Wheel-legged vehicle state estimation method considering slippage and related device

Abstract

The invention discloses a wheel-legged vehicle state estimation method considering slippage and a related device, and relates to the technical field of vehicle control, and the method comprises the steps: determining the contact state of each wheel based on the sensor data of each wheel motor at the previous moment; multi-mode slip detection is carried out on the wheels with the contact state being in contact with the ground, the slip rate of each wheel is obtained, and the average slip rate is calculated according to the slip rates of all the wheels; determining an adaptive weight based on the average slip rate and a wheel confidence index; based on the self-adaptive weight, the first wheel speed at the previous moment and the second vehicle body speed at the previous moment, the fusion speed at the current moment is obtained through calculation; and inputting the fusion speed at the current moment and the actual state observation value at the previous moment into a Kalman filter, and calculating to obtain the vehicle state at the current moment. According to the invention, the state estimation precision and the system robustness of the wheel-legged vehicle under the complex terrain are obviously improved.

Description

Technical Field

[0001] This application relates to the field of vehicle control technology, and in particular to a method and related apparatus for estimating the state of a wheel-legged vehicle that takes slippage into account. Background Technology

[0002] With the continuous development of ground mobile robot technology, wheel-legged vehicles, as a type of platform that integrates the advantages of wheeled and legged structures, are gradually becoming important carriers for performing tasks in complex terrain environments. These platforms combine the efficiency of wheeled motion with the adaptability of legged motion, enabling high-speed movement on flat surfaces and obstacle-heavy areas where they can overcome obstacles and adjust their posture via leg joints, demonstrating superior terrain adaptability and maneuverability. To achieve stable operation in complex terrain, these systems typically employ a multi-actuator independent drive structure, with each wheel-leg unit possessing independent motion control capabilities, thus supporting highly dynamic and multimodal motion modes. However, during highly dynamic motion, the vehicle's state (including position, velocity, attitude, and contact state) exhibits significant nonlinear and time-varying characteristics, placing high demands on the state estimation system for high accuracy, high real-time performance, and strong robustness.

[0003] Currently, state estimation methods for wheel-legged vehicles still have several shortcomings. First, existing methods generally decouple wheel motion from leg motion, failing to fully consider their dynamic coupling characteristics in actual operation. Especially during high-speed movement or drastic terrain changes, the contact forces, constraints, and kinematic relationships between the wheels and legs influence each other. Simple decoupling can easily lead to a state model that does not match the actual dynamics, thus reducing estimation accuracy. Therefore, there is an urgent need to construct a unified state estimation framework that can integrate leg-based and wheel-based odometers to improve estimation performance under complex terrain and high-speed conditions.

[0004] Secondly, most existing state estimation methods rely on the legged platform frame to introduce wheel motion information. While this has certain advantages on smooth roads, it generally fails to model and compensate for wheel slippage behavior. In unstructured or low-adhesion terrain, wheel slippage occurs frequently. Ignoring its impact will lead to distortion in wheel odometer observations, affecting the accuracy and stability of multi-sensor fusion, and even causing cumulative errors or divergence in state estimation. Therefore, achieving real-time perception and adaptive processing of wheel slippage has become a key issue in improving the robustness of state estimation for legged vehicles.

[0005] Furthermore, existing state estimation methods for wheeled vehicles are mostly designed for Ackerman steering vehicles with steering mechanisms, and their estimation accuracy is low for wheeled platforms using speed differential steering. Wheel-leg platforms typically achieve steering through outward-flaring joints in their legs, and their motion mechanism differs from speed differential steering, making existing methods difficult to apply directly and further limiting the effective application of state estimation systems in wheel-leg vehicles. Summary of the Invention

[0006] The purpose of this application is to provide a state estimation method and related apparatus for wheel-leg vehicles that takes slippage into account, which can significantly improve the state estimation accuracy and system robustness of wheel-leg vehicles in complex terrain.

[0007] To achieve the above objectives, this application provides the following solution: In a first aspect, this application provides a method for estimating the state of a wheel-legged vehicle that considers slippage, including: Acquire inertial measurement unit data and sensor data from multiple wheel motors at the previous moment; the sensor data includes encoder data and torque data; Based on the sensor data of each wheel motor in the previous moment, determine the contact state of each wheel; the contact state includes contact with the ground and non-contact with the ground. Multimodal slip detection is performed on wheels in contact with the ground to obtain the slip ratio of each wheel, and the average slip ratio is calculated based on the slip ratios of all wheels; the multimodal slip detection includes wheel lock-up detection, kinematic consistency detection, and contact force verification. Based on the average slip ratio and the preset wheel confidence index, adaptive weights are determined. Based on adaptive weights and the first wheel speed and the second vehicle speed obtained from the previous moment, the fused speed at the current moment is calculated; the first wheel speed is calculated using the wheel kinematics model, and the second vehicle speed is determined based on inertial measurement unit data. The current fusion velocity and the actual observed state from the previous moment are input into the Kalman filter to calculate the vehicle state at the current moment; the vehicle state includes vehicle speed and vehicle position.

[0008] Optionally, the encoder data includes wheel angular velocity and wheel position, and the inertial measurement unit data includes vehicle body linear velocity and vehicle body angular velocity; for wheels in contact with the ground, multimodal slip detection is performed to obtain the slip ratio of each wheel, specifically including: The system determines whether the angular velocity of the target wheel is lower than a preset angular velocity threshold and whether the torque data of the target wheel exceeds a preset torque threshold, and obtains a first determination result; the target wheel is any wheel in contact with the ground in any contact state. If the first judgment result is yes, then it is determined that the target wheel has locked up, and the target wheel is identified as the locked wheel, and the slip ratio of the locked wheel is set to the preset maximum slip ratio. If the first judgment result is negative, the theoretical wheel linear velocity of the target wheel is calculated based on the vehicle body linear velocity, vehicle body angular velocity and the wheel position of the target wheel, and the actual wheel linear velocity of the target wheel is calculated based on the wheel angular velocity of the target wheel. The second judgment result is obtained by judging whether the deviation between the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel exceeds a preset dynamic threshold. If the second judgment result is yes, then the slip ratio of the target wheel is calculated based on the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel. If the second judgment result is negative, then based on the target wheel's wheel angular velocity, theoretical wheel linear velocity, and torque data, the longitudinal traction force of the target wheel is calculated, and it is determined whether the longitudinal traction force is greater than the preset traction force threshold to obtain the third judgment result. If the third judgment result is yes, then the slip ratio of the target wheel is set to the preset maximum slip ratio; If the third judgment result is negative, the slip ratio of the target wheel is calculated based on the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel.

[0009] Optionally, the slip ratio of the target wheel is calculated based on the theoretical and actual linear velocities of the target wheel, specifically: ; in, Slip ratio; For the first i The absolute value of the deviation between the theoretical linear velocity and the actual linear velocity of each wheel; For the first i The theoretical linear velocity of each wheel; For the first i The actual linear velocity of each wheel; This is the minimum speed threshold.

[0010] Optionally, adaptive weights are determined based on the average slip ratio and a preset wheel confidence index, specifically as follows: ; in, It is an adaptive weight; It is a confidence index for wheels. , It is the linear acceleration of the car body. These are empirical normalization parameters; It is the average slip ratio.

[0011] Optionally, based on adaptive weights and the obtained first wheel speed and second vehicle speed from the previous moment, the fused speed at the current moment is calculated, specifically as follows: ; in, It's the fusion speed; It is an adaptive weight; It is the speed of the first wheel; That is the second vehicle speed.

[0012] Optionally, the fusion speed at the current moment and the actual observed state from the previous moment are input into the Kalman filter to calculate the vehicle state at the current moment, specifically including: Based on the fusion speed at the current moment and the actual observation values ​​of the state at the previous moment, the observation vector of the Kalman filter at the current moment is constructed. Based on the vehicle state at the previous moment and the preset discrete-time state equation, the predicted state value at the current moment is calculated. The Kalman gain at the current time is calculated based on the preset observation matrix and observation noise covariance matrix at the current time, as well as the predicted state covariance matrix at the current time. The predicted state covariance matrix at the current time is determined based on the updated predicted state covariance matrix at the previous time. Based on the Kalman gain and the observation vector at the current time, the state prediction value at the current time is corrected to obtain the updated state prediction value at the current time, and the updated state prediction value at the current time is used as the vehicle state at the current time.

[0013] Optionally, the Kalman gain at the current time is calculated based on the observation matrix at the current moment, the obtained predicted state covariance matrix, and the preset observation noise covariance matrix, specifically as follows: ; in, This represents the Kalman gain at the current time k; This represents the predicted state covariance matrix at the current time k; This represents the observation matrix at the current time k; This represents the transpose of the observation matrix at the current time k; This represents the observation noise covariance matrix at the current time k.

[0014] Optionally, based on the Kalman gain and observation vector at the current time, the predicted state value at the current time is corrected to obtain an updated predicted state value at the current time, specifically as follows: ; in, This represents the updated state prediction value for the current time k. This represents the Kalman gain at the current time k; This represents the predicted state value at the current time k. This represents the observation matrix at the current time k; This represents the observation vector at the current time k.

[0015] In a second aspect, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the wheel-leg vehicle state estimation method considering slippage as described above.

[0016] Thirdly, this application provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the wheel-leg vehicle state estimation method considering slippage described above.

[0017] Fourthly, this application provides a computer program product, including a computer program that, when executed by a processor, implements the wheel-leg vehicle state estimation method considering slippage described above.

[0018] According to the specific embodiments provided in this application, this application has the following technical effects: This application provides a method and related apparatus for state estimation of wheel-leg vehicles considering slip. By acquiring data from an inertial measurement unit and sensor data from multiple wheel motors, including encoder data and torque data, it solves the problem of single sensor information sources and difficulty in accurately reflecting the dynamic characteristics of the vehicle body in traditional wheel-leg vehicle state estimation. It ensures the comprehensiveness and accuracy of sensor data, realizes the foundation for multi-source information fusion of vehicles under complex terrain and high-speed conditions, and provides reliable basic data for subsequent slip ratio calculation and vehicle state estimation.

[0019] By using a multimodal slip detection method based on sensor data to calculate the slip ratio of each contacting wheel, this method solves the problems of difficulty in real-time perception of wheel slip, lack of slip compensation leading to observation distortion and accumulation of estimation errors in existing methods. It achieves accurate identification and dynamic modeling of wheel slip state. This multimodal slip detection method combines wheel lock-up detection, kinematic consistency detection and contact force verification, which effectively improves the accuracy and robustness of state estimation.

[0020] By using adaptive weights to fuse the velocity of the wheel kinematic model at the previous moment with the velocity calculated by the inertial measurement unit, the fused velocity at the current moment is obtained. This solves the dynamic inconsistency problem caused by independent estimation of wheel and inertial information, realizes optimal fusion of multi-source velocity information based on slip perception, and improves the continuity and accuracy of vehicle motion estimation.

[0021] By inputting the fusion velocity at the current moment and the actual state observation value at the previous moment into the Kalman filter for state estimation, the problem of insufficient dynamic consistency caused by the decoupling of wheeled and legged motion in existing methods is solved. This achieves unified modeling and fusion estimation of wheeled and legged motion characteristics, thereby significantly improving the state estimation accuracy and system robustness of wheeled and legged vehicles in complex terrain. Attached Figure Description

[0022] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0023] Figure 1 This is an application environment diagram of a wheel-leg vehicle state estimation method considering slippage in one embodiment of this application; Figure 2 A flowchart illustrating a method for estimating the state of a wheel-legged vehicle considering slippage, provided in an embodiment of this application; Figure 3 A schematic diagram of a framework for estimating the state of a wheel-leg type wheel that takes into account tire slippage, provided for one embodiment of the application; Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0024] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0025] Wheel-leg vehicles combine the advantages of wheeled and legged structures, demonstrating excellent terrain adaptability through coordinated wheel-leg movement in complex terrains. Compared to traditional platforms, wheel-leg platforms offer significant advantages in motion freedom, obstacle-crossing ability, and environmental interaction flexibility. To achieve stable operation in complex terrains, such platforms typically consist of multiple independent actuators, each wheel-leg unit capable of independent drive, enabling precise control of high-dynamic motion under various terrain conditions. During high-dynamic motion, the vehicle needs to frequently respond to drastic attitude changes and terrain disturbances, resulting in significant nonlinearity and time-varying vehicle states (including position, velocity, attitude, and contact state). Therefore, to ensure the safe and stable operation of the system and achieve high-performance motion planning and control strategies, a high-precision, real-time, and environmentally robust state estimation system is urgently needed. Furthermore, in complex or low-traction environments, the tires of wheel-leg vehicles are highly susceptible to slippage. Tire slippage leads to accumulated odometer estimation errors and control command execution deviations, thereby affecting the stability of the entire system and task execution efficiency. Therefore, accurately identifying and modeling tire slippage behavior during state estimation is crucial for improving the accuracy of state estimation and enhancing the robustness of the control system. Especially in unstructured environments, failure to effectively consider tire slippage can lead to attitude estimation errors and sensor fusion failures, thereby affecting the reliability of path planning and control strategies. Therefore, to improve the accuracy and robustness of state estimation for wheel-legged vehicles under high-speed driving conditions in complex environments, this application proposes a state estimation method for wheel-legged vehicles that considers slippage. This method determines the slippage state of each wheel through three stages: wheel lock-up detection, kinematic consistency detection, and contact force verification. The final slippage rate is achieved through a normalized speed difference index. The outputs of multiple detection mechanisms are fused and combined with dynamic thresholds under different operating conditions to achieve accurate determination of the slippage state. Furthermore, a dynamic fusion algorithm based on real-time slippage state adaptive weights is proposed to achieve robust state estimation of wheel-legged vehicles in complex environments.

[0026] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0027] The wheel-leg vehicle state estimation method considering slippage provided in this application embodiment can be applied to, for example... Figure 1In the application environment shown, terminal 102 communicates with server 104 via a network. A data storage system can store the data that server 104 needs to process. The data storage system can be set up independently, integrated into server 104, or placed in the cloud or on another server. Terminal 102 can send sensor data to server 104. Server 104 determines the contact state of each wheel based on the sensor data of each wheel motor from the previous moment; for wheels in contact with the ground, it performs multimodal slip detection to obtain the slip ratio of each wheel, and calculates the average slip ratio based on the slip ratios of all wheels; based on the average slip ratio and a preset wheel confidence index, it determines adaptive weights; based on the adaptive weights and the acquired first wheel speed and second vehicle speed from the previous moment, it calculates the fusion speed for the current moment; it inputs the current fusion speed and the actual observed state from the previous moment into a Kalman filter to calculate the vehicle state for the current moment. Server 104 can then feed back the obtained vehicle state to terminal 102.

[0028] The terminal 102 can be, but is not limited to, various desktop computers, laptops, and IoT devices. The server 104 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0029] In one exemplary embodiment, such as Figure 2 As shown, a method for estimating the state of a wheel-legged vehicle considering slippage is provided. This method is executed by a computer device, specifically a terminal or server, or both. In this embodiment, the method is applied to... Figure 1 Taking server 104 as an example, the explanation includes the following steps 201 to 206. Wherein: Step 201: Acquire the Inertial Measurement Unit (IMU) data and sensor data from multiple wheel motors at the previous moment; the sensor data includes encoder data and torque data.

[0030] Step 202: Based on the sensor data of each wheel motor in the previous moment, determine the contact state of each wheel; the contact state includes contact with the ground and non-contact with the ground.

[0031] Step 203: Perform multimodal slip detection on the wheels in contact with the ground to obtain the slip ratio of each wheel, and calculate the average slip ratio based on the slip ratios of all wheels; the multimodal slip detection includes wheel lock-up detection, kinematic consistency detection and contact force verification.

[0032] Step 204: Based on the average slip ratio and the preset wheel confidence index, determine the adaptive weights.

[0033] Step 205: Based on the adaptive weights and the first wheel speed and the second vehicle speed obtained at the previous moment, the fused speed at the current moment is calculated; the first wheel speed is calculated using the wheel kinematics model, and the second vehicle speed is determined based on the inertial measurement unit data.

[0034] Step 206: Input the fusion speed at the current moment and the actual observed state value of the previous moment into the Kalman filter to calculate the vehicle state at the current moment; the vehicle state includes vehicle speed and vehicle position.

[0035] By implementing steps 201 to 206 above, and acquiring inertial measurement unit data and sensor data from multiple wheel motors, including encoder data and torque data, the problem of single sensor information source and inaccurate reflection of vehicle dynamic characteristics in traditional wheel-leg vehicle state estimation is solved. This ensures the comprehensiveness and accuracy of sensor data, realizes the multi-source information fusion basis for vehicles in complex terrain and high-speed conditions, and provides reliable basic data for subsequent slip ratio calculation and vehicle state estimation.

[0036] By using a multimodal slip detection method based on sensor data to calculate the slip ratio of each contacting wheel, this method solves the problems of difficulty in real-time perception of wheel slip, lack of slip compensation leading to observation distortion and accumulation of estimation errors in existing methods. It achieves accurate identification and dynamic modeling of wheel slip state. This multimodal slip detection method combines wheel lock-up detection, kinematic consistency detection and contact force verification, which effectively improves the accuracy and robustness of state estimation.

[0037] By using adaptive weights to fuse the velocity of the wheel kinematic model at the previous moment with the velocity calculated by the inertial measurement unit, the fused velocity at the current moment is obtained. This solves the dynamic inconsistency problem caused by independent estimation of wheel and inertial information, realizes optimal fusion of multi-source velocity information based on slip perception, and improves the continuity and accuracy of vehicle motion estimation.

[0038] By inputting the fusion velocity at the current moment and the actual state observation value at the previous moment into the Kalman filter for state estimation, the problem of insufficient dynamic consistency caused by the decoupling of wheeled and legged motion in existing methods is solved. This achieves unified modeling and fusion estimation of wheeled and legged motion characteristics, thereby significantly improving the state estimation accuracy and system robustness of wheeled and legged vehicles in complex terrain.

[0039] Furthermore, before implementing the wheel-leg vehicle state estimation method, the following steps are also included: The operator issues control commands, the control system receives the control commands and sends them to the platform after optimization. The platform then sends the IMU data from the previous moment and the sensor data of each wheel motor to the state estimation system.

[0040] The control system employs Model Predictive Control (MPC) and Whole Body Control (WBC), which can receive control commands, optimize the trajectory based on system dynamics, and send them to the platform for execution.

[0041] State estimation is divided into three parts: slip detection, adaptive velocity fusion, and Kalman filter. Specifically: Slip detection adopts a multimodal detection framework. For each wheel in contact with the ground, the algorithm consists of three stages: wheel lock-up detection, which determines whether the wheel is locked by using wheel angular velocity and torque thresholds, and assigns the maximum slip ratio when lock-up is detected. By setting the maximum slip ratio, the system can ignore the influence of the wheel during the estimation process. Kinematic consistency detection, which calculates the theoretical wheel tangential velocity (i.e., wheel linear velocity) based on vehicle motion and compares it with the actual wheel tangential velocity to determine the slip situation. Contact force verification, which estimates the longitudinal traction force based on the torque data transmitted from the motor to determine the actual contact condition. Finally, the slip ratio combines these detection mechanisms with dynamic thresholds under different operating conditions through a normalized speed difference index. At the same time, the average slip ratio is introduced to provide a global indication of the severity of slip in the subsequent adaptive fusion.

[0042] Adaptive speed fusion adjusts the fusion weights in real time based on wheel motion, slippage, and vehicle dynamics, establishing a dynamic coupling between the first wheel speed obtained from the wheel kinematics model and the second vehicle speed obtained from IMU data. In the differential kinematics model, we consider the redistribution of vertical force during leg lift. When only one wheel on one side remains in contact (resulting in greater normal force and better traction), the average of the left and right wheel speeds is recalculated before applying the wheel kinematics model. If wheel lockup is detected during wheel speed calculation, the wheel speed is directly set to 0. Then, the fusion weights are determined based on wheel motion, slippage, and vehicle dynamics, and the first wheel speed and the second vehicle speed predicted by the IMU are fused.

[0043] Kalman filters achieve high-precision estimation of the state of wheel-legged vehicles by fusing information from multiple sensor sources. In this system, the Kalman filter integrates odometer data (i.e., actual state observations, including wheel velocity vectors and wheel position vectors) and motion prediction information based on the inertial measurement unit (IMU) to estimate the vehicle's position and velocity. Specifically, the leg odometer provides displacement estimates from foot contact sequences and kinematic models, exhibiting strong terrain adaptability and providing reliable references in cases of wheel slippage or failure; the wheel odometer has high accuracy on smooth road surfaces and is suitable for high-speed driving; while the IMU achieves short-term dynamic prediction of the system through high-frequency acceleration and angular velocity measurements, compensating for the state estimation needs when other sensors have low update frequencies or are temporarily unavailable.

[0044] Furthermore, in step 202, based on the sensor data of each wheel motor from the previous moment, the contact state of each wheel is determined; the contact state includes contact with the ground and non-contact with the ground. Then, multimodal slip detection is performed on wheels in the contact state of contact with the ground, while no slip detection is performed on wheels in the contact state of non-contact with the ground.

[0045] Further, the encoder data includes wheel angular velocity and wheel position, and the inertial measurement unit data includes vehicle linear velocity and vehicle angular velocity. Step 203 involves multimodal slip detection for wheels in contact with the ground to obtain the slip ratio of each wheel, specifically including: The system determines whether the target wheel's angular velocity is lower than a preset angular velocity threshold and whether the target wheel's torque exceeds a preset torque threshold, thus obtaining a first determination result (i.e., wheel lock-up detection). The target wheel is any wheel in contact with the ground in any contact state. The formula is as follows: (1); In the formula, Indicates the first The wheel angular velocity of each wheel; Indicates the first The absolute value of the wheel angular velocity of each wheel. Indicates the angular velocity threshold; Indicates the first Wheel torque of each wheel (i.e. wheel torque data). Indicates the first Maximum wheel threshold for each wheel; Indicates the first The absolute value of the wheel torque of each wheel is considered to be locked when it satisfies formula (1).

[0046] If the first judgment result is yes, then it is determined that the target wheel has locked up, and the target wheel is identified as the locked wheel, and the slip ratio of the locked wheel is set to the preset maximum slip ratio.

[0047] If the first judgment result is negative, then based on the vehicle's linear velocity, angular velocity, and wheel position, the theoretical linear velocity of the target wheel is calculated, and based on the target wheel's angular velocity, the actual linear velocity of the target wheel is calculated. Finally, it is determined whether the deviation between the theoretical and actual linear velocities of the target wheel exceeds a preset dynamic threshold, thus obtaining the second judgment result (kinematic consistency detection), with the formula: (2); In the formula, Indicates the first The theoretical linear velocity of each wheel For the first The actual linear velocity of each wheel This represents the linear velocity of the vehicle body in the world coordinate system. Indicates the angular velocity of the vehicle body. Represents the first in the vehicle body coordinate system The position of each wheel.

[0048] If the second judgment result is yes, then it is determined that there is a slippage. Based on the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel, the slippage rate of the target wheel is calculated.

[0049] If the second judgment result is negative, then based on the target wheel's angular velocity, theoretical wheel linear velocity, and torque data, the longitudinal traction force of the target wheel is calculated, and it is determined whether the longitudinal traction force is greater than a preset traction force threshold to obtain the third judgment result (i.e., contact force verification), the formula is as follows: (3); In the formula, For the first The actual angular acceleration of each wheel; Indicates the time step; Indicates the first The longitudinal traction force of each wheel; Represents the wheel's moment of inertia; This represents the radius of the wheel. If the calculated value is... If the slip ratio exceeds the traction threshold, the slip ratio will be set to the maximum value.

[0050] If the third judgment result is yes, then the slip ratio of the target wheel is set to the preset maximum slip ratio.

[0051] If the third judgment result is negative, the slip ratio of the target wheel is calculated based on the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel.

[0052] Furthermore, based on the theoretical and actual linear velocities of the target wheel, the slip ratio of the target wheel is calculated, specifically as follows: (4); in, Slip ratio; , For the first i The absolute value of the deviation between the theoretical linear velocity and the actual linear velocity of each wheel; For the first i The theoretical linear velocity of each wheel; For the first i The actual linear velocity of each wheel; This is the minimum speed threshold.

[0053] Furthermore, in step 203, the global average slip ratio is calculated based on the slip ratios of all wheels in contact with the ground. This provides a quantitative reference for the severity of slippage in subsequent adaptive fusion.

[0054] Furthermore, in step 204, adaptive weights are determined based on the average slip ratio and a preset wheel confidence index, specifically as follows: (5); in, These are adaptive weights used for vehicle speed calculation; It is a wheel confidence index that reflects the correlation between wheel slippage and the magnitude of acceleration. , It is the linear acceleration of the vehicle body in a local coordinate system. These are empirical normalization parameters, used for normalizing acceleration scale parameters; It is the average slip ratio.

[0055] Furthermore, in step 205, based on the adaptive weights and the obtained first wheel speed and second vehicle speed from the previous moment, the fused speed at the current moment is calculated, specifically as follows: (6); in, It's the fusion speed; It is an adaptive weight; It is the first wheel speed, that is, the average wheel speed calculated using the wheel kinematics model; It is the second vehicle speed, which is the vehicle coordinate system determined based on inertial measurement unit data. x Velocity in the direction of travel.

[0056] Furthermore, in step 206, the fusion speed at the current moment and the actual observed state value from the previous moment are input into the Kalman filter to calculate the vehicle state at the current moment, specifically including: Step a1: Using the fusion velocity at the current moment and the actual state observation value obtained at the previous moment, construct the observation vector of the Kalman filter at the current moment. Input the observation vector at the current moment into the Kalman filter. The actual state observation value at the current moment is obtained by using odometry and includes the wheel velocity vector and wheel position vector. Specifically, the observation vector can be represented as: (7); in, It is the fusion speed after coordinate system transformation. It is the first in the centroid coordinate system i Wheel position vectors of each wheel It is the first i The wheel velocity vectors of each wheel. In this way, the wheel velocity-based observations, along with the joint information and joint velocities from the motor encoder, participate in state estimation, thereby achieving the fusion of multi-sensor data.

[0057] Step a2: Based on the vehicle state at the previous moment and the preset discrete-time state equation, calculate the predicted state value at the current moment.

[0058] Step a3: Calculate the Kalman gain at the current time based on the preset observation matrix and observation noise covariance matrix at the current time, as well as the called prediction state covariance matrix at the current time; the prediction state covariance matrix at the current time is determined based on the updated prediction state covariance matrix of the previous time.

[0059] Step a4: Based on the Kalman gain and the observation vector at the current time, correct the state prediction value at the current time to obtain the updated state prediction value at the current time, and use the updated state prediction value at the current time as the vehicle state at the current time.

[0060] In this application, the specific vehicle state variables are shown in the following formula: (8); in, This represents the position vector of the vehicle body in the world coordinate system; This represents the velocity vector of the vehicle body in the world coordinate system. This indicates the position of the i-th wheel end effector relative to the center of mass; Indicates the position of the wheel in the world coordinate system; This indicates the speed of the wheels.

[0061] The equations of motion in continuous time are: (9); in, This represents the vehicle's acceleration in the world coordinate system. Adaptive process noise arising from complex wheel dynamics.

[0062] The discrete-time state equation is: (10); in, The state matrix, The input matrix is ​​as follows: (11); (12); Where I represents the identity matrix, This represents a zero matrix, where the subscripts indicate the row and column numbers. Represents discrete time.

[0063] The observation equation is: (13); in, The observation matrix is ​​as follows: (14); Further, in step a2, based on the vehicle state at the previous moment and the preset discrete-time state equation, the predicted state value at the current moment is calculated, specifically as follows: (15); in, This is the predicted state value at time k obtained based on the discrete-time state equation. This is the predicted state value of k-1 at the previous time step; The input for the previous time step k-1 refers to the input for the previous time step k-1. .

[0064] Further, in step a3, the Kalman gain at the current time is calculated based on the preset observation matrix and observation noise covariance matrix at the current time, as well as the called prediction state covariance matrix at the current time. Specifically: (16); in, This represents the Kalman gain at the current time k; Let represent the predicted state covariance matrix at the current time k, which represents the uncertainty of the predicted state at the current time obtained based on the state estimate at the previous time. Let represent the observation matrix at the current time k. It maps the state space to the observation space, that is, it describes how to obtain the observation value from the state variable. The observation matrix is ​​the same at each time, i.e., formula (14). This represents the transpose of the observation matrix at the current time k; This represents the observation noise covariance matrix at the current time k. It represents the noise or uncertainty in the observation, which is usually determined by sensor error.

[0065] Furthermore, in step a4, the state prediction value at the current time is corrected based on the Kalman gain and the observation vector at the current time to obtain the updated state prediction value at the current time. This updated state prediction value is then used as the vehicle state at the current time. Specifically: (17); in, This represents the updated state prediction value for the current time k. This represents the Kalman gain at the current time k; This represents the predicted state value at the current time k. This represents the observation matrix at the current time k; This represents the observation vector at the current time k.

[0066] Furthermore, after obtaining the vehicle state, the predicted state covariance matrix at the current time k is updated, and based on the updated predicted state covariance matrix at the current time k, the predicted state covariance matrix at the next time step is determined: (18); in, Represents the identity matrix; This represents the updated predicted state covariance matrix at the current time k.

[0067] This application also provides an application scenario in which the above-described wheel-leg vehicle state estimation method considering slippage is applied. Specifically, the wheel-leg vehicle state estimation method provided in this embodiment can be applied in a wheel-leg vehicle state estimation scenario. The wheel-leg vehicle state estimation scenario includes a wheel-leg vehicle state estimation stage; this stage is used to determine the vehicle state based on sensor data. The wheel-leg vehicle state estimation method provided in this embodiment belongs to the wheel-leg vehicle state estimation stage.

[0068] In one exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram may be as follows. Figure 4 As shown, this computer device includes a processor, memory, input / output (I / O) interfaces, and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is also connected to the system bus via the I / O interfaces. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores processed data. The I / O interfaces are used for exchanging information between the processor and external devices. The communication interface is used for communicating with external terminals via a network connection. When executed by the processor, the computer program implements a method for estimating the state of a wheel-legged vehicle considering slippage.

[0069] Those skilled in the art will understand that Figure 4 The structures shown are merely block diagrams of some structures related to the present application and do not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than shown in the figures, or combine certain components, or have different component arrangements. In an exemplary embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0070] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0071] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0072] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0073] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0074] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0075] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0076] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for estimating the state of a wheel-legged vehicle considering slippage, characterized in that, include: Acquire inertial measurement unit data and sensor data from multiple wheel motors from the previous moment; The sensor data includes encoder data and torque data; Based on the sensor data of each wheel motor in the previous moment, determine the contact state of each wheel; Contact states include contact with the ground and non-contact with the ground; Multimodal slip detection is performed on wheels in contact with the ground to obtain the slip ratio of each wheel, and the average slip ratio is calculated based on the slip ratios of all wheels; the multimodal slip detection includes wheel lock-up detection, kinematic consistency detection, and contact force verification. Based on the average slip ratio and the preset wheel confidence index, adaptive weights are determined. Based on adaptive weights and the first wheel speed and the second vehicle speed obtained at the previous moment, the fusion speed at the current moment is calculated. The first wheel speed was calculated using a wheel kinematics model, while the second vehicle speed was determined based on data from an inertial measurement unit. The current fusion speed and the actual observed state from the previous time are input into the Kalman filter to calculate the vehicle state at the current time. Vehicle status includes vehicle speed and vehicle position.

2. The method for estimating the state of a wheel-leg type vehicle considering slippage according to claim 1, characterized in that, The encoder data includes wheel angular velocity and wheel position, and the inertial measurement unit data includes vehicle linear velocity and vehicle angular velocity; for wheels in contact with the ground, multimodal slip detection is performed to obtain the slip ratio of each wheel, specifically including: The system determines whether the angular velocity of the target wheel is lower than a preset angular velocity threshold and whether the torque data of the target wheel exceeds a preset torque threshold, and obtains a first determination result; the target wheel is any wheel in contact with the ground in any contact state. If the first judgment result is yes, then it is determined that the target wheel has locked up, and the target wheel is identified as the locked wheel, and the slip ratio of the locked wheel is set to the preset maximum slip ratio. If the first judgment result is negative, the theoretical wheel linear velocity of the target wheel is calculated based on the vehicle body linear velocity, vehicle body angular velocity and the wheel position of the target wheel, and the actual wheel linear velocity of the target wheel is calculated based on the wheel angular velocity of the target wheel. The second judgment result is obtained by judging whether the deviation between the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel exceeds a preset dynamic threshold. If the second judgment result is yes, then the slip ratio of the target wheel is calculated based on the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel. If the second judgment result is negative, then based on the target wheel's wheel angular velocity, theoretical wheel linear velocity, and torque data, the longitudinal traction force of the target wheel is calculated, and it is determined whether the longitudinal traction force is greater than the preset traction force threshold to obtain the third judgment result. If the third judgment result is yes, then the slip ratio of the target wheel is set to the preset maximum slip ratio; If the third judgment result is negative, the slip ratio of the target wheel is calculated based on the theoretical wheel linear velocity and the actual wheel linear velocity of the target wheel.

3. The method for estimating the state of a wheel-leg type vehicle considering slippage according to claim 2, characterized in that, Based on the theoretical and actual linear velocities of the target wheel, the slip ratio of the target wheel is calculated as follows: ； in, Slip ratio; For the first i The absolute value of the deviation between the theoretical linear velocity and the actual linear velocity of each wheel; For the first i The theoretical linear velocity of each wheel; For the first i The actual linear velocity of each wheel; This is the minimum speed threshold.

4. The method for estimating the state of a wheel-leg type vehicle considering slippage according to claim 3, characterized in that, Based on the average slip ratio and a preset wheel confidence index, adaptive weights are determined as follows: ； in, It is an adaptive weight; It is a confidence index for wheels. , It is the linear acceleration of the car body. These are empirical normalization parameters; It is the average slip ratio.

5. The method for estimating the state of a wheel-leg type vehicle considering slippage according to claim 3, characterized in that, Based on adaptive weights and the obtained first wheel speed and second vehicle speed from the previous moment, the fused speed at the current moment is calculated as follows: ； in, It's the fusion speed; It is an adaptive weight; It is the speed of the first wheel; That is the second vehicle speed.

6. The method for estimating the state of a wheel-leg type vehicle considering slippage according to claim 1, characterized in that, The current fusion speed and the actual observed state from the previous time step are input into the Kalman filter to calculate the vehicle state at the current time step, specifically including: Based on the fusion speed at the current moment and the actual observation values ​​of the state at the previous moment, the observation vector of the Kalman filter at the current moment is constructed. Based on the vehicle state at the previous moment and the preset discrete-time state equation, the predicted state value at the current moment is calculated. The Kalman gain at the current time is calculated based on the preset observation matrix and observation noise covariance matrix at the current time, as well as the predicted state covariance matrix at the current time. The predicted state covariance matrix at the current time is determined based on the updated predicted state covariance matrix at the previous time. Based on the Kalman gain and the observation vector at the current time, the state prediction value at the current time is corrected to obtain the updated state prediction value at the current time, and the updated state prediction value at the current time is used as the vehicle state at the current time.

7. The method for estimating the state of a wheel-leg type vehicle considering slippage according to claim 6, characterized in that, Based on the preset observation matrix and observation noise covariance matrix at the current time, as well as the retrieved prediction state covariance matrix at the current time, the Kalman gain at the current time is calculated as follows: ； in, This represents the Kalman gain at the current time k; This represents the predicted state covariance matrix at the current time k; This represents the observation matrix at the current time k; This represents the transpose of the observation matrix at the current time k; Let represent the observation noise covariance matrix at the current time k.

8. The method for estimating the state of a wheel-leg type vehicle considering slippage according to claim 6, characterized in that, Based on the Kalman gain and the observation vector at the current time, the state prediction value at the current time is corrected to obtain the updated state prediction value at the current time, as follows: ； in, This represents the updated state prediction value for the current time k. This represents the Kalman gain at the current time k; This represents the predicted state value at the current time k. This represents the observation matrix at the current time k; This represents the observation vector at the current time k.

9. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and capable of running on the processor, characterized in that the processor executes the computer program to implement the wheel-leg vehicle state estimation method considering slippage as described in any one of claims 1-8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the wheel-leg vehicle state estimation method considering slippage as described in any one of claims 1-8.

Images