Methods for determining the direction of travel of a motor vehicle

Abstract

Method in which the forward or reverse travel of a vehicle is determined, wherein the yaw rate of the vehicle is measured as the first measurement and at least one further measurement describing the lateral dynamics of the vehicle is measured simultaneously, characterized in that the correlation of the first and the further measurement is considered in order to determine the direction of travel within a data set of several successive measurements.

Description

[0001] The invention relates to a method according to the preamble of claim 1 and a control unit according to the preamble of claim 10.

[0002] Modern vehicles are increasingly equipped with vehicle stability control systems and / or driver assistance systems. For these systems to function reliably, knowledge of the vehicle's direction of travel is crucial.

[0003] One way to detect the direction of travel is to use special wheel speed sensors on at least one wheel, which can detect the direction of rotation. However, due to the additional effort involved (e.g., using two sensors on one encoder wheel), these sensors are expensive.

[0004] DE 10 2006 001 378 A1 discloses a method for determining the direction of travel of a vehicle equipped with at least one acceleration sensor and at least one wheel pulse generator. The expected value generated from the wheel pulse generator signal for a physical quantity representing the direction of travel, preferably the vehicle's speed, is compared with an actual value determined from the acceleration sensor signal. A similar method, which in particular detects a reverse start-up, is described in DE 10 2009 020 594 A1. Sensors for longitudinal acceleration are not usually present in motor vehicles—with the exception of off-road vehicles—so a method based on this would incur additional costs.

[0005] DE 10 2007 003 013 A1 discloses a method and a device for detecting the direction of travel of a motor vehicle, in which actual signals from at least two wheel speed sensors are acquired. The direction of travel is determined by comparing the temporal sequence of the actual signals with a target signal sequence for at least one direction of travel. A similar method is described in DE 10 2007 030 431 A1. For direction of travel detection, a reference signal sequence must be known, or information from other sensors must be used.

[0006] From EP 1 065 507 B1, a method is known for determining whether a motor vehicle is moving forward or backward. In this method, the steering wheel relative angle, the yaw rate, and the vehicle speed are measured, and a forward and a backward steering angle are calculated from the vehicle speed and yaw rate. The difference between the steering wheel relative angle and the forward or backward steering angle is then calculated and filtered. By comparing the filtered and unfiltered values ​​of the difference between the steering wheel relative angle and the forward or backward steering angle, the total forward and backward differences are determined. These steps are repeated, and the direction of movement is determined by comparing the summed magnitudes of the total forward and backward differences.

[0007] From WO 2011 / 063 280 A2, a method is known for determining the orientation of a sensor unit in a vehicle, for example, when a mobile device is to be used for navigation assistance. For this purpose, the various eigenvectors of an acceleration sensor cluster are determined, and then, based on various correlations, it is selected which of the eigenvectors is assigned to the different principal directions in the vehicle.

[0008] The object of the present invention is to enable reliable detection of a vehicle's forward or reverse movement, even without the vehicle turning. Furthermore, it is desirable that this detection can be performed using the sensors already available for vehicle stability control.

[0009] This problem is solved by the method according to claim 1 and the control unit according to claim 10.

[0010] A method is provided for determining the forward or reverse movement of a vehicle, whereby the vehicle's yaw rate is measured as the first measurement and at least one further measurement describing the vehicle's lateral dynamics is measured simultaneously. According to the invention, the correlation between the first and the further measurement is considered to determine the direction of travel within a data set of several successive measurements.

[0011] "Simultaneous" here means that the measurement of the first and the subsequent measured quantity have at most a minimal time offset, which is preferably at least a factor of 10 less than the time offset of two successive measurements of the same measured quantity.

[0012] Conventional vehicle stability control systems often incorporate a yaw rate sensor to measure the yaw rate, allowing this initial measurement to be reliably determined without additional effort. By considering a data set comprised of multiple consecutive measurements, short-term external disturbances have a negligible impact on the measured values. By examining the correlation between the initial and subsequent measurements, random fluctuations can be significantly suppressed. Thus, even minor steering corrections unconsciously made by a driver while driving straight ahead provide suitable information for longitudinal direction detection, i.e., distinguishing between forward and reverse travel.

[0013] Advantageously, the further parameter describing the vehicle's lateral dynamics is proportional to the steering angle, lateral acceleration, or wheel speed difference of two wheels on the same axis. Wheels on the same axis here means that the wheels under consideration are preferably the left and right front wheels or the left and right rear wheels of the vehicle, or have a negligible longitudinal offset between their axes of rotation. The aforementioned parameters are measured using sensors that are often already present, so no additional costs are incurred. Steering angle, lateral acceleration, and wheel speed difference are independent of whether the vehicle is traveling forwards or backwards when cornering.In contrast, the sign of the yaw rate changes when traversing the same curve, depending on whether this is done in a forward or reverse direction. Therefore, by examining the correlation between the yaw rate and one of the other measured variables, a distinction can be made between forward and reverse travel. If more than one of the aforementioned measured variables is taken, the correlation coefficients are calculated redundantly, enabling plausibility checks and thus further increasing the reliability of longitudinal direction detection.

[0014] Preferably, the size of the data set, i.e., the number of consecutive measurements, is determined as a function of the magnitude of the yaw angle acceleration. By adaptively selecting the data set size, optimal longitudinal direction detection is ensured under the respective driving conditions.

[0015] Preferably, for each data set, the mean, in particular the arithmetic mean, of the first measurement and the mean, in particular the arithmetic mean, of the second measurement, as well as the deviations of the individual measurements of the first and the second measurement from their respective means, are calculated. Averaging reduces random fluctuations in the measurements.

[0016] The correlation coefficient between the deviations of the two measured quantities is particularly favored. Also known as the correlation coefficient, this value quantifies the relationship between the two measurements and can be evaluated with manageable computational effort.

[0017] A forward movement is determined to be particularly preferred if the magnitude of the correlation coefficient exceeds a first limit value and the correlation coefficient is positive, if a counterclockwise rotation corresponds to a positive yaw rate, and if the other measured quantity is positive for a left turn.

[0018] Since two uncorrelated quantities have a correlation coefficient of almost zero, the existence of a correlation between the two measurements is checked by comparing the magnitude of the correlation coefficient with a suitably chosen limit or threshold value. Then, with an appropriate choice of sign convention, forward or reverse movement can be reliably identified.

[0019] Furthermore, a reverse movement is determined to be particularly preferred if the magnitude of the correlation coefficient exceeds a first limit value and the correlation coefficient is negative, provided that a counterclockwise rotation corresponds to a positive yaw rate and the other measured quantity is positive in the case of a left turn.

[0020] It is advantageous if the determined direction of travel is validated using information on the actuation of a drive clutch and / or vehicle speed.

[0021] By using additional independent information, the reliability of the detection can be further increased.

[0022] It is particularly advantageous to accept a change of direction as plausible only if the vehicle speed has fallen below a speed limit within a specified period before the change of direction was determined, and in particular if the drive clutch has been activated.

[0023] If the vehicle has information about the selected gear, this can also be used for plausibility checks. Reversing can be conveniently detected based on the correlation coefficient when a particularly low threshold for the magnitude of the correlation coefficient is exceeded, provided no gear or reverse gear is engaged. Furthermore, it is advantageous if, after the drive clutch is engaged with first gear or reverse gear engaged, or with the clutch disengaged, a change in direction of travel is detected when a particularly low threshold for the magnitude of the correlation coefficient is exceeded.

[0024] The invention further relates to a control unit of an electronically controlled braking system of a motor vehicle, which performs a method according to at least one of the preceding claims.

[0025] Further preferred embodiments will become apparent from the dependent claims and the following description of an exemplary embodiment with reference to figures.

[0026] They show Fig. 1 a vehicle suitable for carrying out the method according to the invention, Fig. 2 a scheme of a procedure for detecting forward or reverse travel in a curve, and Fig. 3 A scheme of the inventive method for detecting forward or reverse travel even when driving straight ahead.

[0027] Fig. Figure 1 shows a schematic representation of a motor vehicle 1 suitable for carrying out the method according to the invention. It has a drive motor 2 that drives at least some of the vehicle's wheels, a steering wheel 3, a brake pedal 4 connected to a tandem master cylinder (THZ) 13, and four individually controllable wheel brakes 10a-10d. The method according to the invention is also feasible if only some of the vehicle's wheels are driven. In addition to or as an alternative to hydraulic friction brakes, electromechanically actuated friction brakes can also be used as wheel brakes on one, several, or all wheels. According to an alternative embodiment of the invention, the vehicle has an electric drive, and the braking torque at at least one wheel is generated at least partially by the electric machine(s) operating as a generator.

[0028] For the detection of vehicle dynamics, a steering wheel angle sensor 12 is used to measure the steering angle δ, and four wheel speed sensors 9a - 9d are used to measure the rotational speeds V. i of the individual wheels, a lateral acceleration sensor 5 for measuring the lateral acceleration a Lat A yaw rate sensor 6 for measuring the yaw rate Ψ̇ and at least one pressure sensor 14 for measuring the brake pressure p generated by the brake pedal and THZ are provided. The pressure sensor 14 can also be replaced by a pedal travel or pedal force sensor if the auxiliary pressure source is arranged such that a brake pressure applied by the driver is indistinguishable from that of the auxiliary pressure source, or if an electromechanical brake actuator with a known relationship between pedal position and brake torque is used. The signals from the wheel sensors are fed to an electronic control unit (ECU) 7, which uses predefined criteria from the wheel rotational speeds V to determine the appropriate braking force.i the vehicle speed V Ref determined.

[0029] The electronic control unit (ECU) 7 receives data from the sensors described above, as well as from any other sensors that may be present, and controls the hydraulic control unit (HCU) 8 to enable the build-up or modulation of brake pressure in the individual wheel brakes independently of the driver. Additionally, the drive torque currently generated by drive motor 2 and the torque requested by the driver are determined. These can also be indirectly determined values, which are derived, for example, from an engine map and transmitted to the electronic control unit 7 via an interface 11, e.g., a CAN or FlexRay bus, from the engine control unit (not shown).

[0030] The handling characteristics of motor vehicle 1 are significantly influenced by the chassis design, with factors such as wheel load distribution, suspension elasticity, and tire properties determining the vehicle's self-steering behavior. In certain driving situations, characterized by a predetermined, desired curve radius and the coefficient of friction between the tires and the road surface, a loss of driving stability can occur, and the steering response desired by the driver cannot be achieved with the given chassis design. The available sensors can detect the driver's input and monitor its implementation by the vehicle. Preferably, the tendency toward a loss of stability is detected.

[0031] For example, the yaw rate is controlled using a vehicle model to prevent the vehicle from skidding. For the control algorithms or driver assistance systems to function correctly, it is necessary to know the vehicle's direction of travel.

[0032] This detection of a two-lane vehicle reversing can be based on known relationships between various measured or determined quantities: In one approach, the differences in wheel speeds on an axle are compared with the yaw rate (Ψ̇) of the vehicle. Yaw rate, or yaw angular velocity, is defined here as the rate of change of the yaw angle about the vertical axis through the vehicle's center of gravity. The sign of the difference in wheel speeds is the same for forward and reverse travel. However, the sign of the yaw rate (Ψ̇) changes depending on the direction of travel.

[0033] The following discussion uses a right-handed Cartesian coordinate system, assuming that the yaw rate Ψ̇ is positive when moving in the "counterclockwise" direction. Then, in a left turn, the yaw rate Ψ̇ is positive when moving forward (Ψ̇>0) and negative when moving backward (Ψ̇<0), while in a right turn, the yaw rate Ψ̇ is negative when moving forward (Ψ̇<0) and positive when moving backward (Ψ̇>0).

[0034] In contrast, when making a left turn, the speed V is the same both when driving forwards and backwards. L The left wheel (on the inside of the curve) is smaller than the right wheel: VL−VR<0

[0035] Accordingly, the speed V is different when turning right. R The speed of the right wheel is smaller than that of the left wheel, both when driving forwards and backwards: VL−VR>0

[0036] In Fig. Figure 2 shows a scheme of a procedure for detecting forward or reverse travel, which is based on the relationships discussed above during cornering.

[0037] At the beginning, in step 20, the wheel speeds V are measured. L and V R of the left and right wheels on a vehicle axle, as well as the yaw rate Ψ̇.

[0038] Then, in step 21, the difference Δ = V is calculated. L - V R The speed of the left and right wheels is calculated.

[0039] If in step 22 it is determined that the magnitude of the difference |Δ| does not exceed a predetermined limit δ1, then steps 20 to 22 are repeated.

[0040] If the condition |Δ| > δ1 is met, step 23 checks whether the magnitude of the yaw rate also exceeds a predefined limit, |Ψ̇| > δ2. If this is not the case, steps 20 to 23 are repeated.

[0041] Then, in step 24, it is checked whether the wheel speed difference Δ has a positive or a negative sign, i.e., whether it is a right or a left turn.

[0042] If the sign of Δ is positive, the sign of the yaw rate is considered in step 25. If Ψ̃ > 0, a reverse movement and a right turn are detected and / or output in step 26.

[0043] If, on the other hand, Ψ̃ < 0, then a forward movement and a right turn are detected and / or output in step 27.

[0044] If the sign of Δ is negative, the sign of the yaw rate is checked in step 28.

[0045] If Ψ̃ > 0, then in step 29 a forward movement and a left turn are detected and / or output.

[0046] If, on the other hand, Ψ̇ < 0, then a reverse movement and a left turn are detected and / or output in step 30.

[0047] Of course, the verification of the conditions can also be carried out in a different order.

[0048] An alternative embodiment of the method described above is based on a measurement of lateral acceleration and is therefore feasible if the vehicle has a lateral acceleration sensor. When switching from forward to reverse travel or vice versa, the sign of the lateral acceleration a changes. Lat No. So if it is agreed that the lateral acceleration a LatIf the lateral acceleration in a right-hand curve has a positive sign, then the sign of the lateral acceleration in a left-hand curve is negative in both forward and reverse driving. This allows for a differentiation between reverse and forward driving based on the Fig. The procedure described in step 2 is used. Instead of measuring wheel speeds in step 20, the lateral acceleration is measured, thus eliminating the need for step 21. Accordingly, step 22 checks whether the lateral acceleration exceeds a minimum value, and step 24 considers the sign of the lateral acceleration. Therefore, at a Lat If the sign is > 0 and the Ψ̇ > 0, a left turn driven in the forward direction can be detected; for the other cases, detection occurs accordingly. In principle, the method can be applied based on any other measured variable whose sign depends on a right or left turn, but not on forward or reverse travel.

[0049] The methods considered so far have the disadvantage that a curve is required to detect the longitudinal direction.

[0050] In Fig. Figure 3 shows a preferred embodiment of the method according to the invention schematically, which can also distinguish between forward and reverse travel during straight-ahead driving.

[0051] First, in step 31, the wheel speeds V are determined. Li of the left and V Ri of the right wheel on a vehicle axle and the yaw rate ψ i measured.

[0052] Then, in step 32, the difference Δ i = V Li -V Ri calculated the rotational speed of the left and right axle-aligned wheels.

[0053] The measurements and calculations of steps 31 and 32 are repeated until the verification in step 33 shows that a data set of N yaw rates and wheel speed differences has been obtained. Short time intervals Δt of constant duration are preferably taken between the individual measurements.

[0054] According to a preferred embodiment of the invention, a data set has a constant length N. An alternative preferred embodiment of the invention selects the length of a data set adaptively depending on the yaw rate ψ̇. i and yaw acceleration ∂ψ˙i∂t. Within a data set, the yaw rate changes as follows: |ΔΨ|=∑i=1NΔti∗|∂ψ˙i∂t|

[0055] For constant time intervals, one obtains: |ΔΨ|=N∗Δt∗|∂ψ˙i∂t|

[0056] To prevent errors caused by excessively changing yaw rate, the condition ΔΨ̇ << Ψ̇ must be satisfied, which limits the maximum length of a data set: N<<ψ˙Δt*|∂ψ˙i∂t|

[0057] Thus, the length of a data set can be determined by examining the instantaneous change in the yaw rate. |∂ψ˙i∂t| chosen, for example by ending the measurement when it exceeds a predetermined threshold.

[0058] Once a data set has been recorded, the calculation of the mean values ​​of yaw rate and wheel speed difference takes place in step 34: Ψ˙¯=∑i=1Nψ˙iN Δ¯=∑i=1NΔiN

[0059] Then, in step 35, the correlation coefficient R is calculated. ΔΨ calculated between the deviations of the yaw rate of a data point from the mean or the wheel speed difference of a data point from the mean: RΔΨ=∑i=1N(ψ˙i−ψ˙¯)⋅(Δi−Δ¯)N⋅A^⋅B^A=(ψ˙i−ψ˙¯)A^=AB=(Δi−Δ¯)2B^=B

[0060] If in step 36 it is determined that the amount of R ΔΨ If a predefined threshold value δ3 is not exceeded, meaning there is no discernible correlation, a data set is measured again.

[0061] If the data set for direction detection is missing due to |R ΔΨ If δ3 is usable, the sign is checked in step 37.

[0062] Provided the correlation coefficient has a positive sign (R ΔΨ >0), is detected in step 38 when reversing.

[0063] Otherwise (R ΔΨ <0) is detected in step 39 when moving forward.

[0064] If the vehicle is equipped with a lateral acceleration sensor, the correlation coefficient R can be used alternatively or additionally. aΨ between a Latand Ψ̇ are used for direction of travel detection, with the algorithm consisting of Fig. 3 will be adjusted accordingly.

[0065] In step 31, the lateral acceleration a is used instead of the wheel speeds. Lat measured, whereupon step 32 can be omitted. Step 34 accordingly involves calculating the average of the lateral acceleration. Lat In the following steps, the correlation coefficient R will be calculated. aΨ between lateral acceleration a Lat and yaw rate Ψ̇ calculated (step 35), checked for sufficient correlation (step 36), and then used the sign of the correlation coefficient to check for backward correlation (R aΨ > δ3) or forward travel (R aΨ <- δ3) recognized.

[0066] If a steering angle sensor is present, a further possibility for detecting a reverse movement is a correlation coefficient R. αΨto use between steering angle α and yaw rate Ψ̇. Here the steps are in Fig. 3. Adjusted accordingly. Of course, the order of the individual steps can also be changed (where appropriate).

[0067] Further improvements in longitudinal stability can be achieved by calculating the correlation coefficient between yaw rate and more than one other measured variable. For example, if the correlation coefficient between yaw rate and steering angle, as well as the correlation coefficient between yaw rate and lateral acceleration, are calculated, these results can be checked for consistency. If all three correlation coefficients are calculated, forward or reverse movement can be determined by a majority decision.

[0068] Another plausibility check is possible when information about clutch operation is available. This information can be considered together with information about the vehicle speed to detect a change in direction. Such a plausibility check is based on the correlation that a change in direction from forward to reverse or vice versa only occurs at low vehicle speeds and, in particular, after clutch operation.

[0069] If information is available about the selected gear and, in particular, the clutch operation, this can also be used for plausibility checks. For example, it can be assumed that a change in longitudinal direction can occur, especially after clutch operation, if first gear, no gear, or reverse gear is engaged.

[0070] Naturally, one or more of the measured variables under consideration can also be filtered in a suitable manner to further increase the reliability of detecting forward or reverse travel.

Claims

[1] Method in which the forward or reverse movement of a vehicle is determined, wherein the yaw rate of the vehicle is measured as the first measurement and at least one other measurement describing the lateral dynamics of the vehicle is measured simultaneously, characterized by , that to determine the direction of travel within a data set of several consecutive measurements, the correlation of the first and subsequent measurements is considered. [2] Method according to claim 1, characterized by , that the further quantity describing the lateral dynamics of the vehicle is proportional to the steering angle or the lateral acceleration or the wheel speed difference of two identical vehicle wheels. [3] Method according to claim 1 or 2, characterized by , that the size of the data set, i.e. the number of consecutive measurements, is determined as a function of the magnitude of the yaw angle acceleration. [4] Method according to at least one of claims 1 to 3, characterized by , that for each data set the mean, in particular the arithmetic mean, of the first measurement and the mean of the further measurement as well as the deviations of the individual measurements of the first and the further measurement from the respective means are calculated. [5] Method according to claim 4, characterized by , that a correlation coefficient is calculated between the deviations of the two measured quantities. [6] Method according to claim 5, characterized by , that forward motion is determined when the magnitude of the correlation coefficient exceeds a first limit value and the correlation coefficient is positive when a counterclockwise rotation corresponds to a positive yaw rate and the further measured quantity is positive in the case of a left turn. [7] Method according to claim 6, characterized by, that a reverse movement is determined if the magnitude of the correlation coefficient exceeds a first limit value and the correlation coefficient is negative if a counterclockwise rotation corresponds to a positive yaw rate and the further measured quantity is positive in the case of a left turn. [8] Method according to at least one of the preceding claims, characterized by , that the determined direction of travel is validated based on information about the actuation of a drive clutch and / or vehicle speed. [9] Method according to claim 8, characterized by , that a change of direction is only accepted as plausible if the vehicle speed has fallen below a speed limit within a specified period before the change of direction is determined and, in particular, an actuation of the drive clutch has been detected. [10] Control unit of an electronically controlled braking system of a motor vehicle, characterized by that this carries out a process according to at least one of the preceding claims.

Images