EP3465651A1

EP3465651A1 - Method, device and system for detecting wrong-way drivers

Info

Publication number: EP3465651A1
Application number: EP17717382.0A
Authority: EP
Inventors: Simon GEISLER
Original assignee: Robert Bosch GmbH
Current assignee: Robert Bosch GmbH
Priority date: 2016-06-07
Filing date: 2017-04-11
Publication date: 2019-04-10
Also published as: CN109313848A; JP2019519043A; US20190145783A1; WO2017211482A1; DE102016210017A1; US10876843B2

Abstract

The invention relates to a method for detecting wrong-way drivers, comprising the following steps: inputting positional data (106) via an interface, the positional data (106) representing a measured position of a vehicle (100); a step of determining a plurality of particles, by using the positional data (106), one particle representing a position occupied by the vehicle (100) and a weighting associated with the occupied position, and a step of determining the actual position of the vehicle (100) on a road network driven by the vehicle (100), based on the plurality of particles, by using a particle filter (532).

Description

Description title

Method Apparatus and system for wrong driver recognition state of the art

The invention is based on a device or a method according to the preamble of the independent claims. The subject of the present invention is also a computer program.

False drivers ("ghost drivers") cause at least substantial property damage in the event of an accident

Navigation device (street class and direction) is too late for most cases, i. the wrong-way driver is already on the wrong lane (at high speed and with a high probability of collision).

Disclosure of the invention

Against this background, with the approach presented here, a method, furthermore a device and a system for wrong-way driver recognition, and finally a corresponding computer program according to the

Main claims presented. The measures listed in the dependent claims advantageous refinements and improvements of the independent claim device are possible.

An example cloud-based forwarder warning can be advantageously realized with a specially adapted to the application detection with a particle filter.

A method for detecting wrong-way drivers comprises the following steps: Reading position data via an interface, the position data representing a measured position of a vehicle;

Determining a plurality of particles using the position data, wherein a particle represents an assumed position of the vehicle and a weight assigned to the assumed position; and

Determining a current position of the vehicle on a road network drivable by the vehicle based on the plurality of particles using a particulate filter.

The vehicle may be a road vehicle. A wrong travel can be understood to mean a journey of the vehicle on a road contrary to a prescribed direction of travel. The measured position may have been measured using a sensor disposed in the vehicle. The plurality of particles may be determined using a method used with known particle filters. The particles may have different assumed positions, which are grouped around the measured position, for example. The current position may represent an estimated position using the particulate filter, which may be assumed to be the actual position of the vehicle. The current position can be used instead of the measured position for detecting a wrong-way of the vehicle.

The method may include a step of determining a wrong-way signal using the current position. In this case, the wrong-way signal can indicate whether a wrong-way drive of the vehicle is present or not present.

For example, the wrong-way signal can be provided only if a wrong-way is assumed.

The method may include a step of reading in map data depicting the road network drivable by the vehicle. In this case, in the step of determining the current position of the vehicle can be determined using the map data. The map data can be used to determine the current position with high accuracy. In the read-in step, the position data can be read in via an interface of a computer cloud, a so-called cloud. This enables a cloud-based solution.

In the step of determining, a plurality of changed particles may be determined based on the plurality of particles using the particulate filter. The current position of the vehicle can be adjusted accordingly

Using the plurality of changed particles are determined. For example, a weighting of the particles can be changed by the particle filter, whereby the current position can in turn be determined more accurately.

In this case, in the step of determining the current position further under

Using a plurality of previous modified particles are determined, which represent certain particles based on a plurality of previous particles using the particle filter. In this way, a history of the movement of the vehicle can be taken into account in the determination of the current position.

A corresponding device for identifying wrong-way drivers is set up to execute steps of said method in corresponding units. For example, such a device may comprise a read-in device, which is designed to read position data via an interface, have a determination device that is configured to determine a plurality of particles using the position data, and a

Determining means configured to determine a current position of the vehicle on a road network drivable by the vehicle based on the plurality of particles using a particulate filter. Accordingly, the device may comprise the particle filter.

A corresponding system for detecting wrong-way drivers comprises at least one transmitting device which can be arranged or arranged in a vehicle and is designed to transmit position data, as well as a named one

False driver recognition device, which is designed to receive the position data transmitted by the at least one transmitting device

received, for example via a wireless connection. Another false-driver detection system includes at least one transmitting device locatable or arranged in a vehicle and configured to transmit position data, the position data representing a measured position of a vehicle, and at least one receiving device locatable or arranged in the vehicle and is configured to receive data from a device which, in accordance with the approach for false driver recognition described here, is designed to receive the position data transmitted by the at least one transmission device.

The method described may be implemented in software or hardware or in a hybrid of software and hardware, for example in a device.

For this purpose, the device can have at least one arithmetic unit for processing signals or data, at least one memory unit for storing signals or data, and / or at least one communication interface for reading in or outputting data that is included in a

Embedded communication protocol. The arithmetic unit can

For example, be a signal processor, a microcontroller or the like, wherein the memory unit is a flash memory, an EPROM or a

magnetic storage unit can be. The communication interface can be designed to read or output data wirelessly and / or by line, wherein a communication interface that can read or output line-bound data, for example, electrically or optically read this data from a corresponding data transmission line or output to a corresponding data transmission line.

In the present case, a device can be understood as meaning an electrical device which processes sensor signals and outputs control and / or data signals in dependence thereon. The device may have an interface, which may be formed in hardware and / or software. In the case of a hardware-based embodiment, the interfaces can be part of a so-called system ASIC, for example, which contains a wide variety of functions of the device. However, it is also possible that the interfaces are separate, integrated circuits or at least partially made of discrete components consist. In a software training, the interfaces may be software modules that are present, for example, on a microcontroller in addition to other software modules. Of advantage is also a computer program product or computer program with

Program code, which may be stored on a machine-readable medium or storage medium, such as a semiconductor memory, a hard disk memory or an optical memory, and for carrying out, implementing and / or controlling the steps of the method according to one of the above

described embodiments is used, in particular when the

Program product or program is executed on a computer or a device.

Embodiments of the approach presented here are shown in the drawings and explained in more detail in the following description. It shows:

Fig. 1 shows a system for Falzfahrererkennung according to a

Embodiment; FIG. 2 is a flowchart of a method for detecting wrong-way drivers according to an embodiment; FIG.

3 shows a Hidden Markov Chain Model; 4 shows a sequence of a particle filter process according to a

embodiment

Fig. 5 shows a system for wrong driver identification according to a

Embodiment;

6 shows a vehicle according to an embodiment;

7 shows a program sequence according to an embodiment; and FIG. 8 shows a program sequence of a particle filter according to a

Embodiment. In the following description of favorable embodiments of the present invention are for those in the various figures

represented and similar elements acting the same or similar

Reference is made to a repeated description of this

Elements is omitted.

Fig. 1 shows a system for wrong driver identification according to a

Embodiment. The system includes a vehicle 100 that has a

Transmission device 102 which is configured to wirelessly using a at least one sensor device 104 arranged in the sensor 100 measured data 106 wirelessly to a device 110 for

Send wrong-way driver recognition. The device 110 is designed to prepare the measurement data 106 into prepared data and to further process the processed data using a particle filter

False drive signal 112 to generate and send out. The wrong-way signal 112, according to one embodiment, indicates that the vehicle 100 whose measurement data 106 has been processed currently makes a wrong-way drive. According to this embodiment, both the transmission device 102 of the vehicle 100 and a transmission device 102 of another

Vehicle 100 is configured to receive the wrong-way signal 112 and, in response to a receipt of the wrong-way signal 112, a

Warning device of the respective vehicle 100, 114 to activate, the

For example, a driver of the respective vehicle 100, 114 before

False drive warns or engages according to an embodiment in an at least semi-automatic control, such as a brake system or steering system, the respective vehicle 100, 114. According to different

Embodiments, the transmission device 102 only as

Transmitter or be designed as a transceiver device.

According to one embodiment, the measurement data 106

Position data acquired using a position-determining device of the vehicle 100 and depicting a current position of the vehicle 100. According to a further exemplary embodiment, the measurement data 106 further comprises movement data, which has been acquired, for example, using at least one acceleration sensor of the vehicle 100, and Information about a current movement of the vehicle 100, for example information about a direction of travel, a longitudinal acceleration, a

Lateral acceleration or via a rotation of the vehicle by one

Vehicle axle include.

In one embodiment, the device 110 is configured to read in map data 116 that maps a road network drivable by the vehicle 100. For example, in one embodiment, the map data 116 includes information about road sections of the road network. According to an exemplary embodiment, the map data 116 with respect to each road section further comprises at least one parameter that defines, for example, a driving direction specification for the respective road section or a course of the respective road section. For example, it can be defined via the parameter whether the road section runs in a straight line or describes a curve. According to one embodiment, the device 110 has a memory device in which the map data 116 are stored.

In one embodiment, the device is 110 or

Function blocks of the device 110 are arranged or realized in a cloud 118.

The described approach can be used in addition to or instead of various methods for detecting a wrong-way driver, in which e.g. the use of a video sensor is used to detect the passage of a "forbidden entry" sign or the use of a digital map is used in conjunction with a navigation to detect a detection of a wrong direction of travel on a road section, which is only passable in one direction Furthermore, the approach can be combined with wireless methods that detect wrong-way drivers by means of infrastructure such as beacons in the lane or at the lane.

In addition to the detection of a wrong-way driver, the described approach offers many possibilities of responding to a wrong-way driver. Examples are the warning of the wrong driver himself via a display or acoustic information. Also, procedures can be applied with which other drivers in the Be warned near a wrong-way driver, eg via vehicle-vehicle communication or by mobile radio. Furthermore, the warning of other road users on the roadside established variable traffic signs is possible. An intervention in the engine control or brake of the wrong-traveling vehicle 100 can also take place.

The approach described makes it possible to detect a wrong-way driver and to warn other road users in the vicinity in time, for which there is very little time available.

The described approach uses for a wrong-way driver detection (Wrong Way Driver Detection) with a client-server solution. As a client, a device can be seen, located on or in a motor vehicle, which has a

Internet connection and has at least access to position coordinates. For example, this may be the transmission device 102. The transmission device 102 may be, for example, a smartphone. In the transmission device 102, the

Sensor device 104 may be integrated. Thus, wrong-driver-specific server-client communication can be implemented with a smartphone as an exemplary client. The smartphone can be connected via a mobile radio network with a gateway (PDN_GW) to the Internet, in which the device 110, for example in the form of a server, can be arranged.

The following key problem areas for this technology result from the possible functions of a wrong-way driver warning with a client-server solution, which are addressed by the approach described here: a) False-positive reduction

False positives, ie misdetections with the correct driving style, must be reduced as far as possible or completely avoided in the case of a self-warning and / or active intervention. Depending on the warning concept, the standards must meet up to ASIL-A. b) Time-critical execution of the trip chain To the endangerment of other road users starting from a

To keep wrong-hand drivers as low as possible should be intervened or warned as quickly as possible. Ie. The complete functional chain from the detection of a critical situation to the detection of a wrong-way driver until intervention or warning should be carried out in as short a time as possible. The utilization and thus the required performance of the server,

For example, the device 110, in a nationwide use of this function plays a very important role. In addition to the trip time, cost-efficiency also plays an important part. C) Communication, data efficiency and power consumption

Communication and power consumption, especially for mobile devices, must be as low as possible to achieve acceptable battery life. The overload of a mobile radio cell or other wireless communication unit must also be ensured by a data-efficient one

Communication be prevented. Also the data volume and the associated costs are, as far as possible, to be limited. The efficiency of communication is an extremely important factor on the server side for computing power reasons.

The approach described above is particularly relevant for the key fields a) "false positive reduction" and b) "time-critical execution of the trigger chain", but also c) "communication, data efficiency and power consumption"

possibly influenced by it. The detection of wrong-way drivers in the cloud 118, based on standard smartphone and connectivity control unit sensor technology is not a trivial undertaking.

FIG. 2 shows a flowchart of a method for wrong-way driver recognition according to one exemplary embodiment. The method may, for example, be carried out using devices of the device for false driver recognition shown with reference to FIG.

The method comprises a step 201, in which position data are read in via an interface. The position data represents a measured position of a vehicle. In a step 203, a plurality of Particles determined using the position data. Each of the particles represents an assumed position of the vehicle and a weight assigned to the assumed position. In this case, according to one exemplary embodiment, the assumed positions are distributed around the measured position. In a step 205, a current position of the vehicle on a road network accessible by the vehicle is determined using a particle filter which is used to process the particles.

According to one exemplary embodiment, the current position is determined using particles that have passed through the particle filters and thereby changed, for example, with regard to their weighting. In addition, in determining the current position according to an embodiment, additional or alternative to a previous time step flow under

Using the same particle filter changed particles. These

Previous particles may have been determined in a previous step 203 using previous position data.

For the wrong driver recognition, it is not decisive which route the wrong-way driver drove. Above all, the information needed is where the wrong-way driver is currently located and whether this is a street opposite

Driving direction. Of course, the history is needed for this determination, but this is not part of the question, but rather the way to the result.

Due to these circumstances, a method based on a particle filter is presented. Similar to the Kalman filter, the particle filter is applicable to systems which are subject to a hidden Markov chain characteristic, ie a Markov chain with unobserved states:

Fig. 3 shows a Hidden Markov Chain Model 320 with state x and observation z at time k and k-1.

That is, the state of a system can not be measured directly, but estimated based on other observations. In this case It is important to estimate the position and thus the current road. For this, the following equation must be solved:

The state at time k is described below with x _k , the previous states are summarized with. Analogous

For x this convention also applies to the control quantities u and observations describes a normalization term, which in the following, however, has no great significance. This equation can be simplified to the following equation:

And these are described in two steps: the forecasting step

and the weighting term:

For a particle filter, the integral over the

Probability distributions with a numerical approximation

and Monte Carlo methods, this describes the weight / probability of the particle. A lot of particles will be with you

described. Thus, each particle has the weight and the condition

4 shows the sequence of a particle filter process according to a

Embodiment. For this purpose, a hidden Markov Chain Model with the state x and the observation z at time k and k-1 is shown in FIG.

Much of the work is a suitable feature for

and

TO

find that optimally map the problem. The basis for this is to define the states x to be estimated.

Block 401 stands for the particulate filter

From block 403, jump is made to block 405 until all values have passed.

In block 405, a new state is calculated:

In block 407, the weight is calculated:

If all values have been traversed in block 403, block 409 is entered. From the block 409, the block 411 is skipped until all values i = 1: J have passed. In block 411, a value is drawn in accordance with.

In block 413, the particle set is added according to

If all values have been traversed in block 409, the program branches to block 415, which represents the end X _k .

Fig. 5 shows a system for wrong driver recognition according to a

Embodiment. The system comprises devices 102, for example in the form of the transmission means referred to with reference to FIG. 1 and a

Device 110 for wrong driver identification, according to this

Embodiment designed as a so-called WDW server. The device 110 is designed to receive data 106 from the device 102,

For example, to receive the measurement data described with reference to FIG. 1 and to provide a warning 112 based on the data 106 and to send back to the devices 102, for example in the form of the wrong-way signal described with reference to FIG.

The apparatus includes pre-processing means 530, particulate filter 532, and warning module 534.

In a simplified cloud-based forklift warning architecture, the particulate filter 532 embeds as shown in FIG.

With the particle filter 532, the probability distribution of the position of the car can be approximated.

FIG. 6 shows by means of a vehicle 100 values that can be included in the model shown with reference to FIG. 5. The values may, for example, be states in the direction of the longitudinal axis x, the transverse axis y, the vertical axis z, as well as a roll p about the longitudinal axis, a pitch q about the transverse axis and a yaw r about the vertical axis.

Regarding map matching using the particulate filter, for the Bayesian filter, referring to FIG.

stand for what the condition (not measured) is, for example the geographic longitude, latitude and altitude, U _{k + 1} stand for how the car 100 is moving, for example in terms of speed and yaw rates and Z k what can be observed, such as a GPS signal or the environment of the vehicle 100 relevant signal (camera, etc.)

Fig. 7 shows a program flow according to an embodiment. The process starts with a block 701. In a block 530, a

Data preprocessing performed, as described for example with reference to FIG. 5. In a block 703, if present, the state is loaded from the previous point. In a block 705, a map matching takes place with the particle filter. In a block 707 is a

Interpretation of the results. In a block 709 it is checked if a

Wrong course exists. If so, a warning is sent in block 534, as described, for example, with reference to FIG. If there is no wrong drive, the end of the program sequence takes place with a block 711.

FIG. 8 shows a program flow of a particle filter according to a

Embodiment. A block 801 stands for a beginning of the particle filter. In a block 803, a displacement of the particles taking into account the sensor inaccuracy, for example, the sensor device described with reference to FIG. 1 takes place. In a block 805, a determination of the

card-related parameters. For example, such a parameter indicates whether a particle is on a road or what its title is. In a block 807, a calculation of the new particle weights takes place. In a block 809 a so-called resampling takes place in which an elimination of the irrelevant regions and / or particles takes place. In a block 811 an interpretation of the individual particles takes place and in a block 813 a return of the possible roads.

By using the particulate filter, the following aspects are improved. On the one hand, a sequential (real-time possible) working method is created, which primarily determines the current position on the road network. Furthermore, a robust estimate of the current position on the road network is possible. An uncertainty about the current Estimate can be determined. This makes it possible to delay the decision on a potential wrong-way reliably to a reasonable extent.

If an exemplary embodiment comprises a "and / or" link between a first feature and a second feature, this is to be read such that the

Embodiment according to an embodiment, both the first feature and the second feature and according to another embodiment, either only the first feature or only the second feature.

Claims

Expectations

1. Method for detecting wrong-way drivers, the method being the

includes the following steps:

Reading in (201) position data (106) via an interface, the position data (106) representing a measured position of a vehicle (100);

Determining (203) a plurality of particles using the position data (106), wherein a particle represents an assumed position of the vehicle (100) and one of the assumed position

assigned weight represents; and

Determining (205) a current position of the vehicle (100) on a road network that can be traveled by the vehicle (100) based on the plurality of particles using a particle filter (532).

2. The method according to claim 1, with a step of determining a wrong-way signal (112) using the current position, the (112) wrong-way signal indicating whether the vehicle (100) is driving the wrong way or not.

3. Method according to one of the preceding claims with a step of reading in map data (116) which depicts the road network that can be traveled by the vehicle (100), wherein in the step of determining the current position of the vehicle (100) using the map data (116 ) is determined.

4. Method according to one of the preceding claims, in which in the step of reading (201) the position data (106) via a

Interface of a computer cloud (118) can be read.

5. The method according to any one of the preceding claims, wherein in the step of determining (205) a plurality of changed particles is determined based on the plurality of particles using the particle filter (532), and the current position of the vehicle (100). Using the majority of modified particles is determined.

6. The method according to claim 5, wherein in the step of determining (205), the current position is further determined using a plurality of previous changed particles which represent certain particles based on a plurality of previous particles using the particle filter (532). .

7. Device (110) for wrong-way driver detection, which is set up to carry out steps (201, 203, 205) of the method according to one of

to carry out the preceding claims in corresponding units (530, 532, 534).

8. System for wrong-way driver detection, the system comprising the following features: at least one transmitting device (102) which can be arranged or is arranged in a vehicle (100) and is designed to transmit position data (106), the position data (106) represent a measured position of a vehicle (100); and a device (110) according to claim 7 for wrong-way driver detection, which is designed to receive the position data (106) emitted by the at least one transmitting device (102).

9. Wrong-way driver detection system, the system comprising the following

Features include: at least one transmitting device (102), which can be arranged or arranged in a vehicle (100), and is designed to transmit position data (106), wherein the position data (106) represents a measured position of a vehicle (100); and at least one receiving device, which can be arranged or arranged in the vehicle (100) and is designed to receive data from a device (110), which according to claim 7 for wrong-way driver detection is designed to detect the data emitted by the at least one transmitting device (102). Receive position data (106).

10. Computer program that is designed to carry out the method according to one of claims 1 to 6.

11. Machine-readable storage medium on which the computer program according to claim 9 is stored.