METHOD AND CONTROL UNIT FOR MONITORING THE LANE OF A VEHICLE

Abstract

A method and a control unit for monitoring the lane of a vehicle, including the steps of ascertaining at least one lane characteristic, ascertaining at least one driving situation variable representing the instantaneous driving situation of the vehicle in an instantaneous position, as well as ascertaining at least one approach variable in a subsequent position of the vehicle. The approach variable is ascertained from the at least one lane characteristic, as well as from the at least one driving situation variable.

Claims

1-14. (canceled)

15. A method for monitoring the lane of a vehicle, the method comprising: ascertaining at least one lane characteristic; ascertaining at least one driving situation variable representing the instantaneous driving situation of the vehicle in an instantaneous position; and ascertaining at least one approach variable in a subsequent position of the vehicle from the at least one lane characteristic and from the at least one driving situation variable.

16. The method of claim 15, further comprising: comparing the at least one approach variable to a threshold value; and outputting an information variable as a function of the comparison of the approach variable to the threshold value.

17. The method of claim 15, wherein the approach value is ascertained predictively, and wherein a prediction length or a prediction time is used as a parameter for the predictive determination of the at least one approach variable.

18. The method of claim 17, wherein the prediction length and the prediction time are linked to one another via the ego-velocity of the vehicle.

19. The method of claim 18, wherein one of the prediction length and the prediction time is variably settable.

20. The method of claim 15, wherein the at least one approach variable is ascertained as a function of the instantaneous distance of a coordinate system of the vehicle at an instantaneous position relative to at least one of the coordinate systems of a roadway boundary, of the instantaneous angle of the coordinate system of the vehicle at the instantaneous position relative to at least one of the coordinate systems of the roadway boundary, of the curvature of at least one roadway boundary, and of the curvature change of at least one roadway boundary as lane characteristics.

21. The method of claim 15, wherein the at least one approach value is ascertained as a function of the instantaneous curvature of a vehicle trajectory, and of the instantaneous curvature change of the vehicle trajectory as a driving situation variable, and wherein the instantaneous curvature is a function of the instantaneous velocity and of the instantaneous yaw rate of the vehicle.

22. The method of claim 15, wherein at least two approach variables are determined, wherein a first approach variable is assigned to a first roadway boundary and a second approach variable is assigned to a second roadway boundary.

23. The method of claim 22, wherein the smaller approach variable of the first approach variable and the second approach variable is ascertained, and the ascertained smaller approach variable is compared to the threshold value.

24. The method of claim 22, wherein the first approach value and the second approach value are provided as output variables to a lane-keeping system and are used as input variables for this system.

25. The method of claim 17, wherein the at least one lane characteristic is detected with the aid of at least one sensor, the sensor including a detection range, and the prediction length is altered.

26. A control unit for monitoring the lane of a vehicle, comprising: a first subunit for ascertaining at least one lane characteristic; a second subunit for ascertaining at least one driving situation variable representing the instantaneous driving situation of the vehicle in an instantaneous position; and a third subunit for ascertaining at least one approach variable in a subsequent position of the vehicle from the at least one lane characteristic and from the at least one driving situation variable.

27. The control unit of claim 26, wherein in the third subunit the at least one approach variable is compared to a threshold value, further comprising: an output unit to output an information variable as a function of the comparison of the approach variable to the threshold value.

28. The control unit of claim 26, wherein in the third subunit, two approach variables are ascertained, the smaller approach variable of the first approach variable and of the second approach variable is ascertained in an intermediate unit, the ascertained smaller approach variable is compared to the threshold value, further comprising: an output unit to output an information variable as a function of the comparison of the approach variable to the threshold value.

29. The method of claim 17, wherein the prediction length and the prediction time are linked to one another via the ego-velocity of the vehicle, in particular, while taking the wheelbase into consideration.

30. The method of claim 18, wherein one of the prediction length and the prediction time is variably settable, in particular, by a driver of the vehicle.

31. The method of claim 17, wherein the at least one lane characteristic is detected with the aid of at least one sensor, the sensor including a detection range, and the prediction length is altered, in particular, reduced, as a function of the expanse of the detection range of the sensor.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0026] FIG. 1 shows the sequence of a method according to the present invention.

[0027] FIG. 2 illustrates variables for characterizing the driving situation based on a depiction of a vehicle on a roadway.

[0028] FIG. 3 shows a control unit for carrying out the method according to the present invention in a first embodiment.

[0029] FIG. 4 shows a control unit for carrying out the method according to the present invention in a second embodiment.

DETAILED DESCRIPTION

[0030] A vehicle is depicted in FIG. 2 in instantaneous position 201 as well as in a subsequent position 202. The vehicle in subsequent position 202 corresponds to vehicle 201 at a later point in time. Expected variables of the vehicle in subsequent position 202 are predicted based on the vehicle in instantaneous position 201.

[0031] Vehicle 201 moves along a trajectory 205. Each vehicle 201, 202 has a separate coordinate system 206, 207. Trajectory 205 of host vehicle 201 is described at each point by curvature κ.sub.act instantaneously present at this point, as well as by the change of curvature κ.sub.act instantaneously present at this point. The change of curvature κ.sub.act is a curvature change along trajectory 205.

[0032] The lane in which the vehicle moves is bound by boundaries 203a,b. The right and left boundaries each exhibit a curvature κ.sup.RI and a curvature κ.sup.LE . In addition, boundaries 203a,b may be described based on the change of their respective curvature {dot over (κ)}.sup.LE and {dot over (κ)}.sup.RI. Boundaries may be roadway markings, roadway boundaries or other profiles indicating the expanse of the roadway. Visual boundaries as well as structural boundaries may be present.

[0033] An angle θ, which describes the angle between the respective vehicle coordinate system 206 and coordinate systems 204a,b of the roadway boundaries, may be established based in each case on the respective vehicle coordinate system 206, 207, and based in each case on right and left coordinate system 204a, 204b of roadway boundary 203a,b.

[0034] Thus, an angle θ.sup.LE between coordinate system 204a of left roadway boundary 203a and vehicle coordinate system 206 may be determined.

[0035] An angle θ.sup.RI between coordinate system 204b of right roadway boundary 203b and vehicle coordinate system 206 may also be determined.

[0036] The vehicle in instantaneous position 201 may be established in its position relative to roadways 203a,b by two distances Δy.sub.act.sup.LE and Δy.sub.act.sup.RI the right and left distance relative to roadway boundary 203a,b, respectively, based on the origin of coordinate system 206 of the vehicle in instantaneous position 201.

[0037] Lane boundaries 203a,b may be detected with the aid of sensors, for example, with the aid of a video camera. The respective curvature of the roadway boundaries may be approximated with the aid of so-called clothoid models.

[0038] For the roadway curvature, this yields

κ.sub.1=κ.sub.0+{dot over (κ)}.Math.S

where κ.sub.0 corresponds to the instantaneous curvature, κ.sub.1 corresponds to the curvature in distance s. Distance s in this case is the distance in the x-direction of coordinate system 204a or 204b.

[0039] The ego-motion of the vehicle is ascertained with the aid of the present yaw rate of the vehicle. Yaw rate refers to the rotation rate of the vehicle about its vertical axis. The ego-motion of the vehicle is also understood to mean the direction in which the vehicle is instantaneously moving along the vehicle trajectory.

[0040] The yaw rate signal may be provided by a vehicle dynamics control system such as, for example, an ESP system. Alternatively, the yaw rate signal may also be calculated from the visual flow ascertained with the video camera. A determination with the aid of a specifically provided sensor is also possible.

[0041] With the aid of the yaw rate and the ego-velocity of the vehicle V.sub.ego, it is possible to determine the instantaneously present curvature of vehicle trajectory 205.

[00001]${\kappa }_{\mathrm{act}}=\frac{\mathrm{yawrate}}{v_{\mathrm{ego}}}$

[0042] Additional vehicle-specific variables, which may be used below, are length l.sub.wheelbase, which refers to the center distance between the front and the rear axle, as well as length l.sub.carwidth, which corresponds to the front wheelbase of wheels 208, 209 of the vehicle. Wheels 208, 209 are provided with reference numerals in FIG. 2 only in subsequent position 202.

[0043] The subsequent position of vehicle 202 may be predetermined based on a calculation of the described variables of the vehicle at its instantaneous position 201, as well as of the vehicle in relation to the roadway boundaries. This predictive ascertainment yields a variable DLC, the so-called “distance to lane crossing”. FIG. 2 shows variable DLC twice: once as DLC.sub.pred.sup.LE and once as DLC.sub.pred.sup.RI.

[0044] DLC.sub.pred.sup.LE refers to the distance of left front wheel 208 to left roadway boundary 203a of the vehicle at subsequent position 202. DLC.sub.pred.sup.RI refers to the distance of right front wheel 209 to right roadway boundary 203b of the vehicle at subsequent position 202.

[0045] Variables DLC represent a measure of the approach of each front wheel 208, 209 to boundaries 203a, 203b. DLC may also be referred to as the distance variable or approach variable. Approach variable DLC is determined with the aid of the clothoid formula as follows, in this case by way of example of the approach variable of the left front wheel.

[00002]${\mathrm{DLC}}_{\mathrm{pred}}^{\mathrm{LE}}=\Delta .Math..Math.y_{\mathrm{BCI}}^{\mathrm{LE}}+\tan \left({\theta }^{\mathrm{LE}}\right).Math.d_{\mathrm{pred}}+\frac{\left({\kappa }^{\mathrm{LE}}-{\kappa }_{\mathrm{act}}\right).Math.d_{\mathrm{pred}}^2}{2}+\frac{\left({\stackrel{.}{\kappa }}^{\mathrm{LE}}-{\stackrel{.}{\kappa }}_{\mathrm{act}}\right).Math.d_{\mathrm{pred}}^3}{6}-\frac{1_{\mathrm{carwidth}}}{2}$

[0046] Thus, approach variable DLC.sub.pred.sup.LE results from an instantaneous distance of the vehicle to roadway boundary Δy.sub.act.sup.LE from angle θ.sup.LE between the roadway boundary and the vehicle, from prediction length d.sub.pred, from curvature difference (K.sup.le−K.sub.act) of roadway boundary 203a and from trajectory 205, from the difference between curvature changes ({dot over (κ)}.sup.LE−{dot over (κ)}.sub.act) and wheelbase l.sub.carwidth.

[0047] More precisely, the approach variable results from a sum of: [0048] instantaneous distance of the vehicle to roadway boundary Δy.sub.act.sup.LE [0049] the tangent of angle θ.sup.LE multiplied by prediction length d.sub.pred [0050] half the curvature difference (κ.sup.LE−κ.sub.act) multiplied by the second power of prediction length d.sub.pred [0051] a sixth of the difference of curvature changes ({dot over (κ)}.sup.LE−{dot over (κ)}.sub.act) multiplied by the third power of prediction length d.sub.pred [0052] minus half vehicle length l.sub.carwidth.

[0053] Prediction length d.sub.pred results from the ego-velocity V.sub.ego of the vehicle, multiplied by prediction time t.sub.pred, which is then added to wheelbase l.sub.wheelbase of the vehicle as follows:

d.sub.pred=V.sub.ego.Math.t.sub.pred+l.sub.wheelbase.

[0054] Prediction time t.sub.pred may be varied. Prediction time t.sub.pred may be set, depending on the latencies prevailing in the overall system or also as a function of customer requirements. One value for prediction time t.sub.pred, for example, is 700 ms. Based on this variable, it is possible, even during use of the vehicle, to adjust to whether an earlier warning or a later warning is desired. The consideration of wheel base l.sub.wheelbase is significant, particularly in longer vehicles. For small angles, the tangent function of the angle may be replaced approximately by the angle itself.

[0055] Prediction length d.sub.pred may also be varied regardless of the calculation shown. A situation may be present in which the sensors for detecting the roadway boundaries are unable to detect or fully detect the boundaries. Such a situation may occur, for example, if other vehicles cover the detection range of the sensors or disrupt the detection, or also if the visibility is poor. In such a situation, it is possible to shorten the prediction length. In this way, an extreme prediction error is prevented, which could result in undesirable system responses. The prediction length is altered as a function of the expanse of the detection range.

[0056] The prediction length may be adapted by forming the minimum from the comparison of the previous prediction length, calculated at least from ego-velocity V.sub.ego and prediction time t.sub.pred, and a detection length d.sub.sens. Detection length d.sub.sens corresponds to the detection width of the sensor used, in other words, the length detectable by the sensor. The determination is as follows: d.sub.pred=min(V.sub.ego.Math.t.sub.pred,d.sub.sens). Variable l.sub.wheelbase may also be considered in this determination. Prediction length d.sub.pred may subsequently be restored to the initial value.

[0057] Variable DLC.sub.pred.sup.LE ascertained in this way is positive if the corresponding vehicle wheel is located within the corresponding roadway boundary, and is negative in the event that it is located outside the corresponding roadway boundary.

[0058] The depicted ascertainment of approach variable DLC.sub.pred.sup.LE of the left front wheel may similarly also takes place for the right front wheel. Accordingly, for DLC.sub.pred.sup.RI, this results in:

[00003]${\mathrm{DLC}}_{\mathrm{pred}}^{\mathrm{RI}}=-\left(\Delta .Math..Math.y_{\mathrm{act}}^{\mathrm{RI}}+\tan \left({\theta }^{\mathrm{RI}}\right).Math.d_{\mathrm{pred}}+\frac{\left({\kappa }^{\mathrm{RI}}-{\kappa }_{\mathrm{act}}\right).Math.d_{\mathrm{pred}}^2}{2}+\frac{\left({\stackrel{.}{\kappa }}^{\mathrm{RI}}-{\stackrel{.}{\kappa }}_{\mathrm{act}}\right).Math.d_{\mathrm{pred}}^3}{6}\right)-\frac{1_{\mathrm{carwidth}}}{2}$

[0059] Variable DLC.sub.pred.sup.RI ascertained in this way is also positive if the corresponding vehicle wheel is located within the corresponding roadway boundary, and negative in the event that it is located outside the corresponding roadway boundary.

[0060] FIG. 1 shows the sequence of the method according to the present invention.

[0061] In a first step 101, the roadway is analyzed based on roadway boundaries 203a and/or 203b. The results for each analyzed roadway boundary 203a and/or 203b based on the analysis are [0062] the instantaneous distance of the coordinate systems of roadway boundary 204 a and/or b relative to coordinate system 206 of vehicle 201: Δy.sub.act.sup.LE and/or Δy.sub.act.sup.RI, [0063] the instantaneous angle between the coordinate systems of roadway boundary 204a and/or b relative to coordinate system 206 of vehicle 201: θ.sup.LE and/or θ.sup.RI, [0064] the curvature: K.sup.RI and/or K.sup.LE, [0065] the curvature change: {dot over (κ)}.sup.LE and/or {dot over (κ)}.sup.RI.

[0066] The cited variables Δy.sub.act.sup.LE, Δy.sub.act.sup.RI, θ.sup.LE, θ.sup.RI, κ.sup.RI, κ.sup.LE, {dot over (κ)}.sup.LE, {dot over (κ)}.sup.RI are all variables, which are assigned to a lane of the vehicle and which may also be referred to as lane characteristics.

[0067] It is not absolutely necessary to analyze both roadway boundaries. A one-sided analysis—with respect only to left roadway boundary 203a or with respect only to right roadway boundary 203b—is conceivable. It is also possible to examine both roadway boundaries 203a and 203b.

[0068] In a second step 102, the instantaneous driving situation of the vehicle is ascertained. This is achieved by determining: [0069] instantaneous curvature K.sub.act from instantaneous ego-velocity V.sub.ego and the instantaneous yaw rate, [0070] instantaneous curvature change {dot over (κ)}.sub.act

[0071] These variables stand for the instantaneous driving situation of vehicle 201 and may be referred to as driving situation variables K.sub.act, V.sub.ego, yaw rate and {dot over (κ)}.sub.act.

[0072] In a step 103, approach variable DLC.sub.pred is ascertained from the lane characteristics and from the driving situation variables, as previously shown above. The ascertainment may take place for only one approach variable DLC.sub.pred.sup.RI or DLC.sub.pred.sup.LE, as well as for two approach variables DLC.sub.pred.sup.RI and DLC.sub.pred.sup.RI together.

[0073] If only one approach variable is ascertained, this variable is set equal to a value DLC.sub.pred in a step 104. The ascertained approach variable may also be directly used again without setting it as DLC.sub.pred.

[0074] The additional relevant variables d.sub.pred, t.sub.pred, l.sub.wheelbase for ascertaining approach values DLC.sub.pred.sup.RI and/or DLC.sub.pred.sup.RI are stored and are used in the method for calculation. Approach variable DLC.sub.pred is compared in step 104 to a threshold value. The method continues to step 105 if the threshold value is not reached.

[0075] If both approach variables DLC.sub.pred.sup.RI and DLC.sub.pred.sup.RI are ascertained in step 103, a minimum value determination DLC.sub.pred=MIN(DLC.sub.pred.sup.LE,DLC.sub.pred.sup.RI) is then made in step 107a by comparing the two approach variables. Thus, value DLC.sub.pred corresponds to the smaller of the two approach values DLC.sub.pred.sup.RI and DLC.sub.pred.sup.RI.Approach value DLC.sub.pred is compared to a threshold value in step 107b. The method continues in step 105 if the approach value does not reach the threshold value.

[0076] If the threshold value for the approach variable is not reached, the method in step 106 then continues again to step 101. The return path to step 101 occurs both in the variant in which only one approach value is used in step 104, as well as in the variant in which two approach values are used in steps 107a and 107b.

[0077] If approach variable DLC.sub.pred does not reach the threshold value, an information signal is then provided in step 105. Subsequent actions may then be taken, based on the information signal. Subsequent actions may be, but are not limited to, for example, warnings to the driver, braking interventions, steering interventions, velocity adjustments, interventions in a longitudinal control or vehicle dynamics interventions.

[0078] The threshold value used in steps 104 or 107b may be variably configured. Based on the threshold value, the system may be adjusted in terms of how soon the information signal should be output. In other words, it may be established based on the threshold value up to which approach variable DLC.sub.pred a driving situation may be classified as still nonhazardous, and as of which approach variable DLC.sub.pred an action (warning and/or subsequent actions) is necessary. Such a threshold value may, in particular, also be set to 0, corresponding to a completed approach to the roadway boundary.

[0079] An adaptation of the prediction length to the detection range of the at least one sensor for detecting the roadway boundaries, as previously described, is not shown in FIG. 2, but is conceivable at any time. The adaptation of prediction length d.sub.pred to the detection range may also be carried out with the aid of a control unit described below.

[0080] FIG. 3 shows a control unit for carrying out the method.

[0081] In a first subunit 301 of the control unit, lane characteristics Δy.sub.act.sup.LE, Δy.sub.act.sup.RI, θ.sup.LE, θ.sup.RI, κ.sup.RI, κ.sup.LE, {dot over (κ)}.sup.LE and {dot over (κ)}.sup.RI and are ascertained.

[0082] In a second subunit 302 of the control unit, driving situation variables κ.sub.act, V.sub.ego, yaw rate and {dot over (κ)}.sub.act are ascertained. First and second subunits 301 and 302 convey the variables to a third subunit 303, in which approach variable DLC.sub.pred is ascertained. Additional relevant variables d.sub.pred, t.sub.pred, l.sub.wheelbase for ascertaining approach values DLC.sub.pred.sup.RI and/or DLC.sub.pred.sup.RI are stored and are used in the method for calculation.

[0083] Approach variable DLC.sub.pred is compared to the previously cited threshold value. As a function of the comparison, an information variable is generated in an output unit 304 when the threshold value is not reached. This variant of the control unit is used to carry out the method according to path 104 with the ascertainment of only one single approach variable for only one roadway boundary of one side.

[0084] The described minimum value ascertainment from path 107 of the method does not have to be implemented in this control unit. It may also be sufficient, therefore, to ascertain only the variables of one side of the roadway boundaries in first subunit 301 of the control unit and to also determine only the approach variable for this side of the roadway boundary in third subunit 303.

[0085] FIG. 4 shows a control unit, which is able to carry out the method according to path 107, i.e. while taking both sides of roadway boundary 203a and 203b into consideration.

[0086] Steps 301 and 302 remain identical to those in the control unit from FIG. 3, except that at this point the variables of both sides of the roadway boundary must be evaluated.

[0087] In third subunit 401, two approach variables are ascertained. The minimum value ascertainment of the approach values, as well as the comparison to the threshold value, takes place in intermediate unit 402. An information variable is generated in output unit 403 if the threshold value is not reached.

[0088] Both control units from FIG. 3 or FIG. 4 are able to receive or be supplied with additional variables, for example, the prediction time to be set, the prediction length to be set or the threshold values to be used, and to process them in the respective appropriate units.

[0089] In addition, the control units may further provide other systems, control units or subunits via interfaces with variables in addition to the information variable, for example, the approach variables for active lane keeping control as previously mentioned.

[0090] The present invention also includes a computer program, which is configured to carry out each step of the method, and an electronic memory medium, on which this computer program is stored. This electronic memory medium is contained in one of the control units described.