METHOD FOR ESTIMATING AND ADJUSTING THE SPEED AND ACCELERATION OF A VEHICLE

Description

Technical field

[0001] The invention relates to the field of motor vehicles. It relates more particularly to a strategy for estimating speed and associated acceleration, ranging from high speeds to very low speeds, while overcoming the limitations of current sensors. State of the art

[0002] In the context of developing control laws, knowledge of precise speed and associated acceleration are very important. For example, control laws used in ADAS (Advanced Driver Assistance Systems) and autonomous vehicles always require speed and acceleration information.

[0003] On current vehicles, speed and acceleration are already accurately calculated above a certain threshold. If the actual speed is below this threshold, speed and acceleration information is not available. This speed range is commonly referred to as "low speed."

[0004] The main problem is that, due to the limitations of the sensors used, the speed cannot be well estimated below the said speed threshold.

[0005] Therefore, the control laws used cannot robustly control the different low-speed systems; for example: Parking systems (known in English as HFPB - Hands Free Parking Brake and auto-park) ACC systems (distance regulator) for "Stop&Start" situations The autonomous car or TJP system (for Traffic Jam Pilot) in traffic jam situations.

[0006] A second problem is the use of the acceleration value from the accelerometer. This value is not very accurate (it has offsets) because of: The position of the accelerometer after factory installation Some external quantities such as the slope and the tilt of the road The roll and pitch of the body.

[0007] There figure 1 represents a graph illustrating the problem encountered. The currently estimated speed of the vehicle is represented by curve 1, the accelerometer value is represented by curve 2, curve 3 represents the "peaks" of the encoder wheels (i.e. the signal peaks they send when passing a tooth, such a wheel being also called a "toothed wheel") and between lines A and B the low speed zone where one is driving below a threshold of 1 km / h. The peaks indicate whether the wheels have turned or not and give an image of the speed by their amplitude.

[0008] In the area between lines A and B, the speed is unknown. For example, in the right part of the low-speed area, we can see that the wheels are turning (presence of peaks in the encoder wheels), but no speed is detected below the 1 km / h threshold.

[0009] Finally, the accelerometer plot (curve 2) shows a shift in the low speed zone (zone without the presence of peaks and constant accelerometer with non-zero value).

[0010] Thus, it becomes necessary to develop a strategy for estimating speed and acceleration in the low speed zone (between A and B), in addition to the speed value already present in the car. An example of such a strategy is known from the document "Improving the Response of a Wheel Speed ​​Sensor by Using a RLS Lattice Algorithm" by W. Hernandez, published in Sensors of June 2006, pages 64-79.

[0011] This document discloses more particularly the use of an adaptive filter using the recursive least squares algorithm (in English " RLS ”) to process a signal relating to the speed of a vehicle.The main advantage of this type of software solution based on adaptive filtering lies in its low cost.

[0012] Also known from US2017 / 219381 A1 is a rotor assembly providing improved low speed sensing, employed by a gyroscope for spacecraft attitude control.

[0013] However, a more significant problem remains beyond the estimation of the speed, namely the discontinuity of the estimated speed and acceleration values ​​when moving from the high speed range, located above the threshold, to the low speed range located below the threshold. The objective of the present invention is in particular to solve this technical problem by proposing a method for estimating the speed and / or acceleration of a vehicle at low speed while being suitable for the precise measurement of the vehicle speed at medium and high speed, without presenting any discontinuity of these values. Description of the invention

[0014] To this end, the invention relates to a method for estimating the speed of a motor vehicle in which: A first speed threshold SV1 is defined corresponding to a minimum speed value provided by a wheel angular speed sensor of the vehicle; A second speed threshold SV2 is defined, greater than SV1; Low speed values ​​are estimated when the vehicle is traveling below SV1 using an adaptive filter type estimation method; High speed values ​​are measured when the vehicle is traveling above SV2 using vehicle speed values ​​provided by the wheel angular speed sensors; In the intermediate zone between SV1 and SV2, a mixing of high speed with low speed is carried out.

[0015] According to the invention, three speed ranges are used: low speed, high speed and an intermediate mixing zone. Using a mixing range makes it possible to avoid discontinuities in speed or acceleration (essential to guarantee the stability of the control laws).

[0016] Advantageously, the adaptive filter is a Kalman filter.

[0017] Advantageously, in the intermediate zone between SV1 and SV2, the mixing is carried out periodically at successive times using a linear mixing method following the formula: vitesse = vitesse kalman basse × SV 2 − vitess e t − 1 SV 2 − SV 1 + vitesse v é hicule haute × vitess e t − 1 − SV 1 SV 2 − SV 1 This linear mixing allows the calculation of the value of the mixed speed ( speed ) using the velocity values ​​from the Kalman method ( vitesse kalman basse ) and vehicle speed ( vitesse v é hicule haute ) Vehicle speed ( vitesse v é hicule haute ) is the speed measured using the angular velocity of the wheels. The speed is the mixed speed at the current time t, speed t- 1 is the mixed speed at the previous mixing time t-1, ( vitesse kalman basse ) is the speed value calculated by the Kalman method at the current time t and ( vitesse v é hicule haute ) is the speed value measured by the angular sensor at the current time t.

[0018] According to a characteristic of the invention, the first threshold SV1 can be 1 km / h.

[0019] According to another characteristic of the invention, the second speed threshold SV2 can be 1.5 km / h.

[0020] Advantageously, in step C) we also estimate the acceleration value using the Kalman filter and in step E) we also mix the acceleration values ​​between SV1 and SV2.

[0021] An advantage of the invention is that the speed is estimated without discontinuity and the value of the associated acceleration can also be taken into account. Brief description of the figures

[0022] The invention will be better understood by reading the following description, of an exemplary embodiment given for illustrative purposes, the description making reference to the appended drawings among which: there figure 1 represents a graph illustrating the problems encountered at low speeds; the figure 2 schematically illustrates the principle of the invention; the Figure 3 and 4 represent examples of results of the use of the method of the invention during the take-off and stopping phases of a vehicle. Detailed description

[0023] There figure 2 schematically represents the general principle of the invention in which 3 speed zones are defined: a low speed zone, below a first threshold SV1 below which the speed and acceleration values ​​are not available. Typically 1 km / h. In this zone, the speed is measured using the Kalman method. This method is known per se to those skilled in the art, but it is recalled below for greater clarity in the description of the invention. A high speed zone above a second threshold SV2 higher than the first threshold SV1, for example 1.5 km / h. In this zone the speed and acceleration values ​​are provided by the vehicle sensors; and a mixing zone located between the two thresholds SV1 and SV2. I. Velocity estimation with the Kalman method I. 1 Classical Kalman filter.

[0024] A Kalman filter takes into account three state variables [x]: x(1) distance traveled since the last instant t; x(2) speed information; x(3) acceleration information.

[0025] The two sensor measurements [z] used for state variable estimation are: z(1) the average of the peaks of the encoder wheels (WT). The peak signals of the 4 wheels are already present in the vehicle's messaging (CAN). This information allows us to have an idea of ​​the displacement of each wheel by counting at each sampling step how many encoder teeth have passed (typically 48 teeth) z(2) the angular speeds of the wheels (WS). The angular speed signals of the 4 wheels are already present in the vehicle's messaging (CAN). The average of the rear wheels will be used in the Kalman (non-driven wheel axis, i.e., less slip during take-off phases)

[0026] The system of equations for the Kalman filter is: 1) Prediction x ^ k k − 1 = F k x ^ k − 1 k − 1 + B k u k − 1 P k k − 1 = F k P k − 1 k − 1 F k T + Q k 2) Correction y ˜ k = z k − H k x ^ k k − 1 S k = H k P k k − 1 H k T + R k K k = P k k − 1 H k T S k − 1 x ^ k k = x ^ k k − 1 + K k y ˜ k P k k = I − K k H k P k k − 1

[0027] There rating used is as follows: x : system state (vector) z : sensor measurements (vector) P : covariance matrix of the estimate F k : state transition matrix uk : command entry B k : command transition matrix H : transition matrix of the measure Q : model noise covariance matrix (accuracy) R : measurement noise covariance matrix (precision) I : identity matrix x̂ : estimated value of variable x x̃ : measured value of variable x

[0028] Noticed : In the implemented Kalman filter the vector u is zero, which simplifies the first equation. I.2 Speed ​​estimation

[0029] At the system input there are two sensor data corresponding to the wheel speeds (WS) and the peaks of the encoder wheels (WT). These data are processed (DP block: "Data processing") then pass into the Kalman filter (block "Estimation") which produces a speed and an acceleration. Step One - “Data Processing”:

[0030] Top of the wheels: The encoder sends the position of the last tooth seen. We will use this increment of the number of teeth [ WT ] during a system sampling step [ You ] (not necessarily at the same rate as the sensor recording). Then the average value between the 4 wheels will be used as a measurement of the [ WT ] . The equivalent value at a linear speed and using the wheel peaks is 2 πR nb _ pic ∗ WT with [ R ] the radius of the wheels assumed to be constant and known (adjustment parameter) and [ nb_pic] the number of encoder teeth. WS: The average of the angular speeds of the rear wheels (non-driven wheel axle, i.e., less slip during take-off phases) will be used in the Kalman filter. This angular speed [WS] will be converted into linear speed from: 2πR60∗WS Second step: “Estimate”:

[0031] The Kalman model used is as follows: Equation of state:

[0032] x k = d k v k a k = d k − 1 + Te * v k − 1 + Te 2 2 * a k − 1 v k − 1 + Te * a k − 1 a k − 1 dk, vk And ak are respectively the distance traveled, the speed and the acceleration at iteration k of the filter Data input vector (“Input data vector” in English):

[0033] z k = 2 πR nb _ pic ∗ WT 2 πR nb _ pic ∗ Wheel _ Top + 2 πR 60 ∗ WS 1 Te ∗ 2 πR 60 ∗ WS nb_pic = 96, the number of encoder increments. You = 0.01 s , the sampling period.

[0034] The equation of state represents the first line of the prediction step shown previously. The assumption made here is a constant evolution of the acceleration.

[0035] The input vector [ z ] corresponds to the insertion of sensor data into the Kalman filter. The data [ WT ] is the sum of the peaks of the encoder wheels divided by four (the number of wheels). The variable [ WS ] it is equal to the sum of the speed of the rear wheels of the vehicle divided by 2.

[0036] Since this latter data is not always available (falls to 0 below SV1), an adaptation of the matrix H (see the system of Kalman equations, phase correction) in the Kalman filter was carried out. II. Speed ​​Mixing

[0037] In the speed zone located between the first threshold SV1 and the second threshold SV2, between 1km / h and 1.5km / h in the example shown figure 2 , a mix of high speed with low speed is achieved.

[0038] More specifically, the mixing can be carried out using a linear mixing method following the formula: vitesse = vitesse kalman basse × SV 2 − vitess e t − 1 SV 2 − SV 1 + vitesse véhicule haute × vitess e t - 1 − SV 1 SV 2 − SV 1

[0039] This linear mixing allows the calculation of the value of the mixed speed ( speed ) using the velocity values ​​from the Kalman method ( vitesse kalman basse ) and vehicle speed ( vitesse véhicule haute )

[0040] Vehicle speed ( vitesse v é hicule haute ) is the speed calculated using the angular speed of the wheels. At low speed the value of the vehicle's high speed is not available.

[0041] To ensure proper mixing, the reference velocity value used is the last velocity value. This value is used to define the weight of each velocity (weight defined between the relative distance to the thresholds). For example, the velocity weight of the Kalman method is defined as ( SV 2 − vitess e t − 1 SV 2 − SV 1 ). t-1 (last value) t=0 (current value) Mixed speed speed t- 1 speed Estimated speed (using the Kalman method) Not used vitesse kalman base Vehicle speed (using angular velocities) Not used vitesse v é hicule haute

[0042] Choosing the reference speed ensures continuity during mixing.

[0043] Using the estimated speed with the Kalman method is not possible because the initial value may be greater than SV2 (due to filter delay). Using vehicle speed is also not possible because it has a discontinuity at low speeds where angular velocity is no longer available. III. Examples of Results Obtained III.1 - Take-off phase

[0044] There figure 3 represents results obtained during the vehicle take-off phase. The green curve corresponds to the speed of the current system, curve 3 shows the WT, curves 4 and 5 show respectively the speed and acceleration calculated according to the method of the invention.

[0045] Concerning the speed, we see that the Kalman filter proposes an increasing speed 4 which joins the speed of the vehicle 1 currently used. The speed 4 calculated by the method of the invention takes off at the level of the first wheel peak detected, that is to say, the first peak of curve 3.

[0046] Regarding acceleration 5, we can make the same observation. The new estimate starts at the first detected peak and converges quite well towards a value that corresponds to that expected for speed 4.

[0047] The gray region corresponds to the transition region between low and high speed. It can be seen that there is no discontinuity and that the estimated speed value presents a consistent transition with respect to the dynamics of the actual vehicle speed. III. 2 In the braking phase until stopping

[0048] There Figure 4 represents results obtained in the stopping phase.

[0049] Looking at the speed, we see that curve 4 follows a speed profile more closely in accordance with the encoder wheel peaks than curve 1. We also detect the stopping of the vehicle more cleanly with the method of the invention.

[0050] Acceleration 5 seems to correspond to the proposed speed 4 and stops at the same time as speed 4.

[0051] The gray region corresponds to the transition region between high and low speed. It can be seen that there is no discontinuity and that the speed value 4 estimated by mixing presents a coherent transition with respect to the dynamics of the Kalman speed.

Claims

1. Method for estimating the speed of a motor vehicle, wherein: A. A first speed threshold SV1 is defined, corresponding to a minimum speed value provided by a vehicle angular wheel speed sensor; B. A second speed threshold SV2 is defined, which is greater than SV1; C. Low speed values are estimated when the vehicle is traveling below SV1 using an adaptive filter estimation method; D. High speed values are measured when the vehicle is traveling above SV2 using vehicle speed values provided by the angular wheel speed sensor; E. In the intermediate zone between SV1 and SV2, high speed is mixed with low speed.

2. Method according to claim 1, wherein the adaptive filter is a Kalman filter.

3. Method according to claim 2, wherein, in the intermediate zone between SV1 and SV2, the mixing is carried out periodically at successive times using a linear mixing method according to the formula: speed = speed kalmam low × SV 2 − speed t − 1 SV 2 − SV 1 + speed vehicle high × speed t − 1 − SV 1 SV 2 − SV 1 where speed is the mixed speed at the current time t, speedt-1 is the mixed speed at the previous mixing time t-1, ( speed kalman low ) is the speed value calculated by the Kalman method at the current time t and ( speed vehicle high ) is the speed value measured by the angular sensor at the current time t.

4. Method according to one of claims 1 to 3, wherein the first threshold SV1 is 1km / h.

5. Method according to one of claims 1 to 4, wherein the second speed threshold SV2 is 1.5km / h.

6. Method according to one of claims 2 to 5, wherein, in step C), the value of the acceleration is also estimated by means of the Kalman filter and, in step E), the values of the acceleration between SV1 and SV2 are also mixed.