Driving wheel longitudinal force estimation method based on wheel speed Kalman filtering-LESO

Abstract

The invention provides a driving wheel longitudinal force estimation method based on wheel speed Kalman filtering-LESO. The method comprises the steps that S1, an unmanned mine car driving wheel speed filter is designed based on a Kalman filter to filter the wheel speed of a driving wheel; and S2, designing a linear expansion observer of the wheel speed of the driving wheel of the unmanned mine car, and obtaining the longitudinal force of the driving wheel in combination with the wheel speed of the driving wheel output in the step S1. According to the method disclosed by the invention, the longitudinal force of the driving wheel of the unmanned mine car is estimated by utilizing the linear expansion observer LESO after the KF noise reduction based on the wheel speed, so that the estimation value precision of the longitudinal force of the driving wheel is improved.

Description

Technical Field

[0001] The invention belongs to the field related to vehicle chassis control, and in particular relates to a method for estimating the longitudinal force of a driving wheel based on wheel speed Kalman filtering-LESO. Background Art

[0002] Unmanned mining vehicles play a vital role in mines, a key resource production area. It not only solves the problem of driver fatigue and high risk of operating errors in the traditional human driving mode, but also optimizes the mining operation process through highly automated and intelligent operations, greatly improving production efficiency. The promotion and application of unmanned mining vehicles indicates that the mining industry is moving towards a safer, more efficient and intelligent direction, which has far-reaching significance and great value.

[0003] However, unmanned mining vehicles face severe weather challenges in their applications, especially when the road surface is slippery due to rain or snow, which may cause safety issues such as drive slippage and brake lock, causing the vehicle to lose control or deviate from the path, greatly increasing transportation risks. Therefore, accurately estimating the adhesion coefficient between the drive wheel and the road surface is crucial to ensure the safe operation of unmanned mining vehicles. By accurately estimating the adhesion coefficient, the drive wheel slip rate can be adjusted in real time to maximize the use of road adhesion and ensure all-round stability of the vehicle.

[0004] In the above, the accurate estimation of the adhesion coefficient depends on the key variable of the longitudinal force of the driving wheel. However, the unmanned mining vehicles currently on the market and put into use do not have the conditions to directly measure the longitudinal force of the driving wheels through on-board sensors. Therefore, how to effectively and accurately estimate this value has become the key to improving the anti-skid and anti-lock braking performance of the mining vehicle drive. Summary of the invention

[0005] The purpose of the present invention is to provide a method for estimating the longitudinal force of a driving wheel based on wheel speed Kalman filtering-LESO. After filtering the data measured by the driving wheel sensor using a Kalman filter (KF), the longitudinal force of the driving wheel is observed and estimated based on a linear expansion observer (LESO).

[0006] A method for estimating longitudinal force of a driving wheel based on wheel speed Kalman filter-LESO comprises the following steps:

[0007] Step S1, designing a driving wheel speed filter for the unmanned mining vehicle based on the Kalman filter to filter the driving wheel speed;

[0008] Step S2: design a linear expansion observer of the driving wheel speed of the unmanned mining vehicle, and obtain the driving wheel longitudinal force in combination with the driving wheel speed output in step S1.

[0009] Preferably, step S1 specifically includes the following steps:

[0010] Step S1.1, based on the Kalman filter, the calculated value containing process noise and the observed value containing observation noise are fused to realize the noise reduction estimation of the state quantity;

[0011] Step S1.2, designing a Kalman filter for the wheel speed of the unmanned mine vehicle;

[0012] Step S1.3: Filter the driving wheel speed.

[0013] Preferably, step S2 specifically includes the following steps:

[0014] Step S2.1, based on the linear expansion observer construction, derive the state observer structure of the first-order system;

[0015] Step S2.2, based on the unmanned mining vehicle, build a quarter vehicle model, regard the expression related to the longitudinal force of the driving wheel as an unknown bounded disturbance, and derive the linear expansion observer expression of the disturbance;

[0016] Step S2.3, input the driving wheel speed outputted in step S1 to the linear expansion observer in step S2.2, and the linear expansion observer outputs a convergence value;

[0017] Step S2.4, inversely calculate the longitudinal force of the driving wheel according to the convergence value.

[0018] Compared with the prior art, the advantages of the present invention are:

[0019] Using Kalman filtering to filter the rotation speed of the driving wheel is beneficial to reducing the influence of sensor system errors, accidental errors and other external noise on the wheel speed and angular acceleration sensors of the unmanned mining wheel; using a linear expansion observer to observe the longitudinal force of the driving wheel can reduce the influence of external disturbances, increase the "reliability" of the data, and obtain a more accurate state value compared to direct calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 is a flow chart of a driving wheel longitudinal force estimation method based on wheel speed Kalman filter-LESO;

[0021] Figure 2 This is the dynamic force diagram of 1 / 4 of the vehicle. DETAILED DESCRIPTION

[0022] The following will be described in more detail with reference to the schematic diagram of the driving wheel longitudinal force estimation method based on wheel speed Kalman filter-LESO of the present invention, wherein the preferred embodiment of the present invention is shown, and it should be understood that those skilled in the art can modify the present invention described herein, while still achieving the advantageous effects of the present invention. Therefore, the following description should be understood as being widely known to those skilled in the art, and not as a limitation of the present invention.

[0023] Step S1: The driving wheel speed of the unmanned mine car is monitored by a wheel speed sensor to obtain the driving wheel speed of the unmanned mine car, and the driving wheel angular acceleration of the unmanned mine car is monitored by a wheel angular acceleration sensor to obtain the driving wheel angular acceleration of the unmanned mine car.

[0024] The real-time data of driving wheel speed and angular acceleration are measured, and a driving wheel speed filter is designed based on the Kalman filter (KF) to filter the driving wheel speed.

[0025] Step S1.1: Based on KF, derive a fusion of the calculated value of the state quantity containing process noise and the observed value of the state quantity containing observation noise to achieve a more accurate denoised estimated value of the state quantity.

[0026] Consider the following first-order discrete linear dynamic system:

[0027]

[0028] In formula (1), x k is the state of the system at time k, x k-1 is the state of the system at time k-1, u k-1 is the system input at time k-1, S k is the process noise of the system at time k, and its covariance is Q, z k is the observed value of the system state at time k, v k is the observation noise of the system at time k, its covariance is R, A is the state transfer matrix, B is the control input matrix, and H is the state observation matrix.

[0029] According to the KF principle, the state prediction equation of the system, also known as the prior estimate, is as follows:

[0030]

[0031] In formula (2) is the prior estimate of the system state at time k, is the estimated value of the system state at time k-1.

[0032] definition is the prior state error at time k, and the formula of this parameter is as follows:

[0033]

[0034] Definition k is the posterior state error at time k, and the formula for this parameter is as follows:

[0035]

[0036] In formula (4) is the estimated value of the state quantity at time k.

[0037] Define p k is the covariance matrix of the prior error at time k. The formula of this parameter is as follows:

[0038]

[0039] Define p k is the covariance matrix of the updated error at time k. The formula of this parameter is as follows:

[0040] p k =E[e k ,e k T ] (6)

[0041] According to the KF principle, the Kalman gain matrix K under the optimal estimation condition can be obtained: k for:

[0042]

[0043] From formula (1)(2)(3)(5) and S k The covariance of is Q, and the prior error covariance matrix is:

[0044]

[0045] According to the KF principle, the posterior estimation expression can be obtained as:

[0046]

[0047] According to equations (1), (6), and (8), the covariance matrix expression of the update error is:

[0048]

[0049] At this point, the filter establishment based on KF for the system in equation (1) has been completed;

[0050] Through the five core formulas (2), (7), (8), (9), and (10), the state quantity x is filtered and calculated, and the estimated value is

[0051] Step S1.2: Based on the five core formulas of KF derived in step S1.1, design a Kalman filter for the wheel speed of the unmanned mine vehicle.

[0052] The driving wheel speed sensor and driving wheel angular acceleration sensor installed on the unmanned mining car measure the driving wheel speed w i and the driving wheel angular acceleration α i, i represents a specific driving wheel.

[0053] The state space equation of the driving wheel speed of the unmanned mining vehicle is established based on Newton's laws of motion. The driving wheel speed and angular acceleration have the following relationship:

[0054]

[0055] The sensor can only obtain the driving wheel angular acceleration information of the previous moment, so the forward Euler method is used to discretize equation (11), as follows:

[0056]

[0057] In formula (12), is the rate of change of driving wheel speed, w i,k-1 is the speed of the driving wheel at time k-1, w i,k is the speed of the driving wheel at time k, α i,k-1 is the angular acceleration of the driving wheel at time k-1, and T is the same observation frame for both the wheel speed sensor and the angular acceleration sensor.

[0058] Based on formula (12) and the wheel speed sensor observation, the system state space equation containing process noise and observation noise of the driving wheel speed is obtained as follows:

[0059]

[0060] In formula (13), w i,k is the speed of the driving wheel at time k, w i,k-1 is the speed of the driving wheel at time k-1, α i,k-1 is the angular acceleration of the driving wheel at time k-1, A is the transfer matrix, value is I, the unit matrix, B is the control input matrix, value is T,z i,k is the observed value of the driving wheel speed at time k, H is the state observation matrix, S k ,v k are process noise and observation noise respectively.

[0061] The five formulas of KF derived in step S1.1 are used to realize the filtered observation of the driving wheel speed.

[0062] The prior prediction value of the driving wheel speed is expressed as follows:

[0063]

[0064] In formula (14), is the prior prediction value of the driving wheel speed at time k, is the estimated value of the driving wheel speed at time k-1. A and B are coefficients, which are the unit matrix I and the same observation frame T of the two sensors.

[0065] The prior error covariance matrix is ​​as shown in formula (8), and the Kalman gain coefficient is as shown in formula (7).

[0066] The posterior estimation of the driving wheel speed is expressed as follows:

[0067]

[0068] In formula (15) is the posterior estimate of the driving wheel speed at time k.

[0069] Update the covariance matrix as shown in formula (10).

[0070] At this point, the driving wheel speed filter based on Kalman filtering has been completed.

[0071] The wheel speed filter based on KF established in step S1 is used to filter and reduce the noise of the wheel speed information, which paves the way for subsequent steps.

[0072] Step S2, design a linear expansion observer of bounded disturbances in a first-order system with system input and unknown bounded disturbances, the goal is to ensure that the observer can quickly and accurately track the trajectory of the actual state quantity, and at the same time be robust to disturbances inside and outside the system. Specifically:

[0073] Step S2.1: Based on the linear expansion observer construction method, derive the linear expansion observer structure for the intermediate state quantity of the following first-order system.

[0074] For the following first-order system:

[0075]

[0076] In formula (16), x is the state of the system, u is the input of the system, y is the output of the system, and A, B and C are coefficient matrices respectively.

[0077] Next, we construct an observer. The main idea is to estimate the state quantity through the measured values ​​of the input quantity and the measured values ​​of the output quantity.

[0078] According to the general construction form of the linear expansion observer, the observer is constructed for the system of formula (16):

[0079]

[0080] In formula (17), is the differential of the estimated value of the state quantity, is the estimated value of the output, and L is the observation coefficient.

[0081] Step S2.2: Based on the unmanned mining vehicle under study, a quarter vehicle dynamics model is constructed to derive the observation expression of the driving wheel longitudinal force based on the linear expansion observer, and the driving wheel longitudinal force is regarded as an unknown and bounded disturbance.

[0082] Depend on Figure 2 According to the driving wheel torque balance, we get:

[0083]

[0084] The parameters in formula (18) are Figure 2 Correspondingly, T is the driving torque acting on the driving wheel, J is the tire moment of inertia, f is the rolling resistance coefficient, r is the tire rolling radius, and F Z is the load carried by the tire, F X is the longitudinal force of the tire.

[0085] According to formula (17), and ignoring the second term in formula (18), formula (18) can be expressed as:

[0086]

[0087] In formula (19), u=T, F X It is both the unknown input of the system and the state quantity to be estimated, and c is the coefficient matrix.

[0088] Let x1 = y, x2 = f, At this time, x2 is the observed quantity, and combined with formula (19), we can get:

[0089]

[0090] make z1 is the observed value of x1, z2 is the observed value of x2,

[0091] c=

[10] . And ignore Item, according to formula (17), construct the observer expression:

[0092]

[0093] From formula (21) and We can get:

[0094]

[0095] According to the bandwidth method, w0 is the observation bandwidth.

[0096] At this point, the longitudinal force observer of the driving wheel of the unmanned mining car based on the linear expansion observer has been completed.

[0097] The posterior estimated value of the driving wheel speed Enter into formula (22), Assign the value to z1.

[0098] The observed value z1 will gradually converge to the actual driving wheel speed, and z2 will converge to Then, the following formula is used for simple transformation and reverse calculation to estimate the longitudinal force of the driving wheel.

[0099]

[0100] Based on the above formula, the ultimate observation of the driving wheel speed of the unmanned mining vehicle is achieved.

[0101] The above is only a preferred embodiment of the present invention and does not limit the present invention in any way. Any technician in the relevant technical field, without departing from the scope of the technical solution of the present invention, makes any form of equivalent replacement or modification to the technical solution and technical content disclosed in the present invention, which does not depart from the content of the technical solution of the present invention and still falls within the protection scope of the present invention.

Claims

1. A method for estimating longitudinal force of a driving wheel based on wheel speed Kalman filter-LESO, characterized in that: The following steps are involved: Step S1, designing a driving wheel speed filter for the unmanned mining vehicle based on the Kalman filter to filter the driving wheel speed; Step S2: design a linear expansion observer of the driving wheel speed of the unmanned mining vehicle, and obtain the driving wheel longitudinal force in combination with the driving wheel speed output in step S1.

2. The driving wheel longitudinal force estimation method based on wheel speed Kalman filter-LESO according to claim 1 is characterized in that: Step S1 specifically includes the following steps: Step S1.1, based on the Kalman filter, the calculated value containing process noise and the observed value containing observation noise are fused to realize the noise reduction estimation of the state quantity; Step S1.2, designing a Kalman filter for the wheel speed of the unmanned mine vehicle; Step S1.3: Filter the driving wheel speed.

3. The driving wheel longitudinal force estimation method based on wheel speed Kalman filter-LESO according to claim 1, characterized in that: Step S2 specifically includes the following steps: Step S2.1, based on the linear expansion observer construction, derive the state observer structure of the first-order system; Step S2.2, based on the unmanned mining vehicle, build a quarter vehicle model, regard the expression related to the longitudinal force of the driving wheel as an unknown bounded disturbance, and derive the linear expansion observer expression of the disturbance; Step S2.3, input the driving wheel speed outputted in step S1 to the linear expansion observer in step S2.2, and the linear expansion observer outputs a convergence value; Step S2.4, inversely calculate the longitudinal force of the driving wheel according to the convergence value.

Images