Wheel speed meter failure analysis system based on observer technology

Abstract

The invention discloses a wheel speed meter failure analysis system based on an observer technology, and the system comprises an accelerometer which is used for measuring an acceleration signal of an electric bicycle; the wheel speed meter is used for measuring a wheel speed signal of the electric bicycle; the global positioning system module is used for measuring the ground speed of the electric bicycle; and the observer module is used for fusing the acceleration signal, the wheel speed signal and the ground speed signal and judging whether the wheel speed meter fails or not. According to the method, the observer is designed, IMU, GPS and wheel speed meter data are fused, more accurate speed measurement is obtained, ebike carrier speed observation is achieved, and failure analysis of the wheel speed meter is carried out.

Description

Technical Field

[0001] The present invention relates to the technical field of vehicle power control and sensor fusion, and in particular to a wheel speed meter failure analysis system based on observer technology. Background Art

[0002] With the widespread use of electric bicycles around the world, their safety and reliability issues have gradually attracted people's attention. As an important sensor for detecting the speed of electric bicycles, the wheel speed meter plays a vital role in the driving control of the vehicle. However, the existing wheel speed meter detection system often faces some challenges in actual use. For example, the wheel speed meter may fail due to sensor aging, external environmental influences or mechanical failures (such as slipping, damage to the speed measuring magnet, fatigue tilt of the Hall sensor, etc.). This failure will not only cause errors in speed measurement, but may even affect the control and safety of the vehicle.

[0003] At present, the detection methods for wheel speed meter failure mainly rely on communicating with the wheel speed meter or using simple filtering algorithms for data analysis. However, these methods have great limitations when dealing with wheel speed meter failure, especially long-term failure and slippage. For example, low-pass filtering has a slow response speed to wheel speed meter slippage, and cannot accurately determine the effectiveness of the wheel speed meter in the case of long-term failure. In addition, the detection accuracy of these methods under complex working conditions (such as non-steady road surface, vehicle acceleration or deceleration) needs to be improved.

[0004] Specifically, solution 1 (effective judgment is made only by communicating with the wheel speed meter):

[0005] Only the speed sensor is used for communication, and failure judgment is made based on whether the speed sensor responds. This method cannot handle the situation where the speed unit is installed fatigue-skewed or the speed sensor is damaged. Logically, the loop cannot be completely closed, and it is impossible to determine whether the wheel speed meter is valid.

[0006] Solution 2 (only using the wheel speed meter data for effective judgment) (Simulation results are as follows Figure 1 shown):

[0007] According to the data of users in non-intense riding, a low-pass filter is designed for the wheel speed meter. Here, a 4th-order Butterworth low-pass filter is designed with a bandwidth of about 0.5HZ. It can be found that this method has the following defects:

[0008] 1. There is a large delay when the wheel speed meter slips, about 0.5 seconds, leaving little time for the controller to respond.

[0009] 2. The success rate of skidding detection is lower.

[0010] 3. After the wheel number meter fails for a long time, it is impossible to determine whether it is valid. It is necessary to generate data changes of the wheel speed meter. It is impossible to handle the situation of fatigue skewness of the speed measurement unit installation and damage of the speed measurement sensor;

[0011] Solution 3 (using only wheel speed meter and IMU accelerometer) (Simulation results as shown in Figure 2 shown):

[0012] As in this scheme, an observer is designed, but only the wheel speed meter and IMU accelerometer are used for observation, which has the following defects:

[0013] 1. Compared with Solution 2, the response speed of slip / failure in this solution is greatly improved, but the delay in slip / failure recovery is still high;

[0014] 2. It is still impossible to determine the effectiveness of the wheel speedometer during long-term slippage / failure;

[0015] 3. There is still room for improvement in the slip / failure detection law. Summary of the invention

[0016] In order to solve the above problems, the present invention provides a wheel speed meter failure analysis system based on observer technology, which obtains a more accurate vehicle speed estimate by fusing the data of the accelerometer, wheel speed meter and GPS module and processing the sensor data using the extended Kalman filter (EKF), and can detect and issue an alarm in time when the wheel speed meter fails or slips. Compared with traditional methods, the system can not only improve the response speed to wheel speed meter failure, but also maintain a high detection accuracy under complex working conditions, thereby significantly improving the driving safety and stability of electric bicycles.

[0017] The present invention is realized by the following technical scheme: A wheel speed meter failure analysis system based on observer technology, comprising:

[0018] An accelerometer, used to measure the acceleration signal of the electric bicycle;

[0019] Wheel speed meter, used to measure the wheel speed signal of the electric bicycle;

[0020] a GPS module to measure the ground speed of the e-bike;

[0021] An observer module, used for fusing the acceleration signal, the wheel speed signal and the ground speed signal to estimate the actual speed of the electric bicycle;

[0022] An extended Kalman filter, embedded in the observer module, is used to:

[0023] Establish the state vector of the system, including the derivatives of the speed and acceleration of the electric bicycle;

[0024] Using the measurements of the accelerometer, wheel speed meter and GPS module, combined with the dynamic model of the system, to perform state prediction and measurement update to generate an estimated vehicle speed;

[0025] Performing real-time adjustment of the covariance matrix of system noise and measurement noise to optimize the estimated value of the vehicle speed;

[0026] The failure analysis module is used to compare the estimated vehicle speed with the measured speed of the wheel speed meter, and when it is detected that the deviation between the two exceeds a preset threshold, determine the failure or slip state of the wheel speed meter.

[0027] As a preferred technical solution, the observer module performs speed estimation through the following steps:

[0028] Initialize the system's state estimate and covariance matrix;

[0029] Use the state transfer matrix to predict the state at the next moment and update the state according to the measured value;

[0030] The Kalman gain is calculated and the predicted value is corrected in combination with the actual measured value to generate an updated speed estimate.

[0031] As a preferred technical solution, the extended Kalman filter dynamically adjusts the weight coefficients in the measurement model.

[0032] As a preferred technical solution, the failure analysis module includes a threshold setting unit, which adjusts the deviation threshold for determining failure or slippage of the wheel speed meter according to the real-time working condition of the electric bicycle.

[0033] As a preferred technical solution, it also includes an output module. When the wheel speed meter is judged to be failed or slipping, the output module generates an alarm signal or a control signal to remind the driver or control the vehicle-mounted system to perform corresponding processing.

[0034] As a preferred technical solution, it also includes a neural network-based adaptive filtering module, which is used to dynamically adjust the noise covariance matrix of the EKF to further improve the system's response speed and accuracy to wheel speed meter failure.

[0035] As a preferred technical solution, a data storage module is also included to store historical data of the accelerometer, wheel speed meter and GPS module, as well as the result data of failure analysis, so as to be used for subsequent data analysis and optimization of system performance.

[0036] As a preferred technical solution, a redundant sensor module is also included. When any one of the accelerometer, wheel speed meter or GPS module fails, the redundant sensor module is started to ensure that the system can continue to perform speed estimation and failure detection.

[0037] The beneficial effects of the present invention are as follows: the wheel speed meter failure analysis system based on observer technology proposed by the present invention realizes accurate estimation of the speed of the electric bicycle and efficient detection of wheel speed meter failure by fusing multi-source data of the accelerometer, wheel speed meter and GPS module and using the extended Kalman filter (EKF) for real-time processing. Compared with the prior art, the present invention has the following significant advantages:

[0038] First, the system can effectively deal with the problems of wheel speed meter slipping and long-term failure under harsh working conditions. By fusing data from different sensors, the accuracy and reliability of speed measurement are significantly improved.

[0039] Secondly, the introduction of the extended Kalman filter enables the system to dynamically adjust the noise weight of the sensor, further optimize the speed estimation, and improve the response speed and accuracy of failure detection;

[0040] Thirdly, the system has a threshold adaptive adjustment function, which can flexibly adjust the failure judgment criteria according to the actual operating conditions of the electric bicycle, thereby improving the robustness and adaptability of the detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0042] Figure 1 It is a simulation schematic diagram of scheme 2 in the prior art;

[0043] Figure 2 It is a simulation schematic diagram of scheme three in the prior art;

[0044] Figure 3 It is a simulation schematic diagram of the present invention. DETAILED DESCRIPTION

[0045] All features disclosed in this specification, or steps in all methods or processes disclosed, except mutually exclusive features and / or steps, can be combined in any manner.

[0046] Any feature disclosed in this specification (including any additional claims, abstract and drawings), unless otherwise stated, may be replaced by other equivalent or alternative features with similar purposes. That is, unless otherwise stated, each feature is only an example of a series of equivalent or similar features.

[0047] A wheel speed meter failure analysis system based on observer technology of the present invention comprises an accelerometer for measuring acceleration signals of an electric bicycle;

[0048] Wheel speed meter, used to measure the wheel speed signal of the electric bicycle;

[0049] a GPS module to measure the ground speed of the e-bike;

[0050] An observer module, used for fusing the acceleration signal, the wheel speed signal and the ground speed signal to estimate the actual speed of the electric bicycle;

[0051] An extended Kalman filter, embedded in the observer module, is used to:

[0052] Establish the state vector of the system, including the derivatives of the speed and acceleration of the electric bicycle;

[0053] Using the measurements of the accelerometer, wheel speed meter and GPS module, combined with the dynamic model of the system, to perform state prediction and measurement update to generate an estimated vehicle speed;

[0054] Performing real-time adjustment of the covariance matrix of system noise and measurement noise to optimize the estimated value of the vehicle speed;

[0055] The failure analysis module is used to compare the estimated vehicle speed with the measured speed of the wheel speed meter, and when it is detected that the deviation between the two exceeds a preset threshold, determine the failure or slip state of the wheel speed meter.

[0056] In this embodiment, the observer module performs speed estimation through the following steps:

[0057] Initialize the system's state estimate and covariance matrix;

[0058] Use the state transfer matrix to predict the state at the next moment and update the state according to the measured value;

[0059] The Kalman gain is calculated and the predicted value is corrected in combination with the actual measured value to generate an updated speed estimate.

[0060] Among them, the extended Kalman filter dynamically adjusts the weight coefficients in the measurement model.

[0061] The failure analysis module includes a threshold setting unit, which adjusts the deviation threshold for determining failure or slippage of the wheel speed meter according to the real-time working condition of the electric bicycle.

[0062] In this embodiment, an output module is also included. When the wheel speed meter is determined to be failed or slipping, the output module generates an alarm signal or a control signal to remind the driver or control the vehicle-mounted system to perform corresponding processing.

[0063] In this embodiment, a neural network-based adaptive filtering module is also included, and the module is used to dynamically adjust the noise covariance matrix of the EKF to further improve the response speed and accuracy of the system to the failure of the wheel speed meter.

[0064] A data storage module is also included for storing historical data of the accelerometer, wheel speed meter and GPS module, as well as result data of failure analysis, for subsequent data analysis and optimization of system performance.

[0065] In this embodiment, a redundant sensor module is also included. When any one of the accelerometer, wheel speed meter or GPS module fails, the entire failure analysis system can continue to operate, but the failure detection performance will be reduced.

[0066] The present invention introduces GPS ground speed data as observer input to improve the effectiveness of speed observation when correcting the long-term error of the wheel speed meter.

[0067] Improve the response speed of slippage / failure. (About 12% faster than Solution 2)

[0068] Improve the response speed of wheel speed meter slip / failure recovery (in this working condition, it is 20-30% higher than that of scheme 3).

[0069] The introduction of GPS ground speed can improve the observation accuracy of the system, make more detailed adjustments to the slip threshold, and improve the detection rate of slip / failure (in this working condition, it is 6% higher than that of Solution 3).

[0070] System Status

[0071] We choose the state vector of the system as:

[0072] x=[vv · ]

[0073] Where, v is the speed of the ebike, v · is the derivative of velocity (i.e. acceleration).

[0074] System dynamics model

[0075] We can use a simple linear model to describe the dynamics of the system:

[0076] x· =Ax+Bu+w

[0077] Where A =

[0010] , B =

[01] , u is the accelerometer measurement (control input), and w is the process noise.

[0078] Measurement Model

[0079] We have three measurement sources: accelerometer, wheel speed meter and GPS. The measurement equation can be expressed as:

[0080] z=Hx+vwhere z=zaccelzwheelzGPS,H=111000

[0081] zaccel, zwheel and zGPS are the velocity measurements of the accelerometer, wheel speed meter and GPS respectively, and v is the measurement noise.

[0082] Extended Kalman Filter Design

[0083] Initialization: Set the initial state estimate x^0 and the initial covariance matrix P0.

[0084] Prediction Steps:

[0085] Compute the state transfer matrix Φ(t) = exp(At) (for linear systems this simplifies to At+I).

[0086] Prediction state: x^k|k-1=Φ(t)x^k-1+Buk.

[0087] Prediction covariance: Pk|k-1=Φ(t)Pk-1Φ(t)T+Q, where Q is the process noise covariance.

[0088] Update steps:

[0089] Calculate the Kalman gain: Kk=Pk|k-1HT(HPk|k-1HT+R)-1.

[0090] Update status: x^k=x^k|k-1+Kk(zk-Hx^k|k-1).

[0091] Update covariance: Pk = (I-KkH)Pk|k-1.

[0092] Output and Control

[0093] Output: The output of the system is the velocity estimate v^, which is the first element of the state vector.

[0094] Control: The accelerometer measurements u can be used as control input to drive the system dynamics model.

[0095] Full-Order State Observer

[0096] Since our state vector already contains the velocity and its derivatives, and we use an Extended Kalman Filter to estimate these states, this filter itself is a full-order state observer. It provides a complete estimate of the system state and can be adjusted as needed to optimize performance.

[0097] The simulation results are as follows Figure 3 shown.

[0098] The present invention is based on a mechanical observer, that is, speed is equal to the integral of acceleration. Acceleration, GPS ground speed and wheel speed meter are "proportionally" fused. A more compromised speed observation value is obtained. In actual working conditions, the three sensors of accelerometer, GPS ground speed and wheel speed meter will have noise. Among them, GPS ground speed will also have certain nonlinear distortion due to environmental reasons. Therefore, EKF is designed to estimate the "proportion". A more accurate speed observation value is obtained, and this speed observation value is subtracted from the wheel speed meter. When the speed observation value deviates too much from the wheel speed meter, it is judged that the wheel speed meter has failed or slipped. Thereby achieving the purpose of failure analysis of failed wheel speed meter.

[0099] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that are not conceived through creative work should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope defined in the claims.

Claims

1. A wheel speed meter failure analysis system based on observer technology, characterized in that: include: An accelerometer, used to measure the acceleration signal of the electric bicycle; Wheel speed meter, used to measure the wheel speed signal of the electric bicycle; a GPS module to measure the ground speed of the e-bike; An observer module, used for fusing the acceleration signal, the wheel speed signal and the ground speed signal to estimate the actual speed of the electric bicycle; An extended Kalman filter, embedded in the observer module, is used to: Establish the state vector of the system, including the derivatives of the speed and acceleration of the electric bicycle; Using the measurements of the accelerometer, wheel speed meter and GPS module, combined with the dynamic model of the system, to perform state prediction and measurement update to generate an estimated vehicle speed; Performing real-time adjustment of the covariance matrix of system noise and measurement noise to optimize the estimated value of the vehicle speed; The failure analysis module is used to compare the estimated vehicle speed with the measured speed of the wheel speed meter, and when it is detected that the deviation between the two exceeds a preset threshold, determine the failure or slip state of the wheel speed meter.

2. The wheel speed meter failure analysis system based on observer technology according to claim 1 is characterized in that: The observer module performs speed estimation through the following steps: Initialize the system's state estimate and covariance matrix; Use the state transfer matrix to predict the state at the next moment and update the state according to the measured value; The Kalman gain is calculated and the predicted value is corrected in combination with the actual measured value to generate an updated speed estimate.

3. The wheel speed meter failure analysis system based on observer technology according to claim 1, characterized in that: The extended Kalman filter dynamically adjusts the weight coefficients in the measurement model.

4. The wheel speed meter failure analysis system based on observer technology according to claim 1, characterized in that: The failure analysis module includes a threshold setting unit, which adjusts the deviation threshold for determining failure or slippage of the wheel speed meter according to the real-time working condition of the electric bicycle.

5. The wheel speed meter failure analysis system based on observer technology according to claim 1, characterized in that: It also includes an output module. When the wheel speed meter is judged to be failed or slipping, the output module generates an alarm signal or a control signal to remind the driver or control the vehicle system to perform corresponding processing.

6. The wheel speed meter failure analysis system based on observer technology according to claim 1, characterized in that: The invention also includes a neural network-based adaptive filtering module, which is used to dynamically adjust the noise covariance matrix of the EKF to further improve the response speed and accuracy of the system to the failure of the wheel speed meter.

7. The wheel speed meter failure analysis system based on observer technology according to claim 1 is characterized in that: A data storage module is also included for storing historical data of the accelerometer, wheel speed meter and GPS module, as well as result data of failure analysis, for subsequent data analysis and optimization of system performance.

8. The wheel speed meter failure analysis system based on observer technology according to claim 1, characterized in that: A redundant sensor module is also included. When any one of the accelerometer, wheel speed meter or GPS module fails, the redundant sensor module is activated to ensure that the system can continue to perform speed estimation and failure detection.

Images