Method for predicting an expected start time of a vehicle

Abstract

Method for predicting an expected start time for a vehicle (1) in a ‘yield’ position comprising the steps: - Determining (100) information about the vehicle's environment (1), - Evaluating (200) the information regarding a future cessation of circumstances which require the vehicle (1) to remain in the ‘yield’ position, and - automatic prediction (400) of the expected arrival time as the time of the future cessation of the circumstances.

Description

[0001] Driver assistance systems are playing an increasingly important role in ensuring driving comfort and preventing traffic accidents.

[0002] A distinction must be made between systems that intervene in the driving dynamics, for example, by automatically initiating emergency braking or an evasive maneuver, and those that are limited to providing the driver with an acoustic, visual, or haptic warning. Due to the high level of safety required and the stringent demands on environmental perception, systems in the former category are only used in selected situations. Warning driver assistance systems, on the other hand, are dependent on the driver's reaction; that is, a warning must be issued at least 1 to 2 seconds before the last possible moment for reaction (e.g., braking).

[0003] In many situations, it is not yet possible to determine the necessity of a warning using conventional means at this time. Fig. Figure 1 illustrates such a situation: The driver of vehicle 1 wants to turn left into the main road. To do this, he first checked to the left and then to the right. At this point, vehicle 3 was not yet approaching from the left, but vehicle 2 was approaching from the right. The driver waits for vehicle 2 to pass from the right and then proceeds without checking to the left again. In the depicted situation, this leads to an accident. To avoid this accident, the driver of vehicle 1 should be warned of vehicle 3 approaching from the left. However, it is unacceptable to warn the driver about every vehicle that has the right of way while waiting. Rather, the warning should only be given when it is foreseeable that the driver is about to start moving. One problem here is that the decision to proceed is only reflected in observable actions (e.g.,This manifests itself (e.g., releasing the brake or pressing the accelerator pedal). Typically, such an action only occurs after the point at which a warning could still prevent the vehicle from moving forward and thus avoid an accident.

[0004] A system that actively prevents starting in a similar situation is known in the prior art. However, this has the disadvantage that the driver no longer has complete control over the vehicle. In situations where the driver can assess the situation better than the vehicle (e.g., because the driver sees that the vehicle is moving forward), the system can prevent the vehicle from moving forward. Fig. (e.g., if vehicle 1 coming from the left turns right), this can negatively affect the acceptance of such a system among users.

[0005] Another relevant state of the art is the so-called start / stop system, designed to save fuel and reduce emissions. In this system, the vehicle's combustion engine restarts when the driver takes an action to resume driving. This action could be, for example, releasing the brake and / or depressing the clutch pedal. Even though modern vehicles restart the engine very quickly after a stop, an unavoidable consequence of the engine stop is a delay before the vehicle actually resumes driving. This reduces the acceptance of start / stop systems among some users. As a result, users deactivate the systems and thus fail to realize their associated benefits.

[0006] DE 102 44 205 A1 deals with warnings and / or braking interventions by vehicles giving right of way when approaching an intersection.

[0007] FR 2 927 720 A1 deals with right-of-way situations in which a warning and / or intervention can take place based on a determined time until a collision.

[0008] US 8 099 232 B2 deals with warnings and / or interventions in uncontrolled intersection situations where right-of-way is assigned situationally.

[0009] JP 2010 - 170 517 A deals with right-of-way situations in which interventions and / or warnings are triggered in the event of imminent collisions. Once the imminent collision situation has ended, the road is cleared.

[0010] It is therefore an object of the present invention to eliminate the aforementioned disadvantages of the prior art.

[0011] The present invention solves the aforementioned problem by means of a method with the features according to claim 1. Additionally, a vehicle, a driver assistance system, and a driving robot are proposed for solving the aforementioned problem. The method according to the invention serves to predict a likely starting time for a vehicle in a "yield" position. The likely starting time can also be understood as a "suitable" starting time and denotes the point in time from which a vehicle could end waiting due to traffic with right-of-way in its vicinity without impairing road safety. The vehicle can, for example, be designed as a road-legal motor vehicle with an internal combustion engine and / or electric drive. According to the invention, information about the vehicle's surroundings is first acquired by sensors.This can be achieved, for example, using sensors already known and intended for other purposes in series production. Such sensors are often based on ultrasound, radar, lidar, and optical sensors. The collected data is then evaluated to determine whether circumstances requiring the vehicle to remain in the "yield" position will cease to exist in the future. This involves, for example, identifying other road users in the vehicle's vicinity and comparing their hierarchy level with the vehicle's own hierarchy level with regard to any existing right-of-way rules. To take the current right-of-way rules into account, traffic signs can be analyzed and / or digital databases (e.g., road maps with right-of-way information) can be consulted.Additionally, approaching vehicles can use car-to-car communication to transmit their presence, direction of travel, and / or information about the current right-of-way rules to the vehicle under consideration. Once the vehicle under consideration has identified no other road users in its vicinity that would prevent it from proceeding, it can continue its journey. By extrapolating the speeds and directions of movement of other road users in the vicinity, the time of departure can also be predicted. Accordingly, the anticipated departure time can be determined as the point at which the circumstances causing the vehicle to stop will cease. A wide range of potential applications for predicting the time of departure are available to those skilled in the art. Individual aspects and applications are explained below without limiting the scope of protection.

[0012] The dependent claims describe preferred embodiments of the invention.

[0013] Preferably, all circumstances that would cause the vehicle to remain in the "yield" position described above are evaluated. This allows the journey to continue without risk of collision. As with the core concept of the present invention, it is important to note that not every detected road user prevents the vehicle from proceeding simply because they would later have to yield the right-of-way. In other words, it is also possible to assess the road users in the vicinity to determine whether the journey can be continued safely even before a potential collision partner approaches.

[0014] The assessment of the traffic situation can involve evaluating a digital database, which is stored either in the vehicle itself or on a stationary server. Additionally, messages exchanged between vehicles can be considered in the evaluation, through which the vehicles inform each other of their presence, direction of travel, and anticipated routes. Both the database and the car-to-car messages exchanged between vehicles can contain information about the current right-of-way rules. All this information improves the basis for predicting when the journey can be resumed.

[0015] One application of the aforementioned method according to the invention involves issuing a warning from a driver assistance system. In this case, a characteristic value for the direction-dependent attention distribution of the vehicle's driver is additionally determined. In other words, the direction-dependent probability distribution is determined for the driver having sufficient information about the traffic situation, i.e., having independently recognized a potential collision partner. If it is determined that the driver has sufficient information, or if it is determined that the driver has (presumably) made a correct assessment of the traffic situation, a warning about the relevant road users or about the entire traffic situation can be omitted.However, if the determined parameter indicates a sufficient probability that the driver will not use the presence of a second road user with the right-of-way as a reason to remain in their current position, a warning of a potential collision with the first road user can be issued once the possibility of a collision with the first road user no longer exists. In other words, according to the invention, it is determined that the driver of the means of transport only considers a subset of the potential collision partners, and a warning is issued to prevent the driver from continuing the journey even though another potential collision partner, whom the driver had not previously considered, has not yet left the danger zone.Here, too, the driver's reaction times can be taken into account, so that a warning of a collision with an unconsidered collision partner is issued within a predefined period before the last potential collision partner considered by the driver is passed. In this way, premature warnings about the unconsidered road user can be avoided, thus increasing user acceptance for a vehicle equipped to execute a method according to the invention.

[0016] The core concept of the present invention can also be used to suppress a warning from a driver assistance system. Here, too, a characteristic value for the direction-dependent attention distribution of the vehicle's driver is determined. In this context, the above explanations regarding the determination of the characteristic value and the characteristic value itself apply accordingly.

[0017] Additionally, the driver assistance system determines the possibility of a collision with a road user who has the right of way. This occurs provided that the driver continues driving before the anticipated time of departure, as determined according to the core of the present invention. According to the invention, the output of the aforementioned warning regarding the determined potential collision is suppressed if the determined parameter indicates a sufficient probability that the driver will stop in response to the road user who has the right of way. In other words, the assistance system does not warn the driver of the collision risk precisely when the evaluation of their behavior suggests that they have already independently recognized and, in particular, correctly assessed the danger.In this way, unnecessary warnings from the driver assistance system can be avoided, which drastically increases the informational value of the warnings that are nevertheless issued to the driver. In particular, this prevents the driver from becoming accustomed to the vehicle's warnings and consequently reducing their own safety precautions. Such a reduction in caution would be critical, as it can increase the risk of an accident in situations where the vehicle on a collision course cannot be detected by the sensors.

[0018] The parameter for direction-dependent attention distribution can be determined by recording and evaluating the driver's gaze direction (distribution) and / or head position (distribution) as a function of time. Corresponding reference values ​​can be stored in a local database to assess currently determined parameters and determine whether the probability of the driver stopping in response to a vehicle with the right-of-way is sufficiently high. The stored reference values ​​for the parameter can be configured according to specific traffic situations. In particular, classes can be defined that include threshold values ​​for acceptable or suitable direction-dependent attention distributions.Possible classes could include, for example, "Turning right at an intersection," "Turning left at an intersection," "Continuing straight ahead at an intersection from a yield position," "Turning right at a T-junction," "Turning left at a T-junction," and so on. Depending on the class, a different appropriate attention distribution is determined for all possible gaze directions and head positions for the driver. By evaluating sensors in the vehicle, suitable parameters can be determined and compared with stored references. Based on this comparison, a decision is then made (yes / no) as to whether there is a sufficient probability that a warning will be displayed by the driver assistance system.

[0019] Another possible application for the method according to the invention is preparing a vehicle in a "yield" position for continued driving. In addition to the steps described above, in accordance with the core concept of the present invention, the approaching time until the determined anticipated start time is compared with a predefined value. Upon reaching the predefined value, an action to prepare the vehicle for continued driving is automatically performed. The predefined value can be determined depending on the action. For example, it can be stipulated that the vehicle's internal combustion engine should be started from a stop state half a second to one second before the anticipated start time in order to provide sufficient power at the anticipated start time.If the inventive method determines at time X that the journey can be resumed two seconds later, the combustion engine can be started after one and a half seconds have elapsed, in order to be able to start moving immediately at time x+2 seconds. This can improve the acceptance of certain functions of the vehicle.

[0020] Alternatively or additionally to automatically starting a combustion engine, the operating state of the vehicle's infotainment system can also be changed in response to reaching a predefined value. For example, the picture of a TV signal can be switched off before the journey resumes. In this way, the user is no longer distracted by the TV signal immediately before resuming the journey and has sufficient time to perceive and assess the current traffic situation. Alternatively or additionally, electrically operated charging systems for a combustion engine can be activated before the journey resumes, allowing their function to assist during acceleration by providing increased engine power.

[0021] According to a second aspect of the present invention, a driver assistance system is proposed which comprises an environmental sensor, an evaluation unit, and, in particular, an interior sensor (depending on the desired functional scope). The environmental sensor is configured to acquire information about the traffic situation in the vehicle's surroundings. The evaluation unit is capable of evaluating the acquired sensor signals and, if necessary, comparing them with reference values. Subsequently, the evaluation unit can generate output values ​​which can be used, for example, to charge a user, to prepare the vehicle for further travel, and / or to suppress the output of a warning signal. In this way, the driver assistance system is configured to execute a method as described above.The features, combinations of features and the advantages arising from them correspond to those described in connection with the method according to the invention, so clearly that reference is made to the above explanations to avoid repetition.

[0022] According to a third aspect of the present invention, a driving robot is proposed which comprises an environmental sensor, an evaluation unit, and an actuator for longitudinal and / or lateral guidance of a vehicle. In other words, the actuator can initiate braking interventions, cause the vehicle to accelerate, and / or initiate steering maneuvers. Such driving robots are sometimes also used for implementing parking assistance and performing autonomous parking maneuvers. The descriptions given in connection with the driver assistance system apply accordingly to the environmental sensor and the evaluation unit. According to the invention, the driving robot is configured to carry out a method as described in detail in connection with the first-mentioned aspect of the invention.The features, combinations of features and the advantages arising from them correspond to those of the method according to the invention in such a way that reference is also made to the above statements with regard to the driving robot.

[0023] According to a fourth aspect of the present invention, a vehicle is proposed which comprises a driver assistance system according to the second aspect of the invention and / or a driving robot according to the third aspect of the invention. The vehicle can, for example, be a road vehicle which, regardless of its engine or intended use, increases road safety by also implementing the features, combinations of features, and advantages of the first three aspects of the invention.

[0024] Further details, features and advantages of the invention will become apparent from the following description and the figures. These show: Fig. Figure 1 depicts a yield situation at a T-junction. The ego vehicle 1 intends to turn left into a road with right-of-way, as indicated by arrow P1. Potential collision partners are a vehicle 2 approaching from the right and a vehicle 3 approaching from the left. According to traffic sign 7 and the green traffic light 5 for the road with right-of-way, the ego vehicle 1 must remain in the depicted position until a safe turn is possible. As indicated by arrows P2 and P3, the intended direction of travel for vehicles 2 and 3 is "straight ahead." In other words, during its intended maneuver, vehicle 1 will cross the trajectories of both vehicles 2 and 3.Based on the highlighted current main viewing angle 4, which serves as an example of the direction-dependent attention distribution of the driver of vehicle 1, it can be assumed that the driver of vehicle 1 has not seen vehicle 3 or is no longer aware of its presence. Although the driver of vehicle 1 waits until vehicle 2 has passed the intersection before proceeding, an immediate continuation of the journey is to be expected, which could lead to a collision 6 with vehicle 3. If vehicle 1 has detected vehicle 3, an immediate warning could be issued by the driver assistance system of vehicle 1 to warn the driver. However, vehicle 2 has not yet passed the intersection, so there is still the possibility that the driver of vehicle 1 will turn their gaze in the direction of vehicle 3 and independently recognize the risk of collision.According to the invention, the warning is therefore initially delayed or suppressed in the depicted situation, or only issued at the latest possible time to avert the collision. If the driver of vehicle 3 subsequently activates the right turn signal, slows down, and the driver of vehicle 1 thus becomes aware of vehicle 3's impending turn, the driver of vehicle 1 is spared an unnecessary warning from their driver assistance system. In this case, vehicle 3's turn can also be used as an opportunity to automatically start the combustion engine of vehicle 1, enabling it to resume its journey as quickly as possible after vehicle 2 has cleared the intersection. Fig. 2 shows components of the in Fig. The vehicle 1 under discussion features a camera 8 as an environmental sensor, which is connected to an electronic control unit 9 as an evaluation unit. A screen 10 of a head unit is located in the dashboard of the vehicle 1, also connected to the electronic control unit 9, and configured to display a visual warning. Additionally, an optical camera 11 is provided as an interior sensor, directed towards the driver, which can detect the driver's head and eye movements and communicate this information to the electronic control unit 9. A data storage device 12 is provided to supply reference values ​​for evaluating driver behavior, as well as a database of the current right-of-way rules, for evaluation by the electronic control unit 9.An actuator 14 of a driving robot is configured to operate the combustion engine 15 and a steering system (not shown) of the vehicle 1, enabling autonomous driving maneuvers. Additionally, the electronic control unit 9 has a connection to an antenna 13 of a wireless communication system, through which it can receive additional information from a server or from other road users in the immediate vicinity. Fig. Figure 3 shows a first flowchart illustrating the steps of an embodiment of the core concept of the present invention. In the first step 100, information about the vehicle's surroundings is acquired. This can be done, for example, using a camera system, a radar system, an ultrasound system, a lidar system, or a laser system. In step 200, the acquired information from the environmental sensor is then evaluated and compared with stored reference values ​​to determine whether circumstances requiring the vehicle to remain in the "yield" position will cease in the future. The stored reference can also include motion models for road users, based on which their future positions can be predicted. In step 300, a digital database is used to determine from which directions in the current vehicle environment road users with the right of way might approach.Subsequently, in step 400, a likely or suitable approach time is automatically predicted as the point in time when the circumstances requiring the vehicle to remain in the "yield" position will cease to exist. Fig. Figure 4 shows a second flowchart illustrating steps of a second embodiment of a method according to the present invention. The first four steps 100 to 400 of the illustrated embodiment correspond to those in Figure 4. Fig. Step 3 was discussed. In step 500, a parameter for the direction-dependent attention distribution of the vehicle's driver is determined. This includes head movements and gaze directions over time, with gaze directions or head positions from further back in time being given less weight than gaze directions from more recent times or current ones. Subsequently, in step 600, the vehicle's driver assistance system issues a warning of possible collisions with a first road user, provided that the determined parameter indicates a sufficient probability that the driver will not stop for a second road user with the right of way once the possibility of a collision with the first road user no longer exists.To assess the probability, a reference stored in a local database can be consulted and compared with the determined characteristic value. Fig. Figure 5 shows a third flowchart illustrating the steps of a third embodiment of a method according to the invention. Steps 100 to 500 correspond to those in connection with Fig. The four steps described above correspond to this. In step 700, the driver assistance system determines a potential collision with a first road user who has the right of way, provided the driver continues driving before the anticipated start time. In other words, the system initially determines the collision relevance of the first road user, regardless of the driver's attention level. In step 800, the system suppresses the warning for the identified potential collision if the calculated parameter indicates a sufficient probability that the driver will take the first road user with the right of way as a reason to remain in a "yield" position. Fig. Figure 6 shows a fourth flowchart illustrating steps of a fourth embodiment of a method according to the invention. Steps 100 to 400 correspond to those in Figure 6. Fig. The three steps shown are used. Additionally, in step 900, the time remaining until the anticipated or suitable departure time is compared with a predefined value. If the time remaining reaches this predefined value (e.g., 0.5 seconds, 1 second, 1.5 seconds, or 2 seconds), an action is automatically performed in step 1000 to prepare the vehicle for departure. This includes switching off the visual output of a TV signal and restarting the combustion engine from a stop state. In this way, the driver and vehicle are prepared in a timely manner for the identified possibility of an imminent departure.

[0025] The method presented within the scope of the present invention can utilize models of a driver's situational awareness to determine, based on the driver's gaze behavior or head position, the probability that the driver has perceived or failed to perceive a specific piece of information. Two publications for corresponding models are cited below. M. Liebner, C. Ruhpanner, F. Klanner, H. Klöden, “Use of generic driver behavior models to predict driver intention”, June 2013, DE 10 2013 212 359 A1 S. Epping, “Predicting driver behavior taking into account situational awareness”, Technical University of Munich, Master's thesis.

[0026] A particularly simple prediction of the expected start time is possible under the assumption that the detected road users are moving at a constant speed. Sensory evaluation of the vehicle's surroundings and corresponding adjustment of the predicted dwell times of road users is, of course, possible at any time. If, according to the invention, it is determined that the expected start time makes sense, the switching off and subsequent restarting of the combustion engine can also be carried out automatically. Reference symbol list: 1 Ego vehicle 2, 3 foreign vehicles 4 perspectives 5 traffic light system 6 Collision point 7 traffic signs 8 Camera 9 Electronic control unit 10 screen 11 Camera 12 Data storage devices 13 Antenna 14 Actuator 15 Internal combustion engine 100 to 1000 process steps P1, P2, P3 trajectories

Claims

[1] Method for predicting the likely start time for a vehicle (1) in a ‘yield’ position comprising the steps: - Determining (100) information about the vehicle's environment (1), - Evaluating (200) the information regarding a future cessation of circumstances which require the vehicle (1) to remain in the ‘yield’ position, and - automatic prediction (400) of the expected arrival time as the time of the future cessation of the circumstances. [2] Method according to claim 1, wherein all circumstances which require the vehicle (1) to remain in the ‘yield’ position are evaluated. [3] Method according to claim 1 or 2 further comprising - Evaluating (300) a digital database (12) comprehensive information on those directions of the vehicle's surroundings (1) from which road users (2, 3) with right of way may approach. [4] Method according to any of the preceding claims, wherein the determination of information about the environment of the vehicle (1) - by means of an environmental sensor (8) and / or - by means of communication signals received, in particular from other road users (2, 3). [5] A method for issuing a warning from a driver assistance system comprising the steps of a method according to any one of the preceding claims and further comprising - Determining (500) a characteristic value for a direction-dependent attention distribution of a driver of the vehicle (1), and - Issuing (600) a warning of possible collisions with a first road user by the driver assistance system, provided that - the determined parameter indicates a sufficient probability that the driver will not use a second detected road user with right of way as a reason to stop after the possibility of a collision with the first road user no longer exists. [6] Method for suppressing a warning from a driver assistance system comprising the steps of a method according to any one of claims 1 to 4 and further comprising - Determining (500) a characteristic value for a direction-dependent attention distribution of a driver of the vehicle, and - Determining (700) a possible collision with a first road user having the right of way by the driver assistance system, provided that the driver continues the journey before the expected time of departure, and - Suppressing (800) an output of the warning of the identified possible collision, provided that the identified parameter indicates a sufficient probability that the driver will take the first road user with right of way as a reason to stop. [7] Method according to one of claims 5 or 6, wherein the characteristic value represents a direction of gaze and / or head position of the driver as a function of time. [8] Method for preparing a vehicle in a ‘yield’ position for a journey onward comprising the steps of a method according to any one of claims 1 to 4, and further comprising the steps - Comparing (900) a time interval approaching the expected (or “suitable”) departure time with a predefined value, and responding to the approaching time interval reaching the predefined value - Automatic execution (1000) of an action to prepare the vehicle for further travel. [9] Method according to claim 8, wherein the action - starting an internal combustion engine (15), and / or - a change in the operating state of an infotainment system (9, 10) of the vehicle, and / or - includes switching off an internal combustion engine (15). [10] Driver assistance system comprehensive - an environmental sensor (8), - an evaluation unit (9), and in particular - an interior sensor (11) wherein the driver assistance system is configured to perform a method according to one of the preceding claims. [11] Driving robots encompassing - an environmental sensor (8), - an evaluation unit (9), and - an actuator (14) for longitudinal and / or lateral guidance of a vehicle (1), wherein the driving robot is configured to perform a method according to one of claims 1 to 4, 7 or 8. [12] Vehicle comprising a driver assistance system according to claim 10 and / or a driving robot according to claim 11.

Images