A method for judging wheel slip based on IMU acceleration

Abstract

The present application relates to the technical field of online monitoring of the motion state of a wheeled carrier, and specifically to a method for judging wheel skidding based on an IMU acceleration, aiming to solve the problems of poor scene adaptability caused by the dependence of existing skidding detection on external sensors, and the accumulation of errors and low detection accuracy caused by the integration of a single inertial measurement unit. The present application adopts a multi-step closed-loop processing procedure to successively complete sensor zero offset calibration, data filtering and noise reduction, coordinate system conversion, trapezoidal integral operation, stationary integral zeroing, periodic integral alignment and skidding judgment. The present application discards various external positioning sensor devices and only relies on the inertial measurement unit and the wheeled odometer to work, and can be adapted to various complex closed scenes. The present application effectively avoids the problem of integral error accumulation through multi-level error suppression and drift correction means, reduces the false detection rate relying on continuous frame judgment logic, and is lightweight in algorithm, can be adapted to real-time operation of vehicle embedded devices, and is suitable for skidding monitoring scenes of various wheeled carriers.

Description

Technical Field

[0001] This invention relates to the field of online monitoring technology for the motion state of wheeled vehicles, and in particular to a method for determining wheel slippage based on IMU acceleration, applicable to real-time wheel slippage detection scenarios for various wheeled equipment such as autonomous passenger vehicles, unmanned engineering vehicles, warehouse AGV robots, and unmanned delivery vehicles in industrial parks. Background Technology

[0002] Existing wheel slippage detection technologies suffer from two major inherent flaws that have long limited the accuracy and applicability of slippage detection: First, traditional slip detection solutions heavily rely on external auxiliary sensors, resulting in poor environmental robustness. The mainstream slip detection methods in the industry fall into three categories: 1) Relying solely on wheel speed sensor differences to determine slippage; vehicle bumps, transmission mechanism clearances, and road surface irregularities can cause instantaneous changes in wheel speed signals, frequently leading to false slippage detections; 2) Combining GPS positioning data with wheel speed mileage comparison to identify slippage; however, satellite signals are lost in scenarios such as tunnels, underground parking garages, high-rise buildings, and indoor factories, rendering the slippage detection function directly ineffective; 3) Using visual odometers to assist slippage detection; however, image features are lost in environments with strong light, darkness, dust, and missing ground texture, making it impossible to output valid displacement data; and when the wheels spin or slip, the mileage output by the wheel odometer deviates significantly from the vehicle's actual travel distance, making it impossible to distinguish between normal driving disturbances and true slippage conditions based solely on odometer data.

[0003] Second, there is a serious problem of error accumulation in calculating vehicle displacement using a single IMU integration. The acceleration and angular velocity signals acquired by the IMU come with a factory-constant zero bias from the sensor and random high-frequency noise from road bumps. Directly integrating the raw IMU data will quickly generate speed and displacement drift errors, and these errors will continue to accumulate over time. When the vehicle is stationary for a long time, the IMU zero bias will continue to be included in the integration result, further amplifying the mileage deviation. Existing technology lacks a complete closed-loop error suppression process and cannot solve the IMU integration drift problem, resulting in low reliability of IMU-calculated mileage and difficulty in stably cooperating with the odometer to complete slippage detection.

[0004] Currently, there is a lack of a complete, lightweight, closed-loop solution that does not rely on external sensors, capable of eliminating various IMU errors layer by layer from the source, controlling integral drift in layers, and accurately filtering instantaneous disturbances to achieve low false detection slippage recognition.

[0005] Therefore, a new technical solution is urgently needed to solve the above-mentioned technical problems. Summary of the Invention

[0006] The purpose of this invention is to overcome the problems of the prior art and provide a method for judging wheel slippage based on IMU acceleration. This method addresses the issues that existing wheel slippage detection technologies mostly rely on external sensors such as satellite positioning or vision, which have weak environmental adaptability and cannot function properly in scenarios without satellite signals or when images are unavailable. Conventional wheel speed detection methods are also susceptible to road bumps and mechanical clearances, leading to misjudgments. Furthermore, the integration operation of a single inertial measurement unit is easily affected by noise and zero bias, resulting in serious accumulation of integration drift errors. After long-term operation, the mileage calculation data becomes distorted, making it difficult to accurately identify wheel slippage conditions. The detection stability and accuracy are low, failing to meet the real-time wheel slippage detection requirements in various complex scenarios.

[0007] The above objectives are achieved through the following technical solutions: A method for determining wheel slippage based on IMU acceleration includes the following steps: S1. Zero bias calibration: After the vehicle is powered on and kept completely stationary, multiple frames of raw triaxial acceleration and triaxial angular velocity samples from the IMU are collected. The inherent zero bias of the accelerometer and gyroscope is calculated by the mean of the samples. The zero bias is used to perform subtraction correction on the raw IMU data to complete the initial error calibration of the sensors. S2. Filtering: The acceleration data after zero bias correction is sequentially fed into the moving average filter and the first-order IIR low-pass filter. The two-stage filtering works together to suppress the random noise from road bumps and the high-frequency abnormal peaks caused by motor vibration, and outputs a smooth acceleration signal without abrupt changes. S3. Coordinate Transformation: Based on the calibrated IMU data, a high-precision attitude angle is obtained through complementary filtering and fusion. A rotation matrix from the carrier coordinate system to the Northeast-East-West world coordinate system is constructed according to the ZYX rotation order. The acceleration of the carrier coordinate system is multiplied by the rotation matrix on the left and the gravity vector component is removed to obtain the linear acceleration of the world coordinate system that only reflects the horizontal motion of the vehicle. S4. Integration Calculation: The trapezoidal integration method is used to iteratively calculate the velocity increment and displacement increment of the linear acceleration in the world coordinate system frame by frame, and the IMU integral mileage is accumulated in real time; the wheel odometer pulse signal is collected synchronously, and the odometer cumulative mileage is continuously counted. S5. Zeroing at rest: Real-time reading of the displacement increment of the odometer in a single frame to determine whether the vehicle is completely stationary; if the vehicle is determined to be stationary, the IMU integral speed and IMU integral mileage are directly set to zero to eliminate the accumulation of integral error caused by the continuous superposition of zero bias during the stationary phase. S6. Periodic Integral Alignment: Set three rigid constraints that must be met simultaneously. When all conditions are met, periodically use the odometer to accumulate mileage to cover and correct the IMU integral mileage, suppressing integral drift caused by long-distance driving and ensuring long-term stability of IMU integral data. S7. Slippage Detection: Calculate the difference between the IMU integral mileage and the odometer cumulative mileage frame by frame, and set the slippage distance threshold and the continuous over-limit count threshold; only when the mileage difference exceeds the slippage distance threshold for multiple consecutive frames is the wheel slippage determined; a single frame instantaneous over-limit does not trigger the determination, and the false slippage signal caused by a single disturbance is filtered out.

[0008] This invention provides a method for determining wheel slippage based on IMU acceleration. By relying solely on the inertial measurement unit and wheel odometer to complete slippage detection, the detection scenario is not limited by the operating conditions of external sensors, and it is adaptable to closed and complex scenarios without satellites or effective images. Through multi-level data calibration and filtering, the original detection error of the sensor is effectively suppressed. A static zeroing and periodic integration alignment mechanism achieves hierarchical control of integral drift error. A continuous multi-frame threshold determination method filters out instantaneous disturbances, significantly reducing the probability of false slippage detection and ensuring accurate and stable detection results. Attached Figure Description

[0009] Figure 1 This is a flowchart of a method for determining wheel slippage based on IMU acceleration, as described in this invention. Detailed Implementation

[0010] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments. The described embodiments are merely some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0011] like Figure 1 As shown, this solution provides a method for determining wheel slippage based on IMU acceleration, including the following steps: Step 1: Zero bias calibration After the vehicle is powered on and started, it is kept completely stationary with no wheel rotation, and 50 frames of raw three-axis acceleration data are continuously collected from the IMU. Three-axis gyroscope raw angular velocity The IMU sampling frequency was 50Hz, and the total acquisition time was 1s. The average value of 50 samples from each acceleration channel and each angular velocity channel was calculated to obtain the accelerometer zero bias. gyroscope zero bias The calculation formula is as follows: Accelerometer bias calculation formula: ; Gyroscope bias calculation formula: ; In the formula, Take 50, subtract the corresponding channel zero bias from each frame of raw data to complete the initial calibration, and output the corrected acceleration (acc) and corrected angular velocity: gyro: , , This step eliminates the constant bias that comes with the sensor from the factory, thus eliminating the continuous interference of fixed errors on subsequent integration calculations from the source.

[0012] Step 2: Filtering Input the corrected acceleration acc from step 1 into the two-stage cascaded filter structure: The first stage is a moving average filter, which maintains a fixed sliding window of length M. For each new frame of acceleration data collected, the earliest frame of data in the window is discarded, and the average value of all accelerations in the window is calculated as the output to smooth out random high-frequency noise caused by road surface bumps and vehicle body vibrations. The second stage is a first-order IIR low-pass filter, configured with a preset filter coefficient α, which filters out instantaneous abnormal peak signals caused by motor vibration and tire impact with road surface bumps, and fully retains the low-frequency effective acceleration components corresponding to vehicle constant speed, acceleration and deceleration, and outputs smooth acceleration data without abrupt changes to the next process.

[0013] Step 3: Coordinate Transformation First, perform complementary filtering attitude calculation: 1) Static attitude calculation: Static roll and pitch angles are calculated using filtered acceleration. , ; 2) Dynamic attitude integration: The dynamic attitude is updated frame by frame based on the frame-by-frame integration of the corrected angular velocity gyro. For IMU single-frame sampling interval: , , ; 3) Attitude-weighted fusion: , ; Based on fused attitude angle Construct the rotation matrix of the carrier to the world coordinate system according to the ZYX rotation order. : , In the formula, To integrate roll angle, To integrate pitch angle, The yaw angle is obtained by integrating the gyroscope. The acceleration in the carrier coordinate system is transformed to the world coordinate system and the gravitational component is removed: , Output The linear acceleration in the horizontal world coordinate system represents the pure vehicle motion after removing vehicle body tilt and gravity interference.

[0014] Step 4: Integral Calculation Using the trapezoidal integral method based on horizontal acceleration in the world coordinate system Iterative updates of speed and displacement frame by frame: The velocity iteration formula is as follows: , The displacement iteration formula is as follows: , Frame-by-frame accumulation Total IMU miles earned Synchronously acquire the output pulses of the wheel odometer, calculate and continuously accumulate them based on the pulse equivalent to obtain the accumulated mileage of the odometer. The two sets of mileage data are transmitted synchronously to subsequent processes.

[0015] Step 5: Zero the system at rest to avoid the accumulation of integration errors. The system reads the displacement increment of each frame output from the odometer in real time and continuously counts the displacement data across multiple frames. If the displacement increment of each frame for a preset number of consecutive frames is less than a preset stationary threshold, the vehicle is determined to be completely stationary. After determining that the vehicle is stationary, the IMU integral speed is immediately calculated. IMU miles All values ​​are zeroed out to clear all IMU drift errors accumulated during the vehicle's static period, preventing excessive mileage deviation when restarting after a long period of parking.

[0016] Step 6: Periodic integration alignment to ensure data stability The system synchronously verifies three constraints in real time, performing mileage alignment correction only when all three conditions are met. If any one condition is not met, the correction is skipped. This prevents the correction from masking the actual slippage conditions and alignment deviations caused by acceleration and deceleration dynamics. It also balances drift suppression and slippage sensitivity to ensure stable and reliable data. Constraint 1: If the system slippage detection flag is false, the vehicle is not currently in a state of wheel slippage; Constraint 2: Real-time world coordinate system acceleration magnitude The vehicle drove smoothly without any sudden acceleration or deceleration. Constraint 3: The time interval between the system timing and the last integral alignment correction is ≥10s; Correct execution logic: Accumulate mileage using the odometer. Directly overwrite and update IMU points and miles It corrects the IMU integral drift that accumulates over long distances, ensuring that the IMU-estimated mileage is stable and reliable in the long term.

[0017] Step 7: Slippage detection, compare IMU (Integrated Memory Unit) mileage with odometer mileage. Read frame by frame for synchronous updates and Calculate the mileage difference: Real-time comparison Set a continuous over-limit counter based on the preset slippage distance threshold: If a single frame Less than or equal to the slippage distance threshold: The continuous over-limit counter is reset to zero, and the vehicle is determined to be driving normally; If a single frame If the distance exceeds the slippage threshold: increment the continuous over-limit counter value by 1; When the continuous over-limit counter accumulates to a preset number of consecutive frames, a stable wheel slippage detection signal is output; if the count has not reached the preset number of consecutive frames... If the distance drops back to within the slippage distance threshold, the counter is reset to zero directly, and no slippage alarm is triggered.

[0018] This multi-frame decision logic can effectively filter out single instantaneous mileage deviations caused by road bumps and transmission mechanism gaps, significantly reducing the probability of false detections of slippage.

[0019] This embodiment uses only two types of vehicle-mounted sensors: IMU and wheel odometer, without requiring any external positioning equipment. The entire seven-step closed-loop error suppression and slippage determination process works in tandem, forming a complete link from the calibration of the sensor's raw data to the final slippage output. It can stably achieve high-precision and low-false-detection real-time wheel slippage detection in tunnels, underground garages, and indoor environments without satellite.

[0020] The above description is merely illustrative of the embodiments of the present invention and is not intended to limit the present invention. For those skilled in the art, any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for determining wheel slippage based on IMU acceleration, characterized in that, Includes the following steps: S1. Zero bias calibration: Collect multiple frames of raw acceleration and angular velocity samples of the IMU when the vehicle is stationary, calculate the mean value of the samples to obtain the sensor zero bias, and use the raw data of the zero bias calibration IMU to complete the initial calibration. S2. Filtering: Perform moving average filtering and low-pass filtering sequentially on the IMU acceleration data after zero bias correction to suppress random noise and abnormal peaks. S3. Coordinate Transformation: Based on the attitude calculation, a rotation matrix is ​​constructed to transform the corrected acceleration in the carrier coordinate system to the world coordinate system and remove the gravity component to obtain the pure linear acceleration of the vehicle. S4. Integral Calculation: The trapezoidal integration method is used to iteratively calculate the velocity increment and displacement increment frame by frame based on the linear acceleration of the world coordinate system, and the IMU integral mileage is updated in real time; the cumulative mileage output by the wheel odometer is also statistically analyzed simultaneously. S5. Zeroing when stationary: Real-time monitoring of the vehicle's stationary state. When the vehicle remains stationary, the IMU integral speed and IMU integral mileage are both set to zero to prevent the accumulation of integral error during the stationary phase. S6. Periodic integration alignment: When multiple preset constraints are met, the IMU integral mileage is periodically corrected by using the odometer to accumulate mileage, ensuring the long-term stability of the integral data. S7. Slippage determination: The difference between the IMU integral mileage and the odometer cumulative mileage is calculated in real time. When the difference exceeds the preset slippage distance threshold for multiple consecutive frames, the vehicle wheels are determined to be slipping.

2. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, The specific implementation method of zero bias calibration in step S1 is as follows: after the vehicle starts, it remains stationary and collects multiple frames of raw IMU data. The zero bias of the accelerometer and the zero bias of the gyroscope are solved by the mean formula respectively. The raw acceleration and raw angular velocity are corrected by the zero bias to obtain the corrected acceleration and corrected angular velocity. Accelerometer bias calculation formula: ; Gyroscope bias calculation formula: ; Corrected acceleration: Corrected angular velocity: ; Where N is the preset number of calibration sampling frames, For the original acceleration, This represents the original angular velocity.

3. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, The moving average filtering in step S2 uses a fixed-length sliding window, taking the average of all acceleration data within the window as the output to smooth random noise; the low-pass filtering is a first-order IIR low-pass filtering, configured with preset filtering coefficients to filter out high-frequency abnormal peaks caused by road vibration and motor jitter, while retaining the low-frequency acceleration components of the vehicle's true motion.

4. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, In step S3, complementary filtering is used to complete the attitude calculation. Specifically, the static roll and pitch angles are calculated using the corrected acceleration, the dynamic attitude angles are obtained by integrating the corrected angular velocity, and the static and dynamic attitude angles are weighted and fused to obtain the fused attitude angles. A rotation matrix from the carrier to the world coordinate system is constructed based on the ZYX rotation sequence. ; Coordinate transformation calculation formula: In the formula Linear acceleration in the world coordinate system This is the acceleration due to gravity.

5. The method for determining wheel slippage based on IMU acceleration according to claim 4, characterized in that, The rotation matrix expression corresponding to the ZYX rotation order is: In the formula To integrate roll angle, To integrate pitch angle, The yaw angle is obtained by integrating the gyroscope.

6. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, The trapezoidal integration method described in step S4 includes velocity iteration formulas and displacement iteration formulas, as follows: Velocity iteration formula: ; Displacement iteration formula: ; In the formula, The sampling time interval for a single frame of the IMU. Earn IMU miles .

7. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, The vehicle stationary determination logic in step S5 is as follows: continuously monitor the displacement increment of the odometer frame for multiple frames. When the displacement increment of a preset number of consecutive frames is less than the stationary threshold, the vehicle is determined to be stationary. After the stationary determination, all IMU integral speed and IMU integral mileage are immediately cleared to zero.

8. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, The periodic integral alignment described in step S6 requires that three constraints be met simultaneously before the correction can be performed: Condition 1: The wheel slippage detection has not been triggered at present; Condition 2: The vehicle's real-time acceleration modulus is less than the preset steady acceleration threshold, and the vehicle is in a steady driving state; Condition 3: The time interval since the last integral alignment correction is greater than or equal to the preset correction interval. The correction method involves using the odometer's accumulated mileage to overwrite the updated IMU integral mileage.

9. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, The formula for calculating the mileage difference in the slippage determination mentioned in step S7 is as follows: In the formula, Accumulate mileage for the odometer. IMU miles; Set the slippage distance threshold and the continuous count threshold, only when When the number of consecutive preset frames exceeds the slippage distance threshold, a wheel slippage judgment signal is output; if the instantaneous difference in a single frame exceeds the limit, the judgment is not triggered.

10. The method for determining wheel slippage based on IMU acceleration according to claim 1, characterized in that, The method uses only two types of sensors—onboard IMU and wheel odometer—to achieve full-process detection of wheel slippage.

Images