Method and system for determining a navigation route for vehicles
Patent Information
- Authority / Receiving Office
- EP · EP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2024-07-30
- Publication Date
- 2026-03-25
Smart Images
Figure EP2024071544_13022025_PF_FP_ABST
Abstract
Description
[0001] Method and system for determining a navigation route for vehicles
[0002] The invention relates to a method for determining a navigation route for vehicles according to the type defined in more detail in the preamble of claim 1 and to a system according to the type defined in more detail in the preamble of claim 12.
[0003] Participation in road traffic always carries an accident risk. This accident risk will be reduced in the future due to the increasing degree of automation of vehicles, up to and including fully autonomous control. The accident risk varies across different road sections and at different times. The accident risk can increase particularly on busy road sections and in challenging environments, such as construction sites or in areas with so-called "sign forests." Furthermore, people are more or less capable of driving their own vehicles safely. The accident risk therefore increases when there are people on the road section who have less control of their vehicles or who tend to drive dangerously.
[0004] Vehicle users strive to avoid stressful driving situations and accidents, which creates the need to provide means to increase safety when participating in road traffic.
[0005] DE 102021 000 596 A1 discloses a method for adapting a vehicle's driving behavior. The document describes observing other road users and determining the vehicle's acceleration capability or deceleration.
[0006] The deceleration capability of individual road users is determined based on the observations made. The vehicle's handling is adjusted to the detected acceleration or deceleration capability. For example, if it is detected that a slow-moving vehicle ahead has a low engine power, an overtaking maneuver can be initiated at an appropriate location. Other road users can be detected by sensors mounted on static infrastructure, on the vehicles of a fleet, or even on the vehicle itself.
[0007] Furthermore, DE 102019 103 554 A1 describes a device for detecting potentially dangerous curves.
[0008] Furthermore, DE 102018217828 A1 discloses a method and a device for adapting the driving behavior of a partially, highly, or fully automated vehicle. The driving behavior is adapted depending on interior data, wherein the interior data describe a safety status of a vehicle occupant, in particular a risk of injury.
[0009] In addition, US 2018 / 0107942 A1 discloses a framework for learning a group-specific driving style for autonomous vehicles. This involves recording the driving behavior of vehicles driving in the vicinity of the autonomous vehicle. Information describing the respective driving behavior is transmitted by the autonomous vehicle to a server. The server classifies the driving behavior based on this information and then tells the autonomous vehicle how it should behave in order to imitate the driving behavior.
[0010] The present invention is based on the object of providing an improved method for determining a navigation route for vehicles, with the aid of which vehicles can be steered particularly safely through road traffic.
[0011] According to the invention, this object is achieved by a method for determining a navigation route for vehicles having the features of claim 1. Advantageous embodiments and further developments as well as a system for implementing the method emerge from the dependent claims.
[0012] A generic method for determining a navigation route for vehicles, wherein other road users are detected with the aid of a sensor system and sensor data generated by the sensor system are evaluated by a computer device to determine the behavior of the respective road users in road traffic, is further developed according to the invention in that
[0013] - the computing device maintains a database comprising behavioral thresholds; - the computing device compares the behavior observed by road users with the behavioral thresholds stored in the database and counts an event on the road section on which the road users are located if the observed behavior of a road user exceeds the corresponding behavioral threshold;
[0014] - the computing device calculates a hazard index for at least one road section depending on the number and severity of the events counted on the respective road section within a specified period; and
[0015] - a navigation unit which takes into account the hazard indices assigned to the respective road sections for route planning, whereby, in order to determine the navigation route, those road sections are avoided whose hazard index exceeds a specified hazard index threshold.
[0016] The method according to the invention allows potentially dangerous road sections to be identified and excluded from the calculation of navigation routes. This guides vehicles to their respective destinations along safe road sections, which not only increases comfort for the drivers, as stressful driving situations are avoided, but also reduces the overall risk of accidents.
[0017] The respective vehicle users can program the navigation route before starting their respective journey. A vehicle-integrated navigation system is preferably used for this purpose. Dedicated mobile devices such as mobile navigation systems or smartphones with a navigation app can also be used. A respective navigation unit, particularly in the form of a navigation system, then establishes a communication connection to the computing device and obtains the respective hazard indices of the road sections potentially suitable for a navigation route.
[0018] The computing device can also be integrated into the vehicle. In this case, several computing devices implemented in different vehicles can interact as a single large computing device. This is particularly preferably a central computing device, i.e., a server or server network.
[0019] The sensors used to monitor road users can be integrated into static infrastructure, the vehicles of a fleet, and / or even the vehicle itself. Examples of possible sensor systems include cameras, laser scanners such as LiDARe, radar systems, and / or ultrasonic sensor systems. Such sensor systems can be used to obtain depth information and thus scan the surroundings. This allows a computer-usable representation of the environment to be generated. Image recognition algorithms, i.e., machine vision, can be used to analyze, classify, and evaluate situations in a particularly sophisticated manner. Other vehicle sensors, such as acceleration sensors, wheel speed sensors, and the like, can also be used. Sensor data can also be retrieved from driver assistance systems such as ABS, ASR, airbag control units, and the like.For example, the driving speed of a vehicle, its acceleration, the occurrence of an accident, the detection of wheel spin and the like can be determined.
[0020] A specific vehicle in traffic can also be identified and tracked based on visual characteristics. Furthermore, external conditions can be identified, such as presence in a construction zone, the detection of a forest of signs, the presence of adverse weather and / or visibility conditions, and the like.
[0021] Other road users can also be vehicles such as cars, trucks, vans, buses, or even small vehicles such as bicycles, scooters, skateboards, etc. Pedestrians can also be identified and considered as road users.
[0022] The behavior of road users can be assessed in a concrete way. To specifically assess behavior, the driving maneuvers (or walking maneuvers) performed by the respective road user can be evaluated. This affects, for example, the acceleration and steering behavior of the road user, i.e., how quickly the respective road user accelerates or decelerates, or how jerkily their steering maneuvers are performed. Furthermore, the distance maintained from other road users and / or the current speed of travel can be recorded and evaluated. The behavior thresholds stored in the database correlate with the detectable behavior. Behavioral thresholds can be defined depending on the respective environmental conditions or driving situation.For example, the current speed limit of the road section being traveled can be used as a behavioral threshold for the tolerable maximum speed of a road user. Different speed limits apply on different road sections, so different behavioral thresholds can also apply to different road sections. At a motorway entrance or exit, correspondingly higher behavioral thresholds can be considered for the tolerated vehicle acceleration than, for example, in city traffic.
[0023] Other examples of observable and assessable behavior include: tailgating, risky overtaking, improper use of high beams, failure to observe a one-way street, crossing lane markings or the central reservation, improper use of the horn, malicious gesticulation, driving on the hard shoulder, driving through a traffic light towards the end of a yellow phase or even when it is red, and the like.
[0024] The driver's own vehicle may be equipped with means to detect the driver's level of fatigue or distraction, so that road sections with a particularly high number of vehicles whose drivers are tired or distracted can be assessed as dangerous.
[0025] The vehicle can be positioned relative to its surroundings in traffic. Using these sensors, other road users can be detected. This depth information can be used to determine the relative distances to other road users over time. From this, the corresponding distances and speeds of other road users, and thus their respective behavior, can be derived.
[0026] The hazard index takes into account both the absolute number of incidents recorded on the respective road section during the specified period, as well as their severity. For example, each incident can be assigned a number to represent its severity. The more serious the incident, the higher the severity can be.
[0027] The specified time period can be chosen differently depending on various boundary conditions. For example, the specified time period can be a few hours, days, weeks, months, or even years. The specified time period can also cover the entire past observation period, i.e., from the beginning of the behavioral observation up to the current point in time. The specified time period can also depend, for example, on traffic density. For example, a different hazard index can be determined for one and the same road section if there are a particularly high number of road users on the road section, for example during rush hour, than compared to a period with low traffic volume. Changes in the road network can also be taken into account.For example, if a specific road section is closed, e.g., due to construction work, this can lead to road users choosing a preferred alternative route. As soon as the closure occurs, the hazard index for the road sections of the preferred alternative route can be recalculated. Accordingly, the specified period coincides with the road closure period.
[0028] The specified period can also be a sliding time horizon, for example, the last 15 minutes. Historical data can also be considered to determine the hazard index. For example, the same time frame can be considered every day, such as 8 a.m. to 10 a.m., and an average can be calculated for the respective time frame over several days.
[0029] The hazard index can be expressed as any numerical value, ranging from 0 to 1, 0 to 10, 0 to 100, or the like. For example, a hazard index of 0 corresponds to a safe road section, and a hazard index of 1, 10, or 100 corresponds to a particularly dangerous road section.
[0030] The hazard index threshold can also be set to different levels depending on various conditions. For example, the hazard index threshold for a road section can be set manually by the navigation unit user. For example, a hazard index threshold of 5, 6, or 7 can be set. The hazard index threshold can also be selected automatically, for example, depending on the number of available alternative routes. If many alternative routes are possible, a particularly low hazard index threshold can be selected to give priority to one of the possible alternative routes. If, on the other hand, fewer potential alternative routes are available, the hazard index threshold can be increased so that preference is still given to driving on the potentially dangerous road section.If the person driving the vehicle deviates from the navigation route set in this way and drives along a section of road whose hazard index exceeds the set hazard index threshold, warnings can be issued in the vehicle, such as the sounding of certain warning tones or the flashing of certain warning lights or the appearance of visual warning messages on a display device.
[0031] Particularly dangerous road sections, i.e., sections that frequently exhibit a comparatively high hazard index, can also be communicated to third parties. For example, traffic authorities can be informed about frequent violations of the driving limits. This allows a traffic authority, for example, to implement appropriate safety measures. For example, more speed enforcement devices, such as speed cameras, can be installed.
[0032] An advantageous development of the method according to the invention provides that the computing device assigns a rating to a respective road user when a behavior threshold is exceeded and takes into account the rating of the road users present on the road section during the specified period to calculate the hazard index. This makes it possible to identify road users who contribute to an increase in the risk or danger of a road section. These road users can be tracked temporally and spatially, whereby the hazard index of a specific road section can be adjusted live and particularly reliably. If, for example, there are currently a particularly large number of road users with a particularly high rating on the respective road section, the corresponding hazard index can also increase. The rating also has a value.For example, the hazard index of a road section can also include the currently accumulated value of all ratings for that particular road section. High ratings, for example depending on the length of the respective road section, can be tolerated before the hazard index is incremented, since there tend to be more vehicles on a long road section than on a short road section. Since the vehicle density remains similar, more "dangerous" drivers can be tolerated on a long road section. According to a further advantageous embodiment of the method according to the invention, the value of the event assigned to a road section and / or the rating assigned to a road user depends on the extent to which the associated behavioral threshold is exceeded.If certain road users exhibit particularly risky behavior, the danger also increases when driving on the section of road on which they are currently traveling. Road sections or road users who exceed the applicable speed limit by, say, 40 percent are assigned a higher-level event or rating than road sections or road users who only exceed the applicable speed limit by, say, 10 percent. This allows for a particularly accurate determination of the respective hazard index for each road section. The hazard index is thus determined, for example, for the respective road section within the specified period by summing the events multiplied by the respective hazard index.
[0033] A further advantageous embodiment of the method according to the invention further provides that the behavioral limits are defined in accordance with the applicable road traffic regulations. In each country, the applicable road traffic regulations regulate which behavior is permissible in road traffic and which must be punished. By selecting the behavioral limits in accordance with the applicable road traffic regulations, applicable limits can be defined for the respective country. For example, different road types such as urban areas, country roads, or motorways have different speed limits, such as 50 km / h, 100 km / h, or 130 km / h. For example, a safety distance dependent on the driving speed can be prescribed, which is taken into account accordingly.A particular road traffic regulation can also, for example, regulate in which situations the horn may be honked, the high beams may be used, or stipulate that suggestive gestures such as showing the middle finger may be punished with a fine.
[0034] According to a further advantageous embodiment of the method according to the invention, the computing device assigns to a road section one of its values based on the current total number of road users on the road section, the average distance between road users, in particular taking into account its standard deviation, and / or an event dependent on the number of lane changes on the road section determined during a defined period. This enables an even more accurate determination of the hazard index of the respective road section and thus the risk assessment of the respective road section. If traffic density on the road section increases, the risk of an accident generally also increases. If the respective road users drive particularly close to one another, the risk of a rear-end collision also increases.If some road users deviate significantly from the behavior of other road users, for example, by maintaining an excessively large or excessively small safety distance, this can further increase the risk of an accident. Changing lanes is a particularly risky maneuver, especially in city traffic, because the driver must carefully monitor the traffic behind them to avoid a collision with another road user in their blind spot. If a particularly high number of lane changes are detected on a particular road section during a certain time interval, this also indicates that particularly risk-averse drivers are present on that road section.
[0035] A further advantageous embodiment of the method according to the invention further provides that the computing device determines situational characteristics present from the sensor data when a behavioral threshold is exceeded and executes a machine learning model that establishes a relationship between the determined situational characteristics and a respective exceedance of a respective behavioral threshold. Situational characteristics describe the boundary conditions present in the respective traffic situation, such as a low sun, a vehicle user operating a radio, being in a construction zone, being distracted by a mobile device such as a smartphone, and the like. Situational characteristics can be determined from the sensor data using proven methods. In particular, machine vision methods can be used for this purpose.It is also possible to use analysis methods based on artificial intelligence.
[0036] With the help of artificial intelligence in the form of a machine learning model, a connection is then established between the situational characteristics present in the respective situation when the behavioral threshold is exceeded. This enables the computer to identify situations associated with a particularly high hazard index and thus an accident risk. For example, if adverse visibility conditions exist, such as due to a wet road surface and a low sun, the computer can increase the hazard index before the respective behavioral thresholds are exceeded.
[0037] According to a further advantageous embodiment of the method according to the invention, this enables the computing device to predict the hazard index of at least one road section for a future time horizon. If a vehicle user plans a journey from a starting point to a destination, the navigation route can be determined by taking into account not only the currently available hazard indices of the road sections, but also the hazard index expected to occur upon reaching the respective road section. This enables an even more accurate assessment of the risk potential of the respective road sections and thus the determination of an even more convenient and safe navigation route.If, for example, the navigation route found contains a road section towards the end on which there is currently a high danger index due to increased traffic volume during rush hour, but the rush hour is over when the vehicle arrives on the road section, the respective road section can still be included in the navigation route due to the low danger index when the vehicle arrives on the road section.
[0038] Predicting hazard indices is also possible without the use of machine learning models. For example, the history of hazard indices occurring on the respective road sections can be taken into account. For example, an average hazard index value can be calculated for a specific period of time on a road section. To do this, for example, for the individual road sections, the hazard index that has occurred on the road section for each hour of the day, for example, from 8 to 9 a.m. and from 9 to 10 a.m., etc., is used, and an average is calculated from this. For example, the hazard index for 10:45 a.m. can be predicted as early as 8:30 a.m.
[0039] If a rating is assigned to a road user, an expected navigation route can also be determined for this road user. For example, the computer can retrieve and analyze the navigation route set in the respective road user's vehicle, or the respective navigation route can be estimated. For example, if a lane change from the left lane to the right lane and a reduction in speed is detected on a motorway, the computer can estimate that the respective road user will soon exit the motorway. Accordingly, when the risk-averse road user leaves the motorway, the hazard index can be reduced for the sections of the motorway following the exit and increased for the sections of a country road following the exit, for example.
[0040] A further advantageous embodiment of the method according to the invention further provides that the computing device determines existing situational characteristics from the sensor data when a behavioral threshold is overwritten and adapts the level of at least one behavioral threshold depending on at least one situational characteristic. This makes it possible to tolerate "poor" or "unlawful" driving behavior depending on the situation if the corresponding violation would be usual or unavoidable due to external conditions. If, for example, vehicles frequently brake particularly abruptly before a certain traffic light, this may be due to the traffic light's brief green phase and not to the risky driving style of the respective road users. To take this into account, the hazard index cannot be increased accordingly for the road section containing this traffic light.However, since this can generally be an unexpected maneuver, the danger index can optionally be increased. The preferred approach can also be set by the user according to their preferences.
[0041] According to a further advantageous embodiment of the method according to the invention, the navigation unit increases the specified hazard index threshold for at least one road section of a navigation route determined disregarding the hazard index if the path-related and / or time-related length of at least one road section of the navigation route calculated taking the hazard index into account exceeds a specified length threshold and then recalculates the navigation route. This allows a user to define preferences for a tolerated detour or the resulting time delay, so that if the detour is too long, the navigation route still leads along potentially dangerous road sections. For this purpose, the "normal" navigation route, i.e., a navigation route calculated according to standard procedures, is first determined.This serves as a comparison to determine any increase in the route length, both in terms of distance and time, compared to the navigation route calculated taking the hazard index into account. The user can specify the extent of tolerated detours in various ways. For example, the user can specify an absolute distance-related and / or time-related tolerance. For example, the user can enter that the navigation route calculated taking the hazard index into account may be a few minutes, hours, or even fractions or multiples thereof longer. This value can also depend on the total length of the navigation route; for example, the navigation route may be up to 5 percent, 10 percent, or even fractions or multiples thereof longer in terms of distance and / or time due to the detour. The user can also specify a ratio of the hazard index to the tolerated increase in length.For example, the driver can specify that routes with a hazard index of > 0.7 should be avoided, resulting in a travel time or distance increase of up to 20 percent. Accordingly, if the navigation route initially calculated taking the hazard index into account is too long, i.e., the length threshold is exceeded, the navigation unit increases the hazard index threshold for at least some road sections. This increases the probability that road sections with a higher hazard index but contributing to a shorter travel time or distance will be included in the newly calculated navigation route.
[0042] If it is not possible to calculate a suitable navigation route while complying with the boundary conditions specified by the user, the navigation unit can also display an error message, such as: “No suitable navigation route could be found with the specified settings”.
[0043] A further advantageous embodiment of the method according to the invention further provides that the vehicle follows the navigation route with at least partially automated control. The vehicle can be at least partially automated or even autonomously controlled. The vehicle is thus capable of following the navigation route determined by the navigation unit, either independently or at least with assistance. The vehicle then travels on road sections characterized by a reduced risk of traffic accidents due to the reduced hazard index. This reduces the risk of the at least partially automated vehicle being involved in an accident. This leads to the safer operation of at least partially automated vehicles.
[0044] According to a further advantageous embodiment of the method according to the invention, the vehicle adopts increasingly defensive driving behavior as the hazard index increases. This also reduces the risk of being involved in an accident for at least partially automated vehicles. For example, the vehicle behaves "particularly cautiously" on dangerous road sections, allowing, for example, a reduced driving speed to be maintained, thereby shortening braking distances and / or increasing a safe distance, thus reducing the risk of rear-end collisions.
[0045] In a system comprising a sensor system, at least one vehicle, a computing device and a navigation unit, according to the invention the sensor system, the vehicle, the computing device and the navigation unit are set up to carry out a method as described above. The sensor system can be part of stationary infrastructure, integrated into the vehicles of a vehicle fleet or even into the vehicle that travels along the navigation route. The vehicle can be any road vehicle, for example a car, truck, van, bus or the like. The navigation unit can be a dedicated navigation system. The navigation unit can also be designed with special navigation software that is executed on the computing device. The navigation unit is able to determine the current position of the vehicle orof the navigation unit. The computing device can be integrated into the vehicle. However, the computing device can also be implemented as a central computing device external to the vehicle, for example, as a cloud server. The computing units integrated into the vehicles of the vehicle fleet can also interact as a large, interconnected computing device.
[0046] Further advantageous embodiments of the method according to the invention for determining navigation routes for vehicles and of the system according to the invention also emerge from the exemplary embodiments which are described in more detail below with reference to the figures.
[0047] The following show: Fig. 1 a schematic plan view of a traffic situation;
[0048] Fig. 2 is a bar chart showing the dependence of a hazard index on the number and severity of violations of behavioral thresholds occurring on a road section during a specified period; and Fig. 3 is a schematic representation of a digital road map with two differently determined navigation routes.
[0049] With the aid of a method according to the invention for determining navigation routes 1 for vehicles, as shown in Figure 3, it is possible to avoid potentially dangerous road sections, allowing the respective vehicle to be steered particularly safely through traffic. The vehicle can be steered manually or at least partially automatically or even autonomously.
[0050] The method according to the invention is based on observing the behavior of road users 2 shown in Figure 1 with the aid of a sensor system 3. Sensor data generated by the sensor system 3 are evaluated by a computing device 4, shown as a cloud server in Figure 1. As Figure 1 shows, the sensor system 3 can be integrated into stationary infrastructure and / or be part of a respective vehicle participating in traffic.
[0051] The computing device 4 comprises a database 5 in which behavioral limits are stored. The respective behavioral limits describe the permissible behavior of road users 2 in traffic using concrete reference values, such as maintaining a certain driving speed, a certain minimum distance from one another, a number of overtaking maneuvers performed per unit of time, limits for tolerated acceleration values, and the like. The behavioral limits are preferably determined in accordance with the road traffic regulations applicable in the respective country.
[0052] In the embodiment shown in Figure 1, a vehicle 6.2 following a leading vehicle 6.3 in a passing lane falls below the minimum distance predefined according to the behavioral thresholds. This exceedance is recorded both by the camera mounted in a stationary infrastructure and by the camera of an ego vehicle 6.1. The following vehicle 6.2 is equipped with a distance radar, which also detects the infringement of the minimum distance.
[0053] The respective road users 2 and the static infrastructure transmit the sensor data generated by the sensor system 3 to the computing device 4 for evaluation. This can be done wirelessly or via a cable. In particular, the road users 2 transmit the sensor data wirelessly, and the stationary infrastructure transmits it via a cable.
[0054] The computing device 4 determines the respective behavior of road users 2 from the sensor data and compares this with the behavioral threshold values stored in the database 5. The computing device 4 then assigns an event E, shown in Figure 2, to the road section on which the behavioral threshold value has been detected being exceeded. The significance of the event E is preferably selected depending on the extent to which the behavioral threshold value has been exceeded. This means, for example, that the further the following vehicle 6.2 falls short of the minimum distance from the vehicle 6.3 in front, the greater the significance of the event E increases. If the following vehicle 6.2 falls short of the minimum distance by one meter, for example, the significance of the event E could be "2". If, on the other hand, the following vehicle 6.2 falls short of the minimum distance by, for example, five meters, the significance of the event E could be "3" or even more.
[0055] The computing device 4 aggregates the respective events E for a respective road section over a period of time individually defined for each road section. Depending on the number and significance, the computing device 4 determines the hazard index Eq. shown in Figure 2.
[0056] Figure 2 shows an example of the dependence of the hazard index Gl on the respective events E multiplied by the value W (ExW). The hazard index Gl is plotted on the abscissa of the diagram and the events E multiplied by the value W on the ordinate. A value of 30 on the ordinate can arise from any conceivable combination of events and associated values, for example from 30 events with a value of 1 or from 10 events with a value of 3. For example, all events E multiplied by the value recorded on a given road section during a specified period of time are summed up. As Figure 2 shows, the hazard index Gl increases gradually with the value of the events E multiplied by the value W. A continuous increase shown by a straight line 7 would also be possible.The hazard index Gl could also increase progressively or degressively (not shown) depending on the events E multiplied by the value W. The hazard index typically resulting from historical data on weekdays between 8 a.m. and 9 a.m. on a given road section is indicated by a hatched bar.
[0057] Various formulas and rules can be defined for determining the hazard index Gl for a given road section, depending on the number and severity of events E. Different rules can also apply to different road sections. For example, an individual rating can be assigned to each road user 2, which is then incorporated into the hazard index Gl for the road section on which the respective road user 2 is currently located.
[0058] A navigation unit takes the hazard index Gl determined in this way into account for route planning. Figure 3 shows an initial navigation route 1.1, i.e., calculated without taking the hazard index Gl into account, and a navigation route 1.2 calculated taking the hazard index Gl into account. The hazard index Gl that applies to each road section is plotted for each road section. For example, 0.6 was stored as the hazard index threshold. This results in the road section with the hazard index Gl 0.7 being omitted from navigation route 1.2, and the corresponding detour being suggested. The respective user can specify how high the respective hazard index threshold should be, in particular as the ratio of a tolerated path- or time-related length increase to the hazard index threshold.In this way, potentially dangerous road sections can still be included in the respective navigation route 1, should the increase in travel distance or travel time due to the inclusion of the detour be unacceptably high.
[0059] The navigation unit can be integrated into a respective vehicle or be designed as a dedicated mobile device. Thus, mobile navigation systems or, for example, a navigation app running on a smartphone can also be used as a navigation unit. The respective navigation unit only requires a connection to the computing device 4 for obtaining the respective hazard indices Gl and a correspondingly adapted route determination algorithm. A corresponding computer program product and a corresponding computer-readable storage medium for storing the computer program product are thus also part of the invention.
Claims
Patent claims 1. A method for determining a navigation route (1) for vehicles, wherein further road users (2) are detected by means of a sensor system (3) and sensor data generated by the sensor system (3) are evaluated by a computing device (4) to determine the behavior of the respective road users (2) in road traffic, characterized in that - the computing device (4) maintains a database (5) comprising behavioral limit values; - the computing device (4) compares the behavior observed by the road users (2) with the behavioral limit values stored in the database (5) and counts an event (E) on the road section on which the road users (2) are located if the respectively observed behavior of a road user (2) exceeds the associated behavioral limit value; - the computing device (4) calculates a hazard index (Gl) for at least one road section as a function of the number and severity of the events counted on the respective road section within a specified period; and - a navigation unit which takes into account the hazard indices (Gl) assigned to the respective road sections for route planning, whereby, in order to determine the navigation route (1), those road sections are avoided whose hazard index (Gl) exceeds a specified hazard index threshold value.
2. Method according to claim 1, characterized in that the computing device (4) assigns a rating to a respective road user (2) when a behavior limit value is exceeded and the rating of the behavior on the road section during the specified period stopping road users (2) are taken into account to calculate the hazard index (Gl).
3. Method according to claim 1 or 2, characterized in that the significance of the event to be assigned to a road section and / or the assessment to be assigned to a road user (2) depends on the extent to which the associated behavioral limit value is exceeded.
4. Method according to one of claims 1 to 3, characterized in that the behavioral limits are determined in accordance with the applicable road traffic regulations.
5. Method according to one of claims 1 to 4, characterized in that the computing device (4) assigns to a road section an event whose value depends on the current total number of road users (2) on the road section, the average distance between the road users (2), in particular taking into account its standard deviation and / or an event dependent on the number of lane changes on the road section determined during a defined period of time.
6. Method according to one of claims 1 to 5, characterized in that the computing device (4) determines situational characteristics present from the sensor data when a behavioral limit value is exceeded and executes a machine learning model which establishes a relationship between the determined situational characteristics and a respective exceedance of a respective behavioral limit value.
7. Method according to one of claims 1 to 6, characterized in that the computing device (4) calculates the hazard index (Gl) of at least one road section for a future time horizon.
8. Method according to one of claims 1 to 7, characterized in that the computing device (4) determines situational features present from the sensor data when a behavioral limit value is exceeded and adapts the level of at least one behavioral limit value depending on at least one situational feature.
9. Method according to one of claims 1 to 8, characterized in that the navigation unit increases the specified danger index threshold value for at least one road section of a navigation route (1.1) determined disregarding the danger index (Gl) if the path-related and / or time-related length of at least one road section of the navigation route (1.2) calculated taking into account the danger index (Gl) exceeds a specified length threshold value and then recalculates the navigation route (1).
10. Method according to one of claims 1 to 9, characterized in that the vehicle follows the navigation route (1) in an at least partially automated manner.
11. Method according to claim 10, characterized in that the vehicle applies increasingly more defensive driving behavior as the danger index (Gl) increases.
12. System comprising a sensor system (3), at least one vehicle, a computing device (4) and a navigation unit, characterized in that the sensor system (3), the vehicle, the computing device (4) and the navigation unit are set up to carry out a method according to one of claims 1 to 11.