Driving a motor vehicle

Abstract

A procedure (300) for determining a lane (115) of a road (110) traversed by a motor vehicle (105) comprises steps of determining (310) hypotheses (120) for the traversed lane (115) using different validators (220); wherein a hypothesis (120) comprises a time, a lane (115) and a probability that the motor vehicle (105) was traversing the specified lane (115) at the specified time; combining (325) the two oldest hypotheses (120) to provide a combined hypothesis (120) relating to the time of the younger of the two hypotheses (120); repeating the combining (325) until only one combined hypothesis (120) remains; and providing (335) the combined hypothesis (120).

Description

[0001] The present invention relates to the control of a motor vehicle. In particular, the invention relates to the control of a motor vehicle with respect to a lane of a roadway traversed by the motor vehicle.

[0002] A motor vehicle comprises a control device configured to steer the motor vehicle along a road. The road comprises multiple lanes and may, for example, comprise a motorway or a similarly constructed expressway. Control can be effected, in particular, in a transverse direction, whereby the motor vehicle can be kept in the currently traversed lane.

[0003] For correct control, it is important to ensure that the lane the vehicle is traveling in has been correctly determined. For example, it is important to avoid the vehicle being unnoticed on an exit leading away from the road.

[0004] The lane can be determined using a validator. Different validators can be based on different data sources. To correctly determine the lane, hypotheses regarding the lane provided by the validators should be combined into a single hypothesis. The individual hypotheses can refer to different points in time. The combination should be performed in such a way that possible errors or inaccuracies are accounted for exactly once in the combination.

[0005] One object underlying the present invention is to provide an improved technique for determining a lane of a roadway traversed by a motor vehicle. The invention achieves this object by means of the subject matter of the independent claims. Subclaims specify preferred embodiments.

[0006] According to a first aspect of the present invention, a method for determining a lane of a road traveled by a motor vehicle comprises steps of determining hypotheses for the traveled lane using mutually different validators; wherein a hypothesis comprises a time, a lane and a probability that the motor vehicle has traveled in the specified lane at the specified time; combining the two oldest hypotheses to provide a combined hypothesis related to the time of the younger of the two hypotheses; repeating the combining until only one combined hypothesis remains; and providing the combined hypothesis.

[0007] By successively relating the oldest hypothesis to the time of the next most recent hypothesis, errors can be avoided from being considered multiple times. Hypotheses that have been included in the combined hypothesis can be dropped. With each combination, the number of hypotheses to be considered can be reduced by one. The final combined hypothesis can represent a meaningful summary of the information from all the hypotheses included in the validators.

[0008] In a preferred embodiment, one of the hypotheses is referenced to the current time before combining. This ensures that the ultimately provided combined hypothesis is related to the current time or the most recent possible time. To reduce errors, it has proven effective to relate the most recent hypothesis to the current time.

[0009] When relating the selected hypothesis to the current time, the probability that the vehicle has taken a known exit from the specific lane can be considered. The exit may lead away from the current road and be determined, for example, based on the vehicle's geographical position and map data relating to the road.

[0010] The probability that the vehicle took an unknown exit from the specific lane can also be considered. This allows for a possible map error regarding the existence or location of an exit to be accounted for.

[0011] In a further preferred embodiment, the combination of probabilities is based on the following relationship: pk = (1-(1-p1) * (1-p2)). Here, p1 and p2 represent probabilities of the hypotheses, and pk represents the probability of the combined hypothesis. In the manner described, it is possible to correctly determine the probability of the combined hypothesis based on the probabilities of the hypotheses that are incorporated into the combined hypothesis.

[0012] It is further preferred that the validators' hypotheses be substantially independent of each other. In particular, it must be ensured that there is no significant error that affects several or all hypotheses of the different validators in the same way.

[0013] For this purpose, it is particularly preferred that the validators use sensor values ​​from different sensors. A validator can use sensor values ​​from multiple sensors. However, it is preferred that no sensor provides sensor values ​​to multiple validators.

[0014] In a further preferred embodiment, the point in time to which a hypothesis refers lies in a predetermined past, extending up to the current point in time. In other words, the hypothesis can have an age that lies within a predetermined range. This range points into the past and can include the current point in time. For example, the point in time relative to a current point in time can be between zero and approximately sixty seconds in the past.

[0015] From a validator's hypotheses, one with the youngest age can be determined. This can have a positive effect on error propagation over time.

[0016] Furthermore, it is preferred that a validator determines a hypothesis with the maximum probability. This probability can be related to the time of determination. Preferably, the probability of the hypothesis being determined is related to the current time. This allows for the most meaningful hypotheses to be selected and combined more effectively.

[0017] These two conditions can be combined in any way. For example, one can determine the most recent hypothesis with the maximum probability. Alternatively, one can determine the one with the youngest age among the highest probabilities.

[0018] According to a further aspect of the present invention, a device for determining a lane of a road traveled by a motor vehicle comprises the following elements: a plurality of validators; wherein a validator is configured to determine a hypothesis for the traveled lane; wherein a hypothesis comprises a time, a lane, and a probability that the motor vehicle traveled in the specified lane at the specified time; a processing device and an output device for providing a combined hypothesis. The processing device is configured to combine the two oldest hypotheses to provide a combined hypothesis related to the time of the younger of the two hypotheses, and to repeat the combining until only a single combined hypothesis remains.

[0019] The processing device is preferably configured to partially or completely execute a method described herein. For this purpose, the processing device can be implemented electronically and, for example, comprise an integrated circuit, a programmable logic module, or a programmable microcomputer. The method can be implemented in the form of a configuration or as a computer program product with program code means for the processing device. The configuration or the computer program product can be stored on a computer-readable data carrier. Features or advantages of the method can be transferred to the device, or vice versa.

[0020] According to yet another aspect of the present invention, a control device comprises a device as described herein. The control device can be configured to control the motor vehicle. The control can be performed, in particular, in the transverse direction and aimed at keeping the motor vehicle in a predetermined lane.

[0021] According to yet another aspect of the present invention, a motor vehicle comprises a device or a control device described herein. The motor vehicle may, in particular, comprise a motorcycle, a passenger car, a truck, or a bus.

[0022] The invention will now be described in more detail with reference to the accompanying drawings, in which: Fig. 1 a motor vehicle on a road; Fig. 2 a system on board a motor vehicle; Fig. 3 a flow diagram of a method; and Fig. 4 illustrates a process for determining a combined hypothesis.

[0023] Fig. 1 shows a system 100 with a motor vehicle 105. The motor vehicle 105 travels on a road 110 with several lanes 115. By way of example, two left lanes 115 run along the road 110, while a lane 115 shown on the right includes an exit leading away from the road 110. To control the motor vehicle 105 on a predetermined lane 115, it is to be determined which lane 115 of the road 110 is being traveled by the motor vehicle 105. For this purpose, different validators can be used, which operate essentially independently of one another. At regular or irregular intervals, a validator can provide a hypothesis 120 that refers to one of the lanes 115 of the road 110. Hypotheses 120 are in Fig. 1 represented by circular areas, with different brightnesses or patterns indicating different validators. Hypotheses 120 with the same brightness or hatching originate from the same validator.

[0024] Hypotheses 120 are typically not provided synchronously. A hypothesis 120 is associated with a lane 115, a time, and a probability that the motor vehicle 105 was in the specified lane 115 at the specified time. A technique presented herein aims to combine hypotheses 120 from different validators to provide a combined hypothesis 120.

[0025] Fig. 2 shows a motor vehicle 105 with a device 205 for determining the lane 115 of a road 110 traveled by the motor vehicle 105. The device 205 can be connected via an interface 210 to a control device 215 for controlling the motor vehicle 105. The control device 215 can be configured to keep the motor vehicle 105 in a predetermined lane 115. In a further embodiment, the device 205 and the control device 215 can also be combined with one another in a single device.

[0026] The device 205 comprises a plurality of validators 220 and a processing device 225. The validators 220 preferably process sensor values ​​from different sensors. This means that a validator 220 can operate based on sensor values ​​from one or more sensors. However, the processed sensor values ​​are preferably not provided to any other of the validators 220.

[0027] A validator 220 determines hypotheses 120 regarding the lane 115 of a road 110 traveled by the motor vehicle 105. The processing device 225 is configured to combine several hypotheses from different validators 220 and to provide a combined hypothesis that can, for example, form the basis for controlling the motor vehicle 105.

[0028] Fig. 3 shows a flowchart of a method 300 for determining a lane 115 of a road 110 traversed by a motor vehicle 105. In a step 305, available validators 220 can be determined, each of which has provided at least one hypothesis 120 regarding a traversed lane 115. If a validator 220 is unavailable or does not provide any hypotheses 120, it can be removed from the determination.

[0029] In a step 310, a hypothesis 120 can be determined for each of the validators 220. From the hypotheses of the validator 220, a hypothesis can be selected that preferably simultaneously has the highest probability and the youngest age. For this purpose, only hypotheses within a predetermined, past time range can be considered. For example, a hypothesis 120 can be determined for each validator 220 that represents the most probable and simultaneously most recent hypothesis within a past time range from zero to approximately sixty seconds.

[0030] In a step 315, one of the selected hypotheses can be updated to the current time. In doing so, the determined probability can be reduced. The reduction can be made relative to an initial probability with which the motor vehicle 105 has left the lane 115 or the road 110 at a known exit since the determination time. Furthermore, a probability with which the motor vehicle has left the lane 115 or the road 110 at an unknown exit can be taken into account.

[0031] The updated hypotheses and all other hypotheses can now be combined. To do this, the oldest hypothesis, i.e., the one relating to the furthest back in time, can be selected in step 320. In step 325, the oldest hypothesis can be updated to the next most recent hypothesis. This can result in a combined hypothesis, and the two processed hypotheses can be discarded. The combined hypothesis receives a probability that can be determined according to the following relationship: pk=(1−(1−p1)⋅(1−p2))

[0032] Where pk represents the combined probability, p1 represents the probability of a first processed hypothesis 120, and p2 represents the probability of a second processed hypothesis 120. By executing step 325, two hypotheses are removed and one combined hypothesis is added. Overall, the number of hypotheses to be processed has thus decreased by one.

[0033] In a step 330, it can be checked whether more than one hypothesis 120 remains. If this is the case, the method 300 can return to step 320 and run again. Otherwise, all hypotheses 120 have been successively combined, and the remaining single hypothesis 120 represents the combined hypothesis 120 of all other hypotheses 120. In this case, the combined hypothesis 120 can be provided in a step 335.

[0034] Fig. Figure 4 shows an example process for determining a combined hypothesis 120. A horizontal line marks a current point in time. Five hypotheses 120 are provided purely as examples, whose points in time are indicated by their vertical positions in Fig. 4. Different brightnesses or patterns of the circular symbols used indicate different assigned validators 220.

[0035] According to step 315 of method 300 of Fig. 3, a hypothesis 120 can first be updated to the current time. In the representation of Fig. 4, the most recent hypothesis 120 is used, which is marked with a white symbol.

[0036] Now, the two oldest hypotheses 120 are combined to form a first combined hypothesis 405. The combining is performed in the manner of step 325 of the method 300. A time assigned to the combined hypothesis 405 corresponds to the time of the younger of the two processed hypotheses 120. The processed hypotheses 120 are discarded.

[0037] The first processed hypothesis 405 is combined with the then oldest hypothesis 120 to form a second combined hypothesis 410 (see step 325 of method 300). The processed hypothesis 120 can also be dropped.

[0038] In a corresponding manner, a third combined hypothesis 415 can now be determined on the basis of the then oldest hypothesis 120 and the second combined hypothesis 410 in the manner of step 325 of the method 300.

[0039] In a final step, the third combined hypothesis 415 can be combined with the then only remaining hypothesis 120 to form a fourth combined hypothesis 420. The processed hypothesis 120 can be dropped, leaving the combined hypothesis 420 as the only hypothesis. Accordingly, the fourth combined hypothesis 420 can be provided as a hypothesis for the lane 115 of the roadway 110 traversed by the motor vehicle 105. Reference symbol 100 systems 105 Motor vehicle 110 road 115 lane 120 Hypothesis 205 Device 210 Interface 215 Control device 220 Validators 225 processing facility 300 procedures 305 Validators determine 310 determine the best and most current hypothesis Update 315 to current time 320 Select oldest hypothesis 325 update to next younger hypothesis 330 >1 hypothesis remaining? 335 provide combined hypothesis 405 first combined hypothesis 410 second combined hypothesis 415 third combined hypothesis 420 fourth combined hypothesis

Claims

[1] Method (300) for determining a lane (115) of a road (110) travelled by a motor vehicle (105), the method (300) comprising the following steps: - determining (310) hypotheses (120) for the lane (115) being traveled by means of mutually different validators (220); - wherein a hypothesis (120) comprises a time, a lane (115) and a probability that the motor vehicle (105) traveled in the specified lane (115) at the specified time; - combining (325) the two oldest hypotheses (120) to provide a combined hypothesis (120) related to the time of the younger of the two hypotheses (120); and - Repeating the combining (325) until only one combined hypothesis (120) remains; and - Providing (335) the combined hypothesis (120). [2] Method (300) according to claim 1, wherein prior to combining (325) one of the hypotheses (120) is related (315) to the current time. [3] The method (300) of claim 2, wherein a probability is taken into account that the motor vehicle (105) has taken a known exit from the particular lane (115). [4] Method (300) according to claim 2 or 3, wherein a probability is taken into account that the motor vehicle (105) has taken an unknown exit from the specific lane (115). [5] Method (300) according to one of the preceding claims, wherein the probabilities are combined as follows: pk = (1- (1-p1) * (1-p2)); where p1 and p2 represent probabilities of the hypotheses (120) and pk represents the probability of the combined hypothesis (120). [6] Method (300) according to one of the preceding claims, wherein the hypotheses (120) of the validators (220) are substantially independent of each other. [7] The method (300) of claim 6, wherein the validators (220) use sensor values ​​from different sensors. [8] Method (300) according to one of the preceding claims, wherein the time is within a predetermined past extending to the current time. [9] Method (300) according to one of the preceding claims, wherein from hypotheses (120) of a validator (220) one with the lowest age is determined (310). [10] Method (300) according to one of the preceding claims, wherein a validator (220) determines (310) a hypothesis (120) whose probability is maximum. [11] The method (300) of claim 10, wherein the probability of the related hypothesis (120) is related to the current time. [12] Device (205) for determining a lane (115) of a road (110) travelled by a motor vehicle (105), the device comprising the following elements: - multiple validators (220); - wherein a validator (220) is configured to determine a hypothesis (120) for the lane (115) being traveled; - wherein a hypothesis (120) comprises a time, a lane (115) and a probability that the motor vehicle (105) traveled in the specified lane (115) at the specified time; - a processing device configured to combine the two oldest hypotheses (120) to provide a combined hypothesis (120) related to the time of the younger of the two hypotheses (120); and to repeat the combining until only one combined hypothesis (120) remains; and - an output device for providing the combined hypothesis (120). [13] Control device (215) comprising a device (205) according to claim 12, wherein the control device (215) is configured to control the motor vehicle (105).

Images