DE102011078946A1 - Method for determining most probable path of car by software modules, involves providing personal and impersonal driving probability data for correcting original path, where data is derived from previous driving behavior of vehicle - Google Patents

Abstract

The method involves determining an original probable driving path i.e. original most probable path (MPP) (10), under consideration of a driving probability data, which is derived from a road map. A supplementary data i.e. allowed speed, is utilized for correcting the path. A software module is created to the path such that the module provides a two-dimensional (2D) horizon based on the path. Personal and impersonal driving probability data are provided by another module for correcting the path, where the personal driving probability data is derived from previous driving behavior of a driver. Independent claims are also included for the following: (1) an arrangement for determining a probable driving path of a vehicle by a module (2) a computer program product comprising a set of instructions for executing a method for determining a probable driving path of a vehicle by a module.

Description

State of the art

The invention relates to a method for determining a most probable driving path of a vehicle by at least one module, in which an original probable driving path MPP (Most Probable Path) is determined in consideration of driving likelihood data derived from a road map, and in which Correction of the original MPPs additional data can be used. In addition, the invention relates to an arrangement and a computer program product.

A generic method is for example from the DE 10 2009 028 299 A1 known. In the known method, for correcting the MPP, additional data are taken into account by a device which influences an actual traffic situation, for example another vehicle or a traffic light.

From the DE 10 2007 043 533 A1 a method is known in which, when the route guidance function is deactivated, the position horizon is checked in the event of traffic jam warnings and at least one alternative route is calculated and output. It is also known to correct the position horizon depending on the traffic jam section and past driver behavior.

A central consideration in future Advanced Driver Assistance Systems (ADAS), which will intervene more strongly and autonomously than before, is to capture not only the proximity of the vehicle, but also information on the upcoming route based on high-quality digital Maps and the current vehicle position. To such card-based driver assistance systems, e.g. A predictive cruise control system or a curve warning assistant to provide information about the route ahead requires an evaluable, electronic horizon. This can be thought of as a virtual sensor that provides information about the environment of the vehicle based on map data of a digitized road network, the current position, and the direction of travel. An on-vehicle module, the horizon provider, provides this electronic horizon; he constantly determines the route ahead on which the vehicle is expected to move. This route is called MPP (Most Probable Path). If the driver has selected a route in the navigation device, this is used as MPP. If the navigation is not activated, the MPP is determined by various heuristic methods. The driver assistance functions receive and evaluate the attributes of the electronic horizon, that is, the MPP, embedded in an electronic horizon, is transmitted to other controllers so that they are able to optimize look-ahead-based functions. A standardization procedure called Advanced Driver Assistance Systems Interface Specifications (ADASIS) is intended to define an interface between navigation systems and ADAS applications, that is to say define the horizon, for example in the form of an ADASIS horizon, preferably via a CAN bus is sent to the applications. Efforts to standardize are concentrated in the so-called ADASIS Forum, with the European organization ERTICO acting as coordinator. Up-to-date information on how to implement this specification in detail on a CAN bus, in particular also for defining the corresponding CAN identifiers, is available on the website www.ertico.com.

The methods mentioned are also used if, instead of a navigation device, a cost-optimized special control device without a user interface is used as the horizon provider.

Based on the map data, current navigation systems determine a most probable travel path MPP by analyzing the turn-off probabilities, which are determined by road classes, turn angles, road numbers or the like. This is however quite inaccurate, since it does not correspond to the typical driving behavior. The reliability of the MPP can be increased by evaluating the previous behavior of the driver and including it in the calculation of the MPP. This results in an 'improved probable driving path'. The improved probable driving path can be used to optimize various functions, such as the targeted control of the front light, which can illuminate the crossing area in a turning process accordingly.

The inclusion of driver behavior is so far but only within the navigation system or the component that provides the horizon (horizon provider). Accordingly, the function of improving the MPP by taking account of stored driver behavior data has hitherto been linked to a component which comprises at least one digital map, if not an operator interface, in a relatively complex and less flexible manner. If impersonal driving data provided by a community server is used to improve the MPP, consideration is given to this Community data so far also within the horizon provider.

Disclosure of the invention

Against this background, an inventive method according to claim 1, an arrangement according to claim 11 and a computer program product according to claim 13 are presented. Further embodiments of the invention will become apparent from the dependent claims and the description.

In the inventive method characterized in claim 1, beyond the generic features, it is provided that a first module prepares the original MPP and provides a 2D horizon on the basis of the original MPP and that a second module corrects the original MPP Provides driving probability data derived from the previous driving behavior of at least one vehicle. In the driving behavior of the vehicles, of course, possibly the driver behavior of several drivers is formed. In particular, if 'personal' or vehicle-specific driving likelihood data are to be provided, it is possible by simple measures to even provide individualized driving likelihood data by using only the driver behavior of a selected driver as the database.

By means of the two separate modules, a spatial or functional decoupling of the memorization and driver behavior evaluation functions from the horizon provider or from the digital map succeeds, so that the possibility of improving the MPP by including the driving behavior according to the invention no longer affects the navigation system or the horizon provider is limited. According to the invention, the second module may, for example, be an autonomous component of a navigation system or another control device in the vehicle independent of the horizon provider. The first or second module, or both modules, may also be located outside the vehicle, in a community, that is, a server-based road user group sharing a common knowledge. For example, the first module, the horizon provider, can sit in the community and receive the current position of the vehicle constantly.

Incidentally, according to the present invention, "road map" should not necessarily be understood to mean a complete digital map, it is sufficient if point-by-point defined maneuvering probabilities exist, Furthermore, as used herein, the term "2D horizon" may also comprise one another, So three-dimensionally arranged driving paths.

According to a first embodiment of the invention, the second module provides personal driving likelihood data by providing driver behavior data, e.g. from previously evaluated horizon data of past journeys of this vehicle originate on the same road segment, or the second module have been made available in an appropriate manner evaluates.

According to a further embodiment of the invention, the second module provides impersonal driving likelihood data by collecting in the community driver behavior data of many vehicles, possibly also of the vehicle in question and this driver behavior data to the second module (in the community or in the vehicle) for generating the driving likelihood data be transmitted.

In order to achieve an optimal correction of the MPP, it is proposed according to a further embodiment of the invention to use not only the impersonal driving probability data from the community but also (locally or centrally stored) personal driving probability data for the correction of the MPP. In addition to the improvement on routes unknown to the individual driver by means of the impersonal data, an already known driving route can thus lead to an individual improvement by means of the personal data in this way. In addition, any data protection requirements may be covered. The implementation of the evaluation of impersonal and personal driving likelihood data can take place by means of a preferably arranged in the community, impersonal driving probability data first part module of the second module, and a personal driving likelihood data providing second submodule of the second module, which can be located both in the community and in the vehicle ,

According to the invention, this driving behavior data can be used in various ways to correct the original MPP.

According to a first embodiment of the invention, the first module transmits the 2D horizon to the second module, whereupon the second module uses the previously collected driver behavior data and derived driving likelihood data to establish a corrected probable travel path in the received 2D horizon and the thus corrected 2D horizon for further use. The second module thus becomes the horizon provider itself. However, the corrected probable travel path can then only be within the framework of the received 2D horizon which may lead to a shortened probable route.

According to a further embodiment of the invention, the first module transmits the 2D horizon to the second module, whereupon the second module uses the driving likelihood data to correct the original MPP contained in the received 2D horizon and transmits the corrected probable travel path back to the first module. From here, various possibilities open up. In a simple further embodiment of this embodiment, the first module may use the received corrected probable travel path to create a corrected probable travel path in the 2D horizon and provide the thus corrected 2D horizon for further use, but possibly again at the price of a shortened probable one driving path.

In a further advantageous embodiment of the embodiment in which the corrected probable travel path is returned to the horizon provider, the first module may use the received corrected probable travel path to create a customized 2D horizon, using the adjusted 2D horizon Probable Fahrpfad is marked with a comparison with the corrected probable travel path enlarged look-ahead length. The first module then makes the 2D horizon so adapted for further use. The 2D horizon adjusted in this way represents a considerable improvement over the only corrected 2D horizon due to the increased look-ahead length of the resulting probable travel path. The resulting probable travel path or the adjusted 2D horizon can be further iteratively further improved by the first Module transmits the adjusted 2D horizon again to the second module for further correction of the current probable Fahrpfads.

In order to avoid bandwidth-critical transmission processes, it is proposed according to a further embodiment of the invention that the second module provides driving likelihood data related to significant intersection coordinates and transmits it to the first module, and that the first module uses the received lane data to generate a corrected probable driving path. Based on this corrected probable travel path, the first module then creates a 2D horizon which is provided for further use or for further correction.

The invention will be described in more detail below with reference to exemplary embodiments. Show

1 a schematic visualization of a 2D horizon based on the original MPP,

2 in the same representation the 2D horizon according to 1 but with a probable travel path corrected according to the invention,

3 a block diagram of a vehicle navigation device as an example of an arrangement for carrying out the method according to the invention.

In the 1 illustrated visualization of the data of the electronic horizon contains the road segments of the (in the 1 and 2 as a double line with solid and parallel dashed lines) original MPPs 10 and optionally further (sub) branches 11 , which can each be enriched with additional data (for example, allowed speeds). If the horizon, as shown, (sub) branches / paths 11 contains, one also speaks of a two-dimensional (2D) horizon. The 2D horizon is typically so pronounced that, as in FIG 1 recognizable, the probable routes, in particular the MPP 10 , have a longer lookahead than the unlikely (secondary) paths 11 ,

The navigation system determines the original MPP 10 and then generates a 2D electronic horizon, as in 1 shown next to the original MPP 10 also the branches and side paths 11 of the road network. This 2D horizon is communicated in several embodiments of the invention to an independent software module, which may be implemented in the navigation system, in another in-vehicle control unit, or in a community-located controller. For example, a community can be constituted by the participating vehicles of a particular car manufacturer. For example, subscribers are connected and registered with the community server via the Internet or direct telephone dial-in. A typical application is traffic jam detection: if many participants report to the community server that they are traveling too long, the server reports the congestion risk to all participants.

The block diagram of 3 shows a simplified illustrated navigation device 1 that a digital map 2 , a locating device 3 , an operating device 4 and a central control 5 includes. A first module 6 , the horizon provider, which is typically a software module, from a computer of the central control of the navigation system 1 is processed, calculated by means of a predetermined algorithm from the present map data 2 and that of the location 3 delivered vehicle position an original MPP 10 , As a result, the horizon provider calculates 6 based on this MPP a 2D horizon, such as in 1 is visualized. A second module, the driver behavior component 7 More specifically, in the case of embodiments with two sub-modules: the driveability component 7 , is preferably also as - from the horizon provider 6 independent software module trained and provides for the correction of the original MPPs 10 suitable driving probability data ready. The two modules 6 and 7 can cooperate according to the invention in different ways, for example, a communication channel 8th for transmission from the horizon provider 6 created 2D horizon to the second module 7 , as well as a return channel 9 for returning the corrected probable travel path from the second module 7 to the horizon provider 6 provided interoperability must be ensured.

The second module 7 , which receives the original 2D horizon, can be used to evaluate and enhance the original MPP 10 to collect own driver-related data of the past routes. The second module 7 stores, for example, data that it has extracted during previous trips from the electronic 2D horizon. These data are stored with a suitable reference so that they can be reassigned later. Based on these data, individual turn off probabilities can be determined for each multiply traveled intersection. With the aid of these driving likelihood data, the turn-off probabilities, ie the original MPP, are checked in the received 2D horizon and improved or individualized. This can be done, for example, by evaluating turn-off frequencies as a function of the day of the week and the time of day.

Thus, in the second module 7 a more accurate, corrected probable travel path available, as well as, according to the embodiment considered here, the original 2D horizon. The initial turn-off probabilities of the original MPP 10 according to 1 are corrected in the original 2D horizon. The changed turn-off probabilities may result in a corrected probable travel path 12 , compare 2 who is now in a sub-branch 11 running. This results in a relatively short corrected probable travel path in this embodiment 12 because the sub-branch 11 in the 2D horizon was not as pronounced as the initial MPP 10 , The original 2D horizon, the second module 7 is given in this embodiment is thus only corrected, but not adapted, that is, not suitable for the corrected probable driving path 12 newly created. The corrected 2D horizon according to 2 is made available modules or a driver assistance system for further use, in particular by means of new CAN IDs (identifiers) in ADASIS format, so that the other systems can work based on the improved data.

In particular, possibly shortened probable driving paths 12 , as in 2 to avoid the driving likelihood data based on the evaluation of the driving behavior, typically the turning probabilities (ie the corrected probable driving path 12 ), from the second module 7 to the horizon provider 6 be returned so that it can provide an optimized 2D horizon based on the improved driving likelihood data. For this purpose, several methods are conceivable.

Required is the specification of a return channel 9 , eg in ADASIS format, to ensure interoperability. The return channel 9 transmits, for example, the location-referenced learned turn probabilities, that is, only the information of the corrected probable travel path relevant to the correction of the original MPP 12 which are then appropriately incorporated in the recalculation of the 2D horizon. The improved probable route 12 is provided, for example, via new CAN IDs in ADASIS format. The horizon provider 6 can then provide a new 2D horizon with adjusted turn-off probabilities where the probable travel path is longer than in 2 is pronounced. The correction process can be iterated, that is, the new, adjusted 2D horizon is provided by the horizon provider 6 back to the second module 7 transmitted. If an MPP deviation is detected again within the new horizon links, the process may be repeated until the probable travel path of the horizon provider 6 is sufficiently long, or the turn-off probabilities are sufficiently accurate.

In order to avoid the bandwidth critical iterative method described above, the driver behavior component, in particular in the case of probable travel paths requiring a large lookahead length, could be used 7 via the ADASIS format (return channel 9 ) Driving likelihood data in the form of certain significant crossing coordinates along which the horizon provider 6 then can express a corrected probable travel path, eg by means of route calculation methods between the coordinate points. In this case, the driver behavior module transmits 7 So without first received 2D horizon, so to speak, initially self-', a, coarse-grained' version of the driver behavior-related driving likelihood data, for example, includes only turn-off probabilities regarding important intersections or departures. The probabilities on the 2D horizon, now pronounced along the corrected probable travel path, may then be adjusted in a further step, as in the embodiment described above, to result in an adjusted 2D horizon.

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• DE 102009028299 A1 [0002]
• DE 102007043533 A1 [0003]

Claims (13)

A method for determining a probable driving path of a vehicle by at least one module in which an initial probable driving path (MPP) is determined in consideration of driving likelihood data derived from a road map and in which to correct the original probable driving path (MPP) Data are used, characterized in that a first module ( 6 ) the original probable route ( 10 ) and based on the original probable route ( 10 ) provides a 2D horizon, and that a second module ( 7 ) for correcting the original probable driving path ( 10 ) Provides driving probability data derived from the previous driving behavior of at least one vehicle. Method according to claim 1, characterized in that the second module ( 7 ) for correcting the original probable driving path ( 10 ) provides personal driving likelihood data derived from the previous drivability of that vehicle. Method according to claim 1, characterized in that the second module ( 7 ) for correcting the original probable driving path ( 10 ) provides impersonal driving likelihood data derived from previous drivability of this and / or at least one other vehicle. Method according to claims 2 and 3, characterized in that the second module ( 7 ) for correcting the original probable driving path ( 10 ) provides impersonal and personal driving likelihood data, wherein a first submodule of the second module, preferably arranged outside this vehicle in a community ( 7 ) the impersonal and a second submodule of the second module, preferably arranged in this vehicle ( 7 ) provides the personal driving likelihood data. Method according to one of claims 1 to 4, characterized in that the first module ( 6 ) the 2D horizon to the second module ( 7 ) and that the second module ( 7 ) the driving likelihood data for creating a corrected probable driving path ( 12 ) in the received 2D horizon and provides the thus corrected 2D horizon for further use. Method according to one of claims 1 to 4, characterized in that the first module ( 6 ) the 2D horizon to the second module ( 7 ) and that the second module ( 7 ) the driving likelihood data for correcting the original probable driving path contained in the received 2D horizon ( 10 ) and the corrected probable travel path ( 12 ) to the first module ( 6 ). Method according to claim 6, characterized in that the first module ( 6 ) the received corrected probable travel path ( 12 ) is used to create a corrected probable travel path in the 2D horizon and provides the thus corrected 2D horizon for further use. Method according to one of claims 1 to 4, characterized in that the second module ( 7 ) provides driving likelihood data related to significant intersection coordinates and to the first module ( 6 ) and that the first module ( 6 ) the received driving likelihood data for creating a corrected probable driving path ( 12 ) used. Method according to claim 6 or 8, characterized in that the first module ( 6 ) the received or created corrected probable travel path ( 12 ) is used to create a customized 2D horizon, wherein in the adjusted 2D horizon, a probable travel path having a (compared to the probable probable path (FIG. 12 ), and that the first module ( 6 ) makes the adjusted 2D horizon available for further use. Method according to claim 9, characterized in that the first module ( 6 ) the adapted 2D horizon for further correction of the current probable travel path according to claim 6 to the second module ( 7 ) transmitted. Arrangement, in particular navigation device ( 1 ), with at least one computer for determining a probable driving path ( 10 . 12 ) according to a method according to any one of the preceding claims. Arrangement according to claim 11, characterized in that the arrangement comprises two separate control devices, in each of whose computers each one of the two modules ( 6 . 7 ) is designed as a software module. Computer program product with electronically readable control signals, which are thus provided with a programmable computer system, in particular a navigation device ( 1 ), that can work together a method according to any one of claims 1 to 10 is carried out.

Images