training data includes training acceleration data and a plurality of groups of training tire strain data as input, and training contact force data and training spatial position as output.

[0055] The output module is configured to train the Gaussian process regression model by using the training data, and model a mapping relationship between the input vector and the output vector by using the Gaussian process regression model, wherein the mapping relationship is represented as follows:

[0056] Y j= k * (K-σ n Ι) -1 y j

[0057] wherein Y j is the output vector, Y j = [F x , F y , F z ] T , y j is the input vector, σ n is the standard deviation, the input vector is represented as x = [s1, s2] T , k * is a variable independent of the input quantity, and K is the covariance matrix.

[0058] On the basis of the above technical scheme, preferably, the output module is configured to input the tire strain data and the acceleration data into the trained Gaussian process regression model, and obtain contact force data and a spatial position of a tire contact point output by the trained Gaussian process regression model, wherein the contact force data includes shear force in the x-axis direction, shear force in the y-axis direction, and normal reaction force, and the spatial position includes a spatial coordinate of the contact point.

[0059] On the basis of the above technical scheme, preferably, the acquisition module is configured to acquire a plurality of resistance data transmitted by the sensor unit, wherein when external pressure acts on the solid tire, the external pressure is transmitted to the polytetrafluoroethylene hose through the solid tire and to the resistance strain gauge, and a resistance value of the resistance strain gauge changes linearly and proportionally with the external pressure.

[0060] The processing module is configured to calculate a resistance change value according to the plurality of resistance data.

[0061] The processing module is configured to calculate a strain value according to the resistance change value to obtain the tire strain data, specifically by the following formula:

[0062]

[0063] wherein ΔR is the resistance change value, R is an initial resistance value, K s is a strain sensitivity, and ε is the strain value.

[0064] In a third aspect of the present application, an electronic device is provided, comprising a processor, a memory, a user interface and a network interface, the memory is configured to store instructions, the user interface and the network interface are configured to communicate with other devices, and the processor is configured to execute the instructions stored in the memory to enable the electronic device to perform the method according to any one of the preceding aspects.

[0065] In a fourth aspect of the present application, a computer-readable storage medium is provided, which stores instructions that, when executed, perform the method according to any one of the preceding aspects.

[0066] In summary, the one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0067] 1. The present application can accurately detect the ground contact information of the wheel-legged robot, because the strain data of the tire and the acceleration data of the ground contact point are combined to establish the mapping relationship between the input vector and the output vector by using the Gaussian process regression model, and the contact force data and the spatial position of the tire contact point are accurately calculated. This multi-source data fusion method can not only capture the force condition when the tire contacts the ground, but also combine the spatial coordinates of the contact point to comprehensively analyze the contact state of the wheel-legged robot, and then accurately reflect the hardness, shape and contact force distribution of the ground, thereby improving the detection accuracy and reliability of the ground contact information.

[0068] 2. By comprehensively considering the contact force data, the contact area, the contact force change rate and the ground deformation data, the ground hardness is accurately calculated, and the classification is performed through the set threshold. Thus, the robot can help to perceive the ground characteristics in real time, such as hard road surface or soft road surface.

[0069] 3. The mapping relationship between the strain data and the acceleration data of the tire and the contact force data and the spatial position of the tire contact point is accurately established by using the Gaussian process regression model, which can efficiently calculate the contact force and the ground contact position of the wheel-legged robot. This method combines multiple sets of sensor data through model training to ensure high-precision detection in various complex terrains and different load conditions.

[0070] 4. The resistance strain gauge is packaged and embedded in the solid tire through drilling, which has the advantages of simple installation and high measurement accuracy. The solid tire detection method has reliable dynamic measurement capability in the tire impact loading scene. BRIEF DESCRIPTION OF DRAWINGS

[0071] Figure 1 is a flowchart of a foot-end ground contact information detection method based on a wheel-legged robot according to an embodiment of the present application;

[0072] Figure 2is an application scenario schematic diagram of a foot end ground contact information detection method based on a wheel-legged robot disclosed by an embodiment of the present application.

[0073] Figure 3 is a simulation result schematic diagram of finite element analysis on solid tire ground contact, disclosed by an embodiment of the present application.

[0074] Figure 4 is a sensor unit arrangement schematic diagram, disclosed by an embodiment of the present application.

[0075] Figure 5 is a sensor unit structure schematic diagram, disclosed by an embodiment of the present application.

[0076] Figure 6 is a different road surface strain signal change schematic diagram, disclosed by an embodiment of the present application.

[0077] Figure 7 is a module schematic diagram of a foot end ground contact information detection device based on a wheel-legged robot, disclosed by an embodiment of the present application.

[0078] Figure 8 is a structure schematic diagram of an electronic device, disclosed by an embodiment of the present application.

[0079] Legend: 201, sensor unit; 202, acquisition module; 203, lower computer; 204, server; 501, polytetrafluoroethylene hose; 502, resistance strain gauge; 5021, metal foil; 5022, resin base; 5023, resin cover sheet; 503, strain gauge lead; 701, acquisition module; 702, processing module; 703, output module; 801, processor; 802, communication bus; 803, user interface; 804, network interface; 805, memory. DETAILED DESCRIPTION

[0080] In order for those skilled in the art to better understand the technical solutions in the specification, the technical solutions in the specification will be clearly and completely described below in conjunction with the drawings in the embodiments of the specification. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments.

[0081] In the description of the embodiments of the present application, the words such as "for example" or "for instance" are used to represent an example, illustration or description. Any embodiment or design scheme described as "for example" or "for instance" in the embodiments of the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the words such as "for example" or "for instance" are intended to present the relevant concept in a specific way.

[0082] In the description of the embodiments of the present application, the term "a plurality of" means two or more. For example, a plurality of systems means two or more systems, and a plurality of screen terminals means two or more screen terminals. In addition, the terms "first", "second", "third", etc. are used only for descriptive purposes and should not be construed as indicating or implying relative importance or indicating the technical features indicated. Therefore, the features defined as "first", "second", etc. can be explicitly or implicitly included one or more features. The terms "include", "contain", "have" and their variants mean "include but are not limited to", unless otherwise specifically emphasized.

[0083] Wheel-legged robots combine the characteristics of wheeled and legged motion, have strong terrain adaptability, and can move flexibly in complex environments. Foot end ground contact information detection is crucial for its stability and adaptability, and can accurately perceive the hardness, inclination and obstacles of the ground, etc. to provide real-time feedback for the robot to adjust the motion mode and optimize the control system. In complex terrain or extreme environments, accurate ground contact information can effectively improve the adaptability and task execution ability of the robot in dynamic environments, and ensure its efficient and safe operation.

[0084] The embodiment discloses a foot end ground contact information detection method based on a wheel-legged robot, referring to Figure 1 , comprising the following steps S110-S140:

[0085] S110, acquiring tire strain data of the wheel-legged robot through a signal acquisition system arranged on the wheel-legged robot.

[0086] The foot end ground contact information detection method based on a wheel-legged robot disclosed in the embodiment of the present application is applied to a server 204. The server 204 includes but is not limited to electronic devices such as mobile phones, tablet computers, wearable devices, PC (Personal Computer), etc. It can also be a background server 204 running a foot end ground contact information detection method based on a wheel-legged robot. The server 204 can be implemented by an independent server 204 or a server 204 cluster composed of multiple servers 204.

[0087] Wheel-legged robots are equipped with two independent control systems and wheel-legged structures, so studying foot end contact information is important for improving its motion stability and decision-making performance. In the design of the wheel-legged robot foot bottom contact unit, the air-filled tire used in the past is easy to be punctured by sharp objects in rough terrain and needs frequent maintenance to maintain normal tire pressure, and is not suitable for long-period special operations in extreme environments. In contrast, the solid tire made of polyurethane has better puncture resistance and smaller deformation under heavy load, making it suitable for long-term load operations of wheel-legged robots in dangerous environments.

[0088] In order to effectively obtain contact information from the solid tire in the dynamic process of the wheel-legged robot, a sensor packaging structure that is lighter, softer and more impact-resistant is needed. The previous research mainly focuses on the wheel-legged robot ground contact state detection method when the pneumatic tire is used as the foot end. Usually, the surface sensor is attached to the bottom or temple side of the tire, and the ground contact state of the wheel-legged robot is determined by measuring the voltage signal output by the sensor. However, the above method is not suitable for the ground contact state detection of the solid tire, and there is currently a lack of effective detection method for the ground contact state of the solid tire wheel-legged robot.

[0089] In one possible implementation, the signal acquisition system includes a plurality of sensor units 201, wherein the plurality of sensor units 201 are embedded in the solid tire of the wheel-legged robot, and are distributed in the tire in a ring shape at equal angles with the tire rotation axis as the center. Each sensor unit 201 includes a polytetrafluoroethylene hose 501 and a resistance strain gauge 502, and the resistance strain gauge 502 includes a resin cover sheet 5023, a metal foil 5021, and a resin substrate 5022. The resistance strain gauge 502 is arranged in the central region of the polytetrafluoroethylene hose 501, and a two-component acrylic ester adhesive is injected into the polytetrafluoroethylene hose 501 to solidify and encapsulate the resistance strain gauge 502. The metal foil 5021 is arranged directly above the resin substrate 5022, and the resin cover sheet 5023 is arranged directly above the metal foil 5021.

[0090] Specifically, referring to Figure 2 , the signal acquisition system lower board, the acquisition module 202, and the plurality of sensor units 201, the lower board and the acquisition module 202 read the sensor signals output by the sensor units 201 from the sensor units 201 through an interface at a sampling frequency of 200 Hz and convert them into 10-bit analog values through amplification. In order to achieve this function, the present application further performs finite element analysis on the solid tire ground contact situation in a simulation environment, referring to Figure 3 , which is a simulation result diagram, and determines the strain sensitive area as the area marked in the figure.

[0091] Based on the above conclusion, the application proposes to embed strain gauges into the solid tire through drilling, to complete the design scheme of sensor unit 201. Multiple sensor units 201 need to be embedded into the solid tire of the wheel-legged robot. The sensor unit 201 is centered on the rotation axis of the tire and uniformly distributed in a ring shape. The distribution angle of each sensor unit 201 is equal, which can ensure comprehensive monitoring of the entire tire contact area. This method effectively utilizes the closed structure of the solid tire and provides a good protection environment for the sensor. However, in the actual application of the foot sensor of the wheel-legged robot, the solid tire has a high material rigidity and produces a small strain under stress, making it difficult for the distal strain gauge to effectively capture the touch signal and reducing the sensitivity of the sensor to the foot end contact state. In addition, the wheel-legged robot faces complex terrain conditions in actual operation and needs to frequently switch between wheel-legged working modes. When the robot switches to foot movement mode, the foot end movement is limited to rotational movement characteristics, which further increases the difficulty of collecting effective signals by a single sensor and affects the accuracy and stability of the touch signal. To address these challenges, the sensor unit 201 design must be optimized for the ability to acquire foot end contact signals to ensure the stability and reliability of the system in complex environments.

[0092] Based on the above needs, the application proposes an improved sensor arrangement scheme, referring to Figure 4 The sensor units 201 are distributed in a circumferential array along the solid tire hub, with one sensor unit 201 inserted every 30°, forming a high-density ring array. This design not only increases the number of sensor arrangements but also significantly improves the spatial coverage and sensitivity of the contact signal. Through the coordinated work of the sensor array, even in complex terrain conditions or frequent switching between wheel-legged working modes, the signal acquisition system can still obtain stable and accurate touch signal responses.

[0093] The principle and benefits of inserting a sensor unit 201 every 30° mainly include the following aspects:

[0094] Firstly, the 30° setting angle can ensure the uniform distribution of the sensor unit 201, ensuring that the tire can be monitored evenly in the entire contact area. Since the tire contact surface is a continuous area, by distributing the sensors at equal angles along the rotation axis of the tire, comprehensive monitoring of the tire contact state can be achieved, avoiding the situation where some areas cannot be monitored due to too loose sensor distribution.

[0095] Secondly, the distribution angle of 30° can effectively balance the relationship between the number of sensors and cost, system complexity. A smaller angle (e.g. 15°) would result in a significant increase in the number of sensors, which not only increases the manufacturing and maintenance costs of the system, but also may introduce signal redundancy, increasing the complexity of data processing. While a larger angle (e.g. 45° or more) may result in a too sparse distribution of sensors, reducing the spatial coverage and sensitivity of the contact signal, thereby affecting the stability and accuracy of the system in complex terrain. Therefore, the 30° angle is a compromise that can ensure high sensitivity without resulting in too many sensors, maintaining the reasonableness and economy of the system.

[0096] Furthermore, the 30° angle allows the sensors to work more efficiently in different terrain conditions. With this spacing, the sensor array can better cover the contact area of the tire, ensuring that even in complex terrain or frequent switching of movement modes, the sensors can still stably and accurately collect the contact signal. The signals of different sensors can complement each other, and when the sensor position changes slightly due to tire deformation, other sensors can still maintain strong signal response, ensuring stable system operation.

[0097] In summary, using a 30° angle as the setting parameter for sensor arrangement is to effectively control the number of sensors, cost and data processing complexity while ensuring high coverage and sensitivity of the contact signal, ultimately achieving the goal of optimizing sensor arrangement, improving system reliability and precision.

[0098] Further, the present application makes full use of the advantages of array sensor distribution, enabling the wheel-foot robot to achieve multi-directional and multi-modal contact sensing capability. This improved scheme not only improves the adaptability of the robot to complex terrain, but also provides important support for its intelligent control and motion planning in unknown environments.

[0099] Further, referring to Figure 5 Each sensor unit 201 is composed of a polytetrafluoroethylene hose 501 and a resistive strain gauge 502. The polytetrafluoroethylene hose 501 material is excellent in bending elasticity, which not only responds sensitively to external pressure changes, but also effectively transmits stress information to the internal strain gauge, ensuring the detection sensitivity of the sensor. The resistive strain gauge 502 includes a resin substrate 5022, a metal foil 5021 and a resin cover sheet 5023, and is placed in the central region of the polytetrafluoroethylene hose 501. In this way, when the tire is under stress, the force is transmitted through the hose to the resistive strain gauge 502, causing it to deform and change the resistance value.

[0100] To ensure the stability and reliability of the resistance strain gauge 502, the resistance strain gauge 502 is fixed and encapsulated in the polytetrafluoroethylene hose 501 by a two-component acrylic adhesive. The two-component acrylic adhesive has high bonding strength and good curing performance, which can firmly fix the strain gauge after rapid curing, while ensuring its stability in actual application. After the adhesive is injected into the hose, it will cure and encapsulate the resistance strain gauge 502, allowing it to work effectively for a long time without being affected by the external environment.

[0101] The specific structure of the resistance strain gauge 502 is that the metal foil 5021 is arranged directly above the resin substrate 5022, and the resin cover sheet 5023 covers the metal foil 5021 directly above. The strain-sensitive area is formed between the metal foil 5021 and the resin substrate 5022, which will deform slightly when external force acts on the tire, causing a change in resistance. Through this design, the resistance strain gauge 502 can accurately sense the deformation of the tire and convert the deformation into a measurable electrical signal. The resistance strain gauge 502 is connected to the acquisition module 202 through the strain gauge lead 503, so as to transmit the electrical signal to the acquisition module 202, and then to the lower computer 203, and finally the lower computer 203 transmits the signal to the server 204.

[0102] This embedded sensor unit 201 design can provide real-time monitoring of tire stress and deformation inside the tire, helping the wheel-legged robot accurately perceive the touch state, thereby providing the necessary ground adaptation information for the robot and supporting its efficient movement in complex environments.

[0103] Further, a plurality of sensor units 201 are installed in the solid tire of the wheel-legged robot, each sensor unit 201 comprising a resistance strain gauge 502 and a polytetrafluoroethylene hose 501, wherein the resistance strain gauge 502 senses the influence of external pressure on the tire through resistance change. When external pressure acts on the tire, the pressure is transmitted through the solid tire to the polytetrafluoroethylene hose 501, and further to the resistance strain gauge 502. Under the working principle of the resistance strain gauge 502, the strain caused by external pressure causes a change in resistance, and the resistance change is linearly related to the applied external pressure.

[0104] The resistance change value of each sensor unit 201 will be transmitted to the server 204 through the strain gauge lead 503. The multiple resistance data received by the server 204 represent the response of the strain gauge at different sensor positions, and then the strain value of the tire is calculated. Next, according to the obtained resistance change value, the strain value can be calculated through the strain sensitivity coefficient and the resistance change rate formula. The specific calculation formula is:

[0105]

[0106] wherein, ΔR is the resistance change value, R is the initial resistance value, K s is the strain sensitivity, and ε is the strain value. The strain value directly reflects the degree of deformation of the tire due to external pressure, and this data is crucial for the wheel-legged robot to determine the tire-ground contact state and assess ground information.

[0107] S120, obtaining acceleration data of the touchdown point when the tire of the wheel-legged robot is in the touchdown state.

[0108] Near the touchdown point of the tire, six-axis acceleration sensors are arranged. These acceleration sensors function to monitor the acceleration information of the touchdown point in real time, including acceleration along the x-axis and y-axis and normal acceleration. The six-axis acceleration sensor can sense acceleration changes in different directions, accurately recording the interaction force between the tire and the ground when touching the ground.

[0109] The arrangement of the sensors should take into account the motion characteristics of the tire and possible force points, ensuring that the sensors can cover the entire touchdown area, especially in the edge area where the tire contacts the ground, in order to capture acceleration changes in all directions. According to specific application requirements, sensors can be evenly distributed or concentrated in key areas to ensure sufficient information is collected.

[0110] S130, according to the established mapping relationship between the input vector and the output vector, the contact force data of the wheel-legged robot and the spatial position of the tire touchdown point are obtained through the tire strain data and the acceleration data.

[0111] In one possible implementation, before obtaining the contact force data of the wheel-legged robot and the spatial position of the tire touchdown point according to the established mapping relationship between the input vector and the output vector through the tire strain data and the acceleration data, the method further comprises: constructing an input vector, wherein the input vector is composed of acceleration data and multiple sets of tire strain data; constructing an output vector, wherein the output vector includes contact force data and spatial position; training a Gaussian process regression model through training data, wherein the training data includes training acceleration data and multiple sets of training tire strain data as input, and training contact force data and training spatial position as output; training the Gaussian process regression model through the training data, and using the Gaussian process regression model to model the mapping relationship between the input vector and the output vector.

[0112] Specifically, due to the non-linear and complex relationship between the readings and the applied force, direct analytical modeling is not possible. Therefore, a Gaussian Process Regression model is used for modeling and analysis. First, the input vector needs to include acceleration data and multiple sets of tire strain data. The acceleration data provides acceleration information when the tire touches the ground, thereby obtaining shear forces along the x and y axes and the normal reaction force, while the tire strain data reflects the deformation of the tire due to external pressure. Each set of input data represents a specific tire state. By combining these data into an input vector, the Gaussian Process Regression model can be provided with multi-dimensional information, thereby helping to accurately estimate the output value.

[0113] The output vector includes contact force data, i.e., shear forces and normal forces on the x, y, and z axes, as well as the spatial position of the tire contact point. The contact force data reflects the interaction force between the tire and the ground, while the spatial position represents the actual position of the tire contact point with the ground. By taking these output data as the target, the model can be trained to predict the corresponding output from the input data.

[0114] The training data set is the basis for model training. The training data includes training acceleration data and training tire strain data as input, which represent the tire ground contact under different ground conditions. At the same time, the output data set includes training contact force data and training spatial position, which represent the force between the tire and the ground and the spatial coordinates of the contact point under the corresponding input data, respectively. These training data should be representative and cover different terrains, different loads, and different motion modes.

[0115] The Gaussian Process Regression model is trained using the training data. Gaussian Process Regression is a non-parametric Bayesian model that can learn the complex relationship between input vectors and output vectors through built-in Gaussian kernel functions. During training, the model learns the mapping relationship between input data and output data and optimizes the model parameters by minimizing errors. After training, the model can predict the output vector contact force and spatial position based on the input vector acceleration data and strain data.

[0116] The mapping relationship between the input vector and the output vector is modeled using the trained Gaussian Process Regression model using the training data. The mapping relationship is represented as:

[0117] Y j =k * (K-σ n Ι) -1 y j

[0118] where Y j is the output vector, Y j =[F xF y ,F z ] T ,y j is the input vector, σ n is the standard deviation, and the input vector is denoted as x = [s1, s2] T , k * is a variable independent of the input, and K is the covariance matrix.

[0119] In the above formula, k * is expressed as follows:

[0120] k * = [k(x * , x1) k(x * , x2)... k(x * , x n )

[0121] where k(x * , x i ) is the covariance value between the input vector x * and the training data point x i , indicating the correlation between the input data.

[0122] The expression of the covariance matrix K calculated from the covariance between all training data points is as follows:

[0123]

[0124] where the square exponential covariance function is used as the kernel k when calculating, and its expression is as follows:

[0125]

[0126] where σ f represents the signal variance, and l represents the length scale. The hyperparameters σ n , σ f , and l are manually adjusted based on the evaluation of the regression results on the validation dataset.

[0127] The strain data and acceleration data of the tire are integrated into an input vector, which is input into a trained Gaussian process regression model. The trained model can predict the contact force data of the tire when in contact with the ground, including the shear forces in the x and y axes and the normal reaction force, based on the learned mapping relationship of the input vector, in addition to predicting the spatial position of the tire contact point, i.e., the three-dimensional coordinates of the contact point. Specifically, the input data is processed through the calculation of the covariance matrix and the kernel function, and the inference process of the Gaussian process regression model to obtain the most relevant contact force and spatial position output. Through this process, the wheel-legged robot can obtain the mechanical characteristics and contact point position of the tire in contact with the ground in real time, thereby achieving terrain adaptation and stable motion control.

[0128] With reference to Figure 6 , there is a significant difference in the behavior of the strain signal between hard and soft road surfaces. With reference to Figure 6 part a, on a hard road surface, the tire strain signal exhibits more drastic changes and higher peaks, indicating concentrated contact force and large instantaneous deformation. This behavior reflects rapid strain changes associated with high-rigidity contact. With reference to Figure 6 part b, on the contrary, on a soft road surface, the strain signal exhibits smooth changes with lower peaks, highlighting the dispersion of contact force and the cushioning effect of the flexible surface. This shows that this solid wheel-foot end detection method is sensitive to changes in road hardness and can effectively reflect the actual strain.

[0129] S140, based on the contact force data and the spatial position, analyzes the ground contact information of the position where the wheel-legged robot is located.

[0130] In one possible implementation, based on the contact force data and the spatial position, the ground contact information of the position where the wheel-legged robot is located is analyzed, specifically including: based on the contact force data, calculating the contact area of the contact area of the wheel-legged robot, the contact area being the area where the tire of the wheel-legged robot contacts the ground; according to the contact force data and the contact area, calculating the change rate of the contact force in the contact area; based on the change of the spatial position, calculating the ground deformation data; according to the change rate and the ground deformation data, calculating the ground hardness; judging the size relationship between the ground hardness and different preset thresholds to determine the ground hardness category of the contact area.

[0131] Specifically, first, according to the normal reaction force and the contact pressure when the tire is in contact with the ground, the contact area can be calculated. The formula for calculating the contact area is:

[0132]

[0133] where F z is the normal reaction force contained in the contact force data, representing the pressure in the vertical direction of the tire; σ contactFor the contact pressure, it can be calculated from the strain sensor data.

[0134] By observing the change of the contact force data, the rate of change of the contact force in the contact area can be calculated. Specifically, based on the change of the normal reaction force, the rate of change can be calculated:

[0135]

[0136] The rate of change here reflects the fluctuation of the contact force between the tire and the ground. When the contact force changes greatly, it usually indicates that the pressure in the contact area is concentrated, and the ground may be hard; when the change is small, it means that the contact force is relatively uniform, and the ground is soft.

[0137] According to the change of the spatial position of the tire when it contacts the ground, especially the vertical deformation, the ground deformation can be calculated. This step needs to be based on the change of the height of the tire at the contact point to estimate:

[0138] Δz = z(new) - z(initial)

[0139] Where Δz is the ground deformation, i.e. the vertical displacement of the tire contact area, indicating the deformation of the ground due to the pressure exerted by the tire. A larger ground deformation indicates a softer ground, and a smaller ground deformation indicates a harder ground.

[0140] By integrating the contact force data, the rate of change of the contact area, and the ground deformation data, the ground hardness can be calculated, specifically by the following formula:

[0141]

[0142] Where H is the ground hardness, F z is the normal reaction force contained in the contact force data, Δz is the ground deformation, A is the contact area, and ΔF z is the change in contact force.

[0143] This formula calculates the ground hardness by integrating the contact force data, the rate of change of the contact area, and the ground deformation data. First, the basic part of the formula The ground hardness is calculated by the normal reaction force, the contact area A, and the ground deformation, which reflects the force exerted per unit contact area and per unit vertical deformation. Next, by adding the correction term Considering the change in contact force, this term reflects the rate of change of the contact force in the contact area, further correcting the hardness calculation, so that this formula can more accurately represent the rigidity and force distribution of the ground. When the contact force changes greatly, the pressure in the contact area is concentrated, and the hardness will increase accordingly; while the contact force changes less, it indicates a softer ground.

[0144] After the ground hardness is calculated, the calculated hardness value needs to be compared with a preset threshold value next. According to the size of the hardness value, it can be judged which category the ground belongs to. If the ground hardness is greater than the set hardness threshold value, the ground is considered as a hard road surface; if the ground hardness is less than or equal to the hardness threshold value, the ground is considered as a soft road surface.

[0145] The embodiment also discloses a foot end ground contact information detection device based on a wheel-legged robot, the device being a server 204, referring to Figure 7 , comprising an acquisition module 701, a processing module 702 and an output module 703, wherein:

[0146] The acquisition module 701 is configured to acquire tire strain data of the wheel-legged robot through a signal acquisition system arranged on the wheel-legged robot.

[0147] The acquisition module 701 is configured to acquire acceleration data of a contact point when the tire of the wheel-legged robot is in a ground contact state.

[0148] The processing module 702 is configured to acquire contact force data of the wheel-legged robot and a spatial position of the contact point of the tire according to a mapping relationship between the input vector and the output vector and through the tire strain data and the acceleration data.

[0149] The output module 703 is configured to analyze ground contact information of a position where the wheel-legged robot is located based on the contact force data and the spatial position.

[0150] In a possible implementation, the processing module 702 is configured to calculate a contact area of a contact region of the wheel-legged robot based on the contact force data, the contact region being a region where the tire of the wheel-legged robot contacts the ground.

[0151] The processing module 702 is configured to calculate a change rate of the contact force in the contact region according to the contact force data and the contact area.

[0152] The processing module 702 is configured to calculate a ground deformation variable based on a change of the spatial position.

[0153] The processing module 702 is configured to calculate ground hardness according to the change rate and the ground deformation data.

[0154] The processing module 702 is configured to judge a size relationship between the ground hardness and different preset threshold values and determine a ground hardness category of the contact region.

[0155] In a possible implementation, the processing module 702 is configured to calculate the ground hardness according to the change rate and the ground deformation variable, specifically through the following formula:

[0156]

[0157] where H is the ground hardness, F z is the normal reaction force contained in the contact force data, Δz is the ground deformation, A is the contact area, ΔF z is the change of the contact force.

[0158] In a possible implementation, the processing module 702 is configured to construct an input vector, where the input vector is composed of the acceleration data and a plurality of sets of tire strain data.

[0159] The processing module 702 is configured to construct an output vector, where the output vector includes the contact force data and the spatial position.

[0160] The processing module 702 is configured to train the Gaussian process regression model by using training data, where the training data includes training acceleration data and a plurality of sets of training tire strain data as input, and training contact force data and training spatial position as output.

[0161] The output module 703 is configured to train the Gaussian process regression model by using the training data, and model the mapping relationship between the input vector and the output vector by using the Gaussian process regression model, where the mapping relationship is represented as follows:

[0162] Y j = k * (K-σ n I) -1 y j

[0163] where Y j is the output vector, Y j = [F x , F y , F z ] T , y j is the input vector, σ n is the standard deviation, the input vector is represented as x = [s1, s2] T , k * is a variable irrelevant to the input, and K is the covariance matrix.

[0164] In a possible implementation, the output module 703 is configured to input the tire strain data and the acceleration data to the trained Gaussian process regression model, and obtain the contact force data and the spatial position of the tire contact point output by the trained Gaussian process regression model, where the contact force data includes the x-axis direction shear force, the y-axis direction shear force, and the normal reaction force, and the spatial position includes the spatial coordinates of the contact point.

[0165] In a possible implementation, the acquisition module 701 is configured to acquire a plurality of resistance data transmitted by the sensor unit 201, wherein when external pressure acts on the solid tire, the external pressure is transmitted to the polytetrafluoroethylene hose 501 through the solid tire, and is transmitted to the resistance strain gauge 502, and the resistance value of the resistance strain gauge 502 changes linearly and proportionally with the external pressure.

[0166] The processing module 702 is configured to calculate a resistance change value according to the plurality of resistance data.

[0167] The processing module 702 is configured to calculate a strain value according to the resistance change value to obtain tire strain data, specifically by using the following formula:

[0168]

[0169] wherein ΔR is the resistance change value, R is an initial resistance value, K s is a strain sensitivity, and ε is the strain value.

[0170] It should be noted that the device provided in the above embodiment is only used as an example to divide the above functional modules to achieve its functions, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the above described functions. In addition, the device and method embodiments provided in the above embodiment belong to the same concept, and the specific implementation process is described in the method embodiment, which will not be repeated here.

[0171] The embodiment also discloses an electronic device, which refers to Figure 8 The electronic device can include at least one processor 801, at least one communication bus 802, a user interface 803, a network interface 804, and at least one memory 805.

[0172] The communication bus 802 is configured to realize the connection and communication between the components.

[0173] The user interface 803 can include a display screen (Display) and a camera (Camera), and the optional user interface 803 can further include a standard wired interface and a wireless interface.

[0174] The network interface 804 can optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

[0175] The processor 801 can include one or more processing cores. The processor 801 connects various parts within the server through various interfaces and lines, performs various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 805, and calling data stored in the memory 805. Alternatively, the processor 801 can be implemented in at least one of a hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 801 can integrate a combination of one or more of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes operating systems, user interfaces, and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; and the modem is used for processing wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 801, but can be realized by a separate chip.

[0176] The memory 805 can include a random access memory (RAM) and a read-only memory (ROM). Optionally, the memory includes a non-transitory computer-readable storage medium. The memory 805 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 805 can include a program storage area and a data storage area, wherein the program storage area can store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area can store data involved in the above-mentioned various method embodiments, etc. The memory 805 can also be at least one storage device located away from the aforementioned processor 801. As a kind of computer storage medium, the memory 805 can include an operating system, a network communication module, a user interface 803 module and an application program based on the foot end information detection method of the wheel-legged robot.

[0177] In Figure 8The electronic device shown, the user interface 803 is mainly used for providing the interface for the user to input, obtaining the data input by the user; and the processor 801 can be used to call the application program stored in the memory 805 and storing a foot end ground information detection method based on a wheel-legged robot, when executed by one or more processors 801, so that the electronic device executes the method of one or more of the above embodiments.

[0178] It should be noted that, for the foregoing method embodiments, in order to simply describe, they are all described as a series of action combinations, but those skilled in the art should know that the present application is not limited to the order of the actions described, because according to the present application, certain steps can be performed in other order or at the same time. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0179] In the above embodiments, the description of each embodiment has its own emphasis, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0180] In several embodiments provided by the present application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of units is only a logical function division. There can be another division manner in actual implementation, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between the units shown or discussed can be indirect coupling or communication connection through some services interfaces, devices or units, and can be electrical or other forms.

[0181] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, that is, they can be located in one place, or can be distributed on a plurality of network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the embodiment scheme.

[0182] In addition, each functional unit in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of software functional unit.

[0183] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable memory. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a memory 805 and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the embodiments of the present application. The aforementioned memory 805 includes: a U disk, a mobile hard disk, a magnetic or optical disk and various program code storage media.

[0184] The present application also discloses a computer readable storage medium, which stores instructions. When executed by one or more processors 801, the instructions cause an electronic device to perform the method of one or more of the above embodiments.

[0185] The above is only exemplary embodiments of the present disclosure, which cannot limit the scope of the present disclosure. That is, any equivalent changes and modifications made in accordance with the teachings of the present disclosure are still within the scope of the present disclosure. Other embodiments of the present disclosure will be readily apparent to those skilled in the art upon considering the specification and practice of the present disclosure. The present application is intended to cover any variations, uses or adaptive changes of the present disclosure that follow the general principles of the present disclosure and include common knowledge or conventional techniques in the art that are not described in the present disclosure. The specification and examples are only considered as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A method for detecting foot contact information of a wheel-legged robot, characterized in that: The method is applied to a server (204), and comprises: Acquiring tire strain data of the wheeled-leg robot through a signal acquisition system provided on the wheeled-leg robot; Acquiring acceleration data of a contact point when a tire of the wheeled robot is in a contact state; According to the established mapping relationship between the input vector and the output vector, the contact force data of the wheel-legged robot and the spatial position of the tire contact point are obtained through the tire strain data and the acceleration data; Analyzing the ground contact information of the wheeled robot based on the contact force data and the spatial position; The analyzing the ground contact information of the wheeled robot based on the contact force data and the spatial position specifically includes: Calculating a contact area of ​​a contact region of the wheeled robot based on the contact force data, wherein the contact region is an area where a tire of the wheeled robot contacts the ground; Calculating a rate of change of the contact force in the contact area based on the contact force data and the contact area; Calculating the ground deformation based on the change in the spatial position; Calculating ground hardness based on the change rate and the ground deformation data; Determine the relationship between ground hardness and different preset thresholds to determine the ground hardness category of the contact area; The ground hardness is calculated based on the change rate and the ground deformation, specifically using the following formula: ； Wherein, H is the ground hardness, F z is the normal reaction force contained in the contact force data, Δz is the ground deformation, A is the contact area, ΔF z is the change of contact force; The signal acquisition system comprises a plurality of sensor units (201), wherein: A plurality of sensor units (201) are embedded in the solid tire of the wheel-legged robot and are distributed in a circular manner with equal angles inside the tire, with the tire rotation axis as the center. Each sensor unit (201) comprises a polytetrafluoroethylene hose (501) and a resistance strain gauge (502). The resistance strain gauge (502) comprises a resin covering sheet (5023), a metal foil (5021), and a resin base (5022). The resistance strain gauge (502) is arranged in the central area of ​​the polytetrafluoroethylene hose (501), and a two-component acrylic adhesive is injected into the polytetrafluoroethylene hose (501) to cure and encapsulate the resistance strain gauge (502); The metal foil (5021) is arranged directly above the resin base (5022), and the resin covering sheet (5023) covers directly above the metal foil (5021).

2. The method for detecting foot contact information of a wheel-legged robot according to claim 1, characterized in that: Before obtaining the contact force data of the wheel-legged robot and the spatial position of the tire contact point through the tire strain data and the acceleration data according to the established mapping relationship between the input vector and the output vector, the method further includes: constructing the input vector, wherein the input vector is composed of the acceleration data and multiple sets of tire strain data; constructing an output vector, wherein the output vector includes the contact force data and the spatial position; Training a Gaussian process regression model using training data, wherein the training data includes training acceleration data and multiple sets of training tire strain data as inputs, and training contact force data and training spatial positions as outputs; The Gaussian process regression model is trained using the training data, and a mapping relationship between the input vector and the output vector is modeled using the Gaussian process regression model, wherein the mapping relationship is expressed as follows: ； Among them, Y j is the output vector, ,y j is the input vector, σ n is the standard deviation, the input vector is represented as , is a variable independent of the input, and K is the covariance matrix.

3. The method for detecting foot contact information of a wheel-legged robot according to claim 1, characterized in that: The step of obtaining the contact force data of the wheel-legged robot and the spatial position of the tire contact point by using the tire strain data and the acceleration data based on the established mapping relationship between the input vector and the output vector specifically includes: The tire strain data and the acceleration data are input into a trained Gaussian process regression model to obtain contact force data and the spatial position of the tire contact point output by the trained Gaussian process regression model, wherein the contact force data includes shear force in the x-axis direction, shear force in the y-axis direction, and normal reaction force, and the spatial position includes the spatial coordinates of the contact point.

4. The method for detecting foot contact information of a wheel-legged robot according to claim 1, wherein: The sensor unit (201) further includes a strain gauge wire (503), and the server (204) and the sensor unit (201) are communicatively connected based on the strain gauge wire (503); The step of obtaining tire strain data of the wheeled robot by a signal acquisition system provided on the wheeled robot specifically includes: Acquiring a plurality of resistance data transmitted by the sensor unit (201), wherein when external pressure acts on the solid tire, the resistance data is transmitted through the solid tire to the polytetrafluoroethylene hose (501) and then to the resistance strain gauge (502), and the resistance value of the resistance strain gauge (502) changes linearly with the external pressure; Calculating a resistance change value based on the plurality of resistance data; The strain value is calculated according to the resistance change value to obtain the tire strain data, which is specifically calculated using the following formula: ； in, is the resistance change value, R is the initial resistance value, K s is the strain sensitivity, and ε is the strain value.

5. An electronic device, characterized in that: The electronic device comprises a processor (801), a communication bus (802), a user interface (803), a network interface (804) and a memory (805), wherein the memory (805) is used to store instructions, the user interface (803) and the network interface (804) are both used to communicate with other devices, the communication bus (802) is used to realize connection and communication between components in the electronic device, and the processor (801) is used to execute the instructions stored in the memory (805) so that the electronic device executes the method according to any one of claims 1 to 4.

6. A computer-readable storage medium, characterized in that The computer-readable storage medium stores instructions, and when the instructions are executed, the method according to any one of claims 1 to 4 is executed.

Images