DE102015222964A1 - A driver intention prediction device and method and vehicle comprising such a device - Google Patents

Abstract

A device (10) adapted to predict the driver's intention at an intersection, comprising an intersection situation identification unit (20) arranged to receive intersection situation information, adapted to Identifying an intersection situation based on the received intersection situation information, and being arranged to provide an intersection situation identification output, at least one indicator providing unit (30, 32, 34) arranged to receive at least one sensor signal and configured to provide a driver intent indicator based on the at least one received sensor signal, and a distance-to-intersection determining unit (50) arranged to: determine the distance of the driver's vehicle to the intersection, wherein the driver intention prediction device (10) is adapted to the driver Intenti on at an intersection based on the intersection situation identification, the driver intention indicators and the distance of the driver's vehicle to the intersection to predict.

Description

The present invention relates to a device for predicting the intention of the driver of a vehicle approaching an intersection.

Such a device is for example off JP 2010-097 335 A known. Based on the position of the vehicle detected by a Global Positioning System (GPS) as well as map data provided by a map database stored in the vehicle, the known driver intention prediction device determines whether the vehicle is at an intersection approaching or not, and in the event that the driver fails to actuate the turn signal of the vehicle, predicts the driver's intention based on the driver's line of sight and, if necessary, communicates the prediction result to other vehicles in the vicinity of the vehicle ,

Another device for predicting the intention of the driver of a vehicle approaching an intersection is known from German patent application 10 2015 200 059.1 which was filed in the name of the present applicant and discloses a driver intention estimation device for estimating driver intention at an intersection based on at least one driver behavior indicator. The entire revelation of German patent application 10 2015 200 059.1 is hereby incorporated by reference into the present application.

In view of the above, it is the object of the present invention to provide a further improved driver intention prediction apparatus.

This object is achieved by a device which is adapted to predict the driver's intention at an intersection, comprising an intersection situation identification unit which is adapted to receive intersection situation information, is adapted to Identifying an intersection situation based on the received intersection situation information, and configured to provide an intersection situation identification output, at least one indicator providing unit configured to receive at least one sensor signal , and configured to provide a driver intent indicator based on the at least one received sensor signal, and a distance-to-intersection determining unit configured to increase the distance of the driver's vehicle to the intersection determine, wherein the driver intention prediction device is adapted to the driver-in tion at an intersection based on the intersection situation identification output provided by the intersection situation identification unit and the driver intention indicators provided by the at least one indicator providing unit, as well as the distance of the driver's vehicle to the intersection determined by the distance-to-intersection determination unit.

The distance-to-intersection determination unit may determine the distance to the intersection (to be specific: the distance to the center of the intersection), for example, based on data obtained from a vehicle position detection device and / or or a map database and / or by evaluating images taken by a camera located in the driver's vehicle and / or from the intersection itself if the intersection is provided with intelligent intersection technology.

The present invention was made based on the fact that the inventor (s) became aware that the driver's behavior changes when approaching an intersection depending on the distance to the intersection. This may be illustrated with respect to a situation in which the driver is approaching an ordinary 90 ° intersection on the road with the right of way at the intersection. In this case, the driver has the three options "continue straight", "turn left" and "turn right". Further assume that the driver's vehicle is the only vehicle approaching the intersection and that there are no pedestrians or the like at the intersection.

In this situation, the cumulative, ie average, driver-head yaw angle may be used as a driver behavior indicator. The information about the driver-head yaw angle may be obtained, for example, by evaluating images of the driver taken by an on-board camera of the vehicle using pattern recognition methods. The head yaw angle thus obtained may be cumulated for a predetermined time period or for a predetermined driving distance of the vehicle. An average head yaw angle α, which is -α ₀ ≤ α ≤ + α ₀ , could be used as an indication that the driver tends to continue straight ahead, whereas an average head yaw angle α of α <-α ₀ could be taken as an indication that the driver tends to turn right, and an average head yaw angle α of α > + α ₀ could be taken as an indication that the driver tends to turn left.

On the basis of these considerations, it will be readily understood that the at least one indicator providing unit may be configured to adjust the distance of the driver's vehicle to the intersection upon providing the driver intent indicator based on the at least one received sensor signal. To consider signal. For example, the closer the driver gets to the intersection, he will be more focused on the direction in which he intends to drive, ie, the smaller the value of the limit angle α ₀ that is used to evaluate the driver behavior indicator "Head". Yaw angle "is used.

In addition or as an alternative, however, the evaluation strategy of the driver intention prediction apparatus may also be changed depending on the distance to the intersection. For example, the driver intention prediction apparatus may further provide at least two prediction units, each configured to provide a driver intent prediction output based on at least one driver intent indicator, which is characterized by the at least one indicator Provisioning unit is provided, and comprise a selection unit, which is adapted to select one of the prediction units based on the intersection situation identification output to the driver intention prediction output of the selected prediction unit as provide the driver intention prediction result of the device, wherein the selection unit is adapted to consider the distance of the driver's vehicle to the intersection in selecting one of the prediction units.

In order to facilitate the consideration of the distance to the intersection by the at least one indicator providing unit and / or by the selection unit, the distance to the intersection may be subdivided into a predetermined number of pitch intervals, each having a predetermined size, and the distance-to-intersection determination unit may be configured to determine in which distance interval the driver's vehicle is located. Taking into account the respective size of the distance intervals, the total number of distance intervals may be used, for example, to determine the distance to the intersection at which the beginning of the operation of the driver intention prediction device is triggered.

Further, the total number of the interval intervals and / or the respective size of at least one of the distance intervals may be selected depending on the intersection situation identification output provided by the intersection situation identification unit. For example, the more complicated the intersection situation, the sooner the driver intentions prediction should be started, and the shorter the distance intervals to the intersection should be.

For example, it is obvious that driver behavior will vary depending on whether there are traffic lights at the intersection, if there are traffic signs at the intersection, or if the right of way is decided based on the "right to left" rule. The driver intention prediction apparatus according to the present invention analyzes the intersection situation, identifies the specific intersection situation, and chooses the most appropriate subdivision intervals and, for each distance interval, the most appropriate prediction unit, depending on the result of the identification Predicting the driver's intention in this specific situation. Accordingly, the driver intention prediction apparatus according to the present invention provides a more reliable prediction result.

It should be noted that in the context of the present invention, the term "driver intention" refers, for example, to whether the driver will turn left or right at the intersection or continue straight ahead, and / or the driver accelerate his vehicle or will continue or will continue at the same speed.

The intersection situation information may include information about the stationary situation of the intersection as such, preferably as seen from the direction from which the driver's vehicle approaches the intersection. In the context of the present invention, the term "stationary situation of the intersection as such" refers to characteristics of the intersection which do not change over time, for example how many roads meet at the intersection and / or whether or not separate and / or combined tracks for the possible directions (eg, one track for "left-straight-ahead" or two tracks, a first track for "left-straight" and a second track for "right" or a first track for "left" and a second track for "straight-ahead" or three tracks, a first track for "left", a second track for "straight ahead" and a third track for "right"), and / or whether or not traffic lights are present at the intersection, and / or whether or not separate traffic lights are present for separate lanes, and / or whether or not traffic signs are present at the intersection, and / or which type of vehicle n traffic sign is provided at the intersection, and the like information.

Additionally or as an alternative, the intersection situation information may include information about the transient situation of the intersection as such, preferably as seen from the direction the driver's vehicle approaches the intersection. In the context of the present invention, the term "transient situation of the intersection as such" refers to characteristics of the intersection which may change over time regardless of the traffic situation at the intersection, e.g. Whether or not the traffic lights are in operation and / or what the current state of the traffic lights is (red-yellow-green) and the like information.

In addition or as an alternative, the intersection situation information may include information about the driver's vehicle approaching the intersection. For example, the driver behavior may depend on the type of lane on which the driver's vehicle is approaching the intersection.

Additionally or as an alternative, the intersection situation information may include information about other moving objects. For example, the driver behavior of whether or not other moving objects exist, e.g. As automobiles, motorcycles, bicycles, pedestrians, and the like, and / or on the position and / or direction of movement of the other moving objects, and the like, depend.

It should be noted, however, that at least some of this information, eg. For example, the latter information regarding the presence of other moving objects need not be taken into account in identifying the crossing situation. Instead, this information may be used in predicting the driver's intent, since for any maneuver the driver may intend to perform, other moving objects, e.g. For example, a vehicle that travels in a direction that extends across the direction of travel of the driver's vehicle may have right of way.

To obtain at least some of the above-mentioned information, the driver intention prediction apparatus may include an interface for obtaining intersection situation information from a vehicle position detection apparatus and / or a map database.

In addition or as an alternative, at least some of the above-mentioned information can be obtained by evaluating images taken by a camera arranged in the driver's vehicle. Consequently, the driver-intention prediction apparatus may further include an interface for obtaining intersection-situation information from a pattern recognition apparatus receiving images from a camera.

In addition, or as an alternative, at least some of the above information may be obtained from the intersection itself if the intersection is equipped with an intelligent intersection technology system.

At least one indicator providing unit may calculate the driver intention indicator based on at least one received sensor signal.

According to a first example, an indicator providing unit may receive a sensor signal providing information about the above-mentioned head yaw angle, cumulating that sensor signal for a predetermined time period or for a predetermined driving distance of the vehicle, and the average head yaw angle as one of the driver intention indicators.

• • eine geschätzte Zeit zu der Kreuzung, welche auf der Grundlage des Abstands zu der Kreuzung und der momentanen Geschwindigkeit des Fahrzeugs des Fahrers berechnet werden kann,
• • eine Gangwechselbetätigung, welche auf Grundlage der Variation über die Zeit eines Sensor-Signals bestimmt wird, welches Informationen über den momentanen Gang des Fahrzeugs des Fahrers bereitstellt,
• • eine Fahrzeuggeschwindigkeit und eine Geschwindigkeitsänderung, welche auf Grundlage von Informationen über die momentane Geschwindigkeit des Fahrzeugs und/oder die momentane longitudinale Beschleunigung des Fahrzeugs und/oder die momentane Betätigungsmenge des Gaspedals des Fahrzeugs und/oder den momentanen Bremsdruck und/oder den momentanen Gang und/oder die geschätzte lokale Krümmung der Straße, welche zu der Kreuzung führt, und/oder den momentanen Zustand der Ampeln bestimmt werden,
• • das Fahrzeug-Folgeverhalten, welches auf Grundlage von Informationen über den Abstand zu und/oder die Relativgeschwindigkeit in Bezug auf ein vorausfahrendes Fahrzeug bestimmt wird,
• • die Art einer Fahrbahnmarkierung (z. B. „gestrichelt” oder „durchgezogen”), welche beispielsweise durch eine Muster-Erkennungs-Vorrichtung, welche Bilder von einer Kamera empfängt, bestimmt werden kann,
• • das Vorhandensein einer oder mehrerer benachbarter Spuren, welches auf Grundlage von Informationen über die Position und/oder die Relativgeschwindigkeit und/oder den Bewegungszustand eines oder mehrerer Objekte bestimmt wird, welche von einem Radar erfasst werden.

Other examples of driver intentions indicators may be:

• An estimated time to the intersection, which can be calculated based on the distance to the intersection and the current speed of the driver's vehicle,
• A gear change operation that is determined based on the variation over time of a sensor signal that provides information about the current gear of the driver's vehicle,
• A vehicle speed and a speed change, which are based on information about the current speed of the vehicle and / or the current longitudinal acceleration of the vehicle and / or the current actuation amount of the accelerator pedal of the vehicle and / or the instantaneous brake pressure and / or the current gear and / or the estimated local curvature of the road leading to the intersection and / or the current state of the traffic lights are determined,
• The vehicle following behavior, which is determined based on information about the distance to and / or the relative speed with respect to a preceding vehicle,
• The type of lane marking (eg, "dashed" or "solid") that can be determined, for example, by a pattern recognition device that receives images from a camera,
• The presence of one or more adjacent lanes determined on the basis of information about the position and / or the relative speed and / or the state of motion of one or more objects detected by a radar.

It should be noted that one and the same sensor signal may be evaluated according to various evaluation methods to provide two or more different driver intention indicators. For example, the vehicle speed could be evaluated to provide a current value of the vehicle speed and / or a current value of the acceleration / deceleration of the vehicle.

Additionally or alternatively, the at least one indicator providing unit may provide the sensor signal as received as the driver intention indicator. For example, the current gear, as obtained from a gear change sensor, or the turn signal status, as obtained from a turn signal sensor, may be provided as one of the driver intentional indicators.

The selection unit determines which of the prediction units is most suitable for predicting the driver's intention based on the specific characteristics of the crossing situation identified by the crossing situation identification unit.

In this connection, it should be mentioned that a similarity of a plurality of crossing situations may result in the same prediction unit being selected for a plurality of different crossing situations. For example, the behavior on a straight priority road at a traffic signal intersection is identical to the behavior at a green traffic light intersection: in all situations, the driver's vehicle has the right of way to go straight and turn right, giving right of way if it does turns left. Therefore, the same prediction unit can be selected for all these situations. As a consequence, the plurality of possible crossing situations can be divided into a smaller plurality of crossing situation classes, for each of which a specific prediction unit can be selected.

It should also be noted that for a specific intersection situation, not all driver intentional indicators need to be required to provide a reliable prediction result.

According to a first example, the gear change at an intersection without traffic lights, but with a stop sign or a right-of-way sign, or at an intersection without traffic lights and without traffic signs, d. H. at an intersection, where the "right to left" rule applies, no advantage. In either of these situations, slowing down to near zero speed, or even stopping the vehicle, which is accompanied by a downshift into the first or second gear, will be required due to the priority control.

According to a second example, the speed and speed change at an intersection without traffic lights but with a stop sign or a right-of-way sign provide no benefit. In this situation, slowing down or even stopping the vehicle will be required due to right of way regulation.

According to a third example, the vehicle following behavior provides no benefit at a crossroads without traffic lights but with a stop sign or a right-of-way sign. In this situation, the behavior of the preceding vehicle will be determined by the right of way regulation.

According to a fourth example, the lane marking is at an intersection without traffic lights but with a stop sign or a right-of-way sign, at an intersection without traffic lights and without traffic signs, d. H. at an intersection, where the "right to left" rule applies, no advantage. In these situations, the road on which the vehicle approaches the intersection usually only has one lane per direction of travel. Moreover, the lane mark at an intersection without traffic lights but with traffic signs approaching the driver's vehicle on a priority road turning right makes no advantage because the lane mark detection fails for roads having a large curvature.

According to a fifth example, the presence of an adjacent lane at an intersection without traffic lights, but with a stop sign or a right-of-way sign, at an intersection without traffic lights and without traffic signs, ie at an intersection where the "right to left" Rule applies, no benefit. In these situations, the road on which the vehicle approaches the intersection usually only has one lane per direction of travel. Consequently, there is no adjacent lane. Moreover, the presence of an adjacent lane at an intersection without traffic lights, but with traffic signs approaching the driver's vehicle on a priority road turning right, does not provide an advantage since the detection of an adjacent lane by radar for roads with a high curvature on the ground missing information of the further course of the road is not reliable.

In principle, it would be conceivable nevertheless to provide all available driver intention indicators to all existing prediction units and to rely on the fact that irrelevant driver intention indicators are automatically excluded when training the prediction units. However, in order to reduce the complexity of the prediction units, it is preferable to forward to each of the prediction units only the respectively relevant driver intention indicators.

According to a first alternative, the driver intention prediction apparatus according to the present invention may comprise a second selection unit which is adapted to display the intersection situation identification output from the intersection situation identification unit and the driver's intention prediction apparatus. Receive intentional indicators from the at least two indicator providing units, and which is arranged, based on the intersection situation identification output, to send at least one driver intention indicator to the prediction unit, which is indicated by the first selection indicator. Unit is selected to provide. According to this first alternative, the prediction units have a variable connection with the indicator providing units, and therefore the driver intention prediction apparatus according to the present invention may comprise a reduced number of prediction units, since one and the same prediction Unit can be used for different intersection situations using different sets of driver intention indicators for each of the intersection situations.

However, according to a second alternative, it is also conceivable that each prediction unit is permanently connected to at least one associated indicator providing unit.

According to a specific embodiment, at least one prediction unit may comprise a Bayesian network. Bayesian networks provide a compact representation of the required data and can be trained using machine learning techniques based on data obtained through hands-on experience. In principle, however, any suitable other type of classification device may be used, for example, a fuzzy logic-based classification device, a classification vector based on a support vector machine, an artificial neural-based classification device Network, to name just a few types of suitable classification devices. Which type of classification device is best suited to be used as a prediction unit for a specific class of intersection situations may be decided, for example, with respect to the compactness of the representation of the required data and / or the complexity of their training become.

According to another embodiment, the driver intention prediction apparatus according to the present invention may be configured to output the driver intention prediction result to at least one driving assistance apparatus. A first type of driver assistance device may be an active driver assistance device that actively assists the driver, for example by activating the turn signal if the driver forgets to activate it, and / or by activating the brakes if there is a risk that the driver could disregard a right of way of another vehicle, whereas a second type of driving assistance may be a passive driving assistance device, which passively assists the driver, for example by alerting him to a potentially dangerous situation, for example by visual and / or acoustic signals.

According to another embodiment, the driver intention prediction apparatus according to the present invention may include a triggering apparatus configured to initiate the start of the operation of the driver intention prediction apparatus if a predetermined condition has been satisfied. For example, operation of the driver intention prediction device may be triggered if the distance to the intersection approaching the driver's vehicle is less than a predetermined distance, e.g. B. less than 40 m. However, according to an alternative embodiment, the operation of the driver intention prediction device may be triggered if the time to the intersection approaching the driver's vehicle is less than a predetermined time.

In another aspect, the present invention relates to a method for predicting driver intent at an intersection, comprising the steps of receiving intersection situation information, identifying the intersection situation based on the received intersection situation information, receiving at least one sensor Signal, providing at least one driver intent indicator based on the at least one received sensor signal, providing the distance of the driver's vehicle to the intersection based on the at least one received sensor signal, the driver intention at the intersection Basis of the identified crossing situation, the at least one Driver Intentions indicator and the distance of the driver's vehicle to the intersection is predicted.

Other embodiments of the driver intent prediction method according to the invention may be taken from the above description of the driver intention prediction apparatus according to the present invention.

In the following, two embodiments of the present invention will be explained in more detail with reference to the accompanying drawings.

1 FIG. 12 is a block diagram of a first embodiment of a driver intention predicting apparatus according to the present invention; FIG.

2 FIG. 12 is a block diagram of a second embodiment of a driver intention prediction apparatus according to the present invention; FIG. and

3 to 11 show a plurality of specific crossing situations, wherein

3 showing a two-lane intersection for the driver's vehicle's three possible directions of travel, a first lane to turn left and a second lane to drive straight or right, and with separate traffic lights for the separated lanes, with only the traffic light to the left Turning is shown;

4 shows a three-lane intersection for the driver's vehicle's three possible directions of travel, a first lane to turn left, a second lane to drive straight ahead, and a third lane to turn right, and with separate traffic lights for the separated lanes, only the traffic light is shown turning left;

5 a three-lane intersection showing the driver's vehicle's three possible directions of travel, a first lane to turn left, a second lane to go straight, and a third lane to turn right, and a common traffic light for all three lanes, where the green light is on;

6 shows a junction with a combined lane for the three possible directions of travel of the driver's vehicle, and with a common traffic light for all lanes;

7 an intersection without traffic lights, but with traffic signs, showing a general ascent sign or (indicated by a slash) a right-to-left intersection sign for the direction of travel of the driver's vehicle;

8th an intersection without traffic lights but with traffic signs shows, namely a stop sign or (indicated by a slash) a right-of-way sign for the direction of travel of the driver's vehicle;

9 an intersection shows where the right of way is decided on the basis of the "right to left"rule;

10 showing an intersection with traffic lights, the red light illuminating, and with a green arrow-on-black-ground indication of the direction of travel of the driver's vehicle; and

11 an intersection shows where the priority road makes a right turn; and

12 to 14 show schematic drawings to illustrate the distance-to-intersection concept.

According to 1 includes a driver intention prediction device 10 a crossing situation identification unit 12 ,

On its input side, which is shown on its left side, is the intersection situation identification unit 12 with input devices 14 . 16 and 18 connected. As a non-limiting example, the input device 14 a vehicle position detecting device, for example a GPS device, may be the input device 16 a map database device and may be the input device 18 be a pattern recognition device that evaluates images received from a camera (not shown) disposed in the vehicle V of the driver (see 3 to 11 ).

The intersection situation information may include information about the stationary situation of the intersection as such.

As mentioned above, the term "stationary situation of the intersection as such" may refer to characteristics of the intersection that are not changed over time, for example, how many roads meet at the intersection (see eg. 3 ) and / or whether or not there are separate and / or combined traces for the possible directions, e.g. For example, a track for "left-straight-right" (see, eg. 6 . 8th and 9 ) or two tracks, a first track for "left-straight" and a second track for "right" (see, for example, 10 ) or a first track for "left" and a second track for "straight ahead-right" (see eg. 3 ) or three tracks, namely a first track for "left", a second track for "straight ahead" and a third track for "right" (see eg. 4 and 5 ) and / or whether or not traffic lights are present at the intersection (see eg 3 to 6 and 10 ) and / or whether or not separate traffic lights are available for separate lanes (see eg 3 and 4 ) and / or whether or not there are traffic signs at the intersection (see eg 7 to 9 and 11 ) and / or what type of traffic sign is provided at the intersection, and the like information.

As such, the intersection situation information about the stationary situation of the intersection may be provided by the vehicle position detection device 14 and the map database device 16 to be obtained. However, you can also use the pattern recognition device 18 to be obtained.

Further, the intersection situation information may include information about the transient state of the intersection as such.

As mentioned above, the term "transient situation of the intersection as such" refers to characteristics of the intersection, which may change over time irrespective of the traffic situation at the intersection, for example whether or not the traffic lights are in operation and / or what the traffic light is current status of the traffic lights is (red-yellow-green, see eg 5 and 10 ), and the like information.

As such, the intersection situation information about the transient situation of the intersection may be provided by the pattern recognition device 18 to be provided.

Further, the intersection situation information may include information about the vehicle V of the driver approaching the intersection.

As mentioned above, the driver behavior depends on the type of lane on which the driver's vehicle V approaches the intersection (see, for example, FIG. 3 to 5 and 10 ).

As mentioned above, the information about the vehicle V of the driver approaching the intersection may be obtained by, for example, the pattern recognition device 18 to be provided.

The input devices 14 . 16 and 18 are further with a distance-to-intersection determining unit 50 which determines the distance d of the driver's vehicle V to the intersection (see eg 12 . 13 and 14 ).

The intersection situation identification result and the distance traveled by the distance-to-intersection determination unit 50 is determined are to a first selection unit 20 and to a second selection unit 22 forwarded to the output side of the intersection situation identification unit 12 are connected as shown on the right side thereof.

Based on the intersection situation identification result obtained from the intersection situation identification unit 12 is received, and the distance, which from the distance-to-intersection determining unit 50 is received, determines the first selection unit 20 which of the prediction units 24 . 26 and 28 is best suited to provide a reliable prediction result of the driver intention at the intersection. The prediction units 24 . 26 and 28 may be formed by any known type of classification device, for example, a classification device comprising a Bayesian network, a classification device based on a fuzzy logic, a classification device based on a support vector machine, a Classification device based on an artificial neural network, to name just a few types of suitable classification devices. It should also be noted that the driver intention prediction device 10 according to the present invention is not limited to include three prediction units. Instead, it may comprise less than three or more than three prediction units.

On its input side, which is shown on its left side, is the second selection unit 22 with a plurality of indicator providing units 30 . 32 and 34 which are again shown at their respective input sides, which are shown on their respective left side, with input signal providing units 36 and 38 are connected. Again, note that the driver intention prediction device 10 according to the present invention is not limited to comprising three indicator providing units. Instead, it may comprise fewer than three or more than three indicator providing units. Moreover, it is not limited to encompassing two input signal providing units. Instead, it may include more than two input signal providing units.

The input signal providing unit 36 is with two indicator providing units 30 and 32 to illustrate, by a non-limiting example, that one and the same input signal may be evaluated according to various evaluation methods, by two or more different driver intention indicators provide. For example, the vehicle speed could be evaluated to provide a current value of the vehicle speed and / or a current value of the acceleration / deceleration of the vehicle.

However, it is also conceivable to evaluate a specific input signal only in a specific way to provide a specific driver intention indicator. For example, the indicator providing unit 34 receive an input signal that provides information about the head yaw angle of the driver's head, cumulate that sensor signal for a predetermined time period or for a predetermined driving distance of the vehicle, and provide the average head yaw angle as one of the driver intention indicators. The information about the head yaw angle may be from, for example, a pattern recognition unit 38 which evaluates images of the driver taken by an on-board camera of the driver's vehicle using pattern recognition methods. Similarly, an average driver's line of sight may be provided as one of the driver intentional indicators.

Further, it should be noted that at least one indicator providing unit receives the sensor signal as received from the sensor unit, i. H. without any calculation, than can provide the driver intent indicator. For example, the current gear, as obtained from a gear change sensor, or the turn signal status, as obtained from a turn signal sensor, may be provided as one of the driver intentional indicators.

From the above, it is obvious that an input signal providing unit may be either a sensor unit, for example, a speed change sensor, a turn signal sensor, a vehicle speed sensor, or the like, or an evaluation unit that evaluates the output signals generated by the at least one sensor unit provided, for example, a pattern recognition unit that evaluates images of the driver taken by an on-board camera to provide a head yaw angle.

Another type of input signal may include information about other moving objects O (see, eg, FIG. 9 and 10 ), since the driver's behavior will depend on whether his vehicle V (see eg 9 ) or one of the other moving objects O (see eg 10 ) has the right of way.

Based on the intersection situation identification result obtained from the intersection situation identification unit 12 is received, and the distance from the distance-to-intersection determination unit 50 is received, determines the second selection unit 22 which of the indicators provided by the indicator provisioning units 30 . 32 and 34 to which of the prediction units 24 . 26 and 28 to be forwarded. This means that, depending on the crossing situation identification result, one and the same prediction unit can receive different sets of indicators.

Last is the prediction unit, which is determined by the first selection unit 20 is selected, the driver intention prediction result to at least one driving assistance device. As a non-limiting example, in 1 only a driver assistance device 40 shown. A first type of driver assistance device may be an active driver assistance device that actively assists the driver, for example by activating the turn signal if the driver forgets to activate it, and / or by activating the brakes if there is a risk. that the driver could disregard a right of way of another vehicle, whereas a second type of driving assistance may be a passive driving assistance device, which passively assists the driver, for example by alerting him to a potentially dangerous situation, for example by visual and / or acoustic signals.

Regarding 12 . 13 and 14 It should be noted that the first and the second selection unit 20 and 22 the distance d from the distance-to-intersection determination unit 50 may be evaluated according to a predetermined number of distance intervals, each having a specific size, wherein the predetermined number and / or the specific size of each distance interval may be selected depending on the intersection situation identification result obtained from the intersection -Situations identifier unit 12 Will be received.

For example, shows 12 four pitch intervals I1, I2, I3 and I4, all of the same size, namely 10 meters, whereas according to FIG 13 six pitch intervals I'1, I'2, I'3, I'4, I'5 and I'6 of the same size are provided. Last shows 14 an example with six pitch intervals I''1, I''2, I''3, I''4, I''5 and I''6 of different sizes. It should be noted that, since the focus of 12 to 14 is at the intervals of intervals, the details of the respective intersection have been omitted for the sake of simplicity.

For the sake of completeness, it should be mentioned that the driver intention prediction device 10 according to the present invention, a triggering device 42 which is adapted to start the operation of the intersection situation identification unit 12 and therefore the entire driver intention prediction device 10 trigger if a predetermined condition has been met. For example, the operation may be triggered if the distance to an intersection approaching the driver's vehicle is less than a predetermined distance, e.g. B. less than 40 m. The distance to the intersection may be determined, for example, based on data collected from a vehicle position detecting device and / or a map database and / or by evaluating images taken by a camera disposed in the driver's vehicle , and / or obtained from the intersection itself if the intersection is provided with intelligent intersection technology.

2 shows a second embodiment of a driver intention prediction apparatus according to the present invention, which is different from the driver intention prediction apparatus 10 , what a 1 shown is only slightly different. Therefore, components of the second embodiment will be denoted by components of the first embodiment with the same reference numerals as in FIG 1 designated, but increased by 100 , In addition, the driver intention prediction device becomes 110 out 2 only insofar as they are described by the driver intention prediction device 10 out 1 different. For the remaining description, please note explicitly the description of 1 directed.

The driver intention prediction device 110 , what a 2 is different from the driver intention prediction device 10 , what a 1 is shown only in that it does not have a second selection unit corresponding to the second selection unit 22 out 1 includes. Instead, the indicator deployment units are 130 . 132 and 134 with the respective prediction units 124 . 126 and 128 permanently connected. That means that each of the prediction units 124 . 126 and 128 always receives the same set of indicators provided by the indicator providing units 130 . 132 and 134 based on the input signals provided by the input signal providing units 136 and 138 be received.

With regard to triggering the operation of the intersection situation identification unit 112 through the trigger unit 142 , with reference to the identification of the crossing situation by the crossing situation identification unit 112 based on the input signals provided by the input devices 114 . 116 and 118 with respect to the determination of the distance of the vehicle V of the driver to the intersection by the distance-to-intersection determination unit 150 , with reference to the selection of the most suitable prediction unit by the selection unit 120 and with reference to the output of the prediction result to the driving assistance device 140 Turn to the description of the embodiment 1 directed.

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

• JP 2010-097335A [0002]
• DE 102015200059 [0003, 0003]

Claims (19)

Contraption ( 10 ), which is adapted to predict the driver's intention at an intersection, comprising: • a crossing situation identification unit ( 20 ) which is • arranged to receive intersection-situation information, • is arranged to identify the intersection situation on the basis of the received intersection-situation information, and • is arranged to perform a intersection-situation identification • provide at least one indicator providing unit ( 30 . 32 . 34 ) configured to receive at least one sensor signal, and configured to provide a driver intent indicator based on the at least one received sensor signal, and a distance-to-intersection determination device; Unit ( 50 ), which is adapted to determine the distance of the driver's vehicle to the intersection, wherein the driver intention prediction device ( 10 ) is arranged to control the driver intention at an intersection based on the intersection situation identification output generated by the intersection situation identification unit ( 20 ) and the driver intention indicators provided by the at least one indicator providing unit ( 30 . 32 . 34 ), and the distance of the driver's vehicle to the intersection, which is determined by the distance-to-intersection determination unit ( 50 ) is to predict. The driver intention predicting apparatus according to claim 1, wherein the distance to the intersection is divided into a predetermined number of pitch intervals, each having a predetermined size, and wherein the distance-to-intersection determining unit (10) 50 ) is arranged to determine in which distance interval the driver's vehicle is located. The driver intention prediction apparatus according to claim 2, wherein the total number of the distance intervals and / or the respective size of the at least one of the distance intervals is dependent on the intersection situation identification output generated by the intersection situation identification unit. 20 ) is selected. A driver intention prediction device according to any one of claims 1 to 3, wherein the at least one indicator providing unit (16) 30 . 32 . 34 ) is arranged to account for the distance of the driver's vehicle to the intersection when the driver intent indicator is provided based on the at least one received sensor signal. A driver intention prediction apparatus according to any one of claims 1 to 4, further comprising: • at least two prediction units ( 24 . 26 . 28 each configured to provide a driver intention prediction output based on at least one driver intent indicator provided by the at least one indicator providing unit, and a selection unit 20 ), which is adapted to one of the prediction units ( 24 . 26 . 28 ) on the basis of the intersection-situation identification output to determine the driver-intention prediction output of the selected prediction unit as the driver-intention prediction result of the device ( 10 ), wherein the selection unit is arranged to take into account the distance of the driver's vehicle to the intersection when one of the selection units ( 24 . 26 . 28 ) is selected. A driver intention predicting apparatus according to any one of claims 1 to 5, wherein the intersection situation information includes information about the stationary situation of the intersection as such. A driver intention prediction apparatus according to any one of claims 1 to 6, wherein the intersection situation information includes information about the transient situation of the intersection as such. The driver intention predicting apparatus according to any one of claims 1 to 7, wherein the intersection situation information includes information about the driver's vehicle approaching the intersection. The driver intention predicting apparatus according to any one of claims 1 to 8, wherein the intersection situation information includes information about other moving objects. A driver intention prediction apparatus according to any one of claims 1 to 9, comprising at least one of the following interfaces: An interface for obtaining intersection situation information from a vehicle position detection device (FIG. 14 ) and / or a map database ( 16 ), An interface for obtaining intersection-situation information from a pattern recognition device (FIG. 18 ) which receives images from a camera, • an interface for obtaining intersection-situation information from an intersection equipped with an intelligent intersection technology system. A driver intention prediction apparatus as claimed in any of claims 1 to 10, wherein at least one indicator providing unit ( 30 . 32 . 34 ) calculates the driver intent indicator based on at least one received sensor signal. A driver intention prediction device according to any one of claims 1 to 11, wherein at least one indicator providing unit ( 30 . 32 . 34 ) provides the sensor signal as received as the driver intent indicator. A driver intention prediction device according to any one of claims 1 to 12, wherein it comprises a second selection unit ( 22 ), which is adapted to the intersection situation identification output from the intersection situation identification unit ( 12 ) and the driver intention indicators from the at least two indicator providing units ( 30 . 32 . 34 ) and is adapted to provide at least one driver intent indicator based on the intersection situation identification output to the prediction unit (16). 24 . 26 . 28 ) generated by the first selection unit ( 20 ) is selected. A driver intention prediction apparatus according to any one of claims 1 to 12, wherein each prediction unit ( 124 . 126 . 128 ) with at least one associated indicator providing unit ( 130 . 132 . 134 ) is permanently connected. A driver intention prediction apparatus according to any one of claims 1 to 14, wherein at least one prediction unit ( 24 . 26 . 28 ) comprises a classification device. A driver intention prediction apparatus according to any one of claims 1 to 15, wherein it is arranged to apply the driver intention prediction result to at least one driving assistance apparatus ( 40 ). A driver intention prediction device according to any one of claims 1 to 16, wherein it comprises a triggering device ( 42 ), which is adapted to start the operation of the driver intention prediction device ( 10 ), if a predetermined condition has been met. Vehicle comprising a device according to one of claims 1 to 17. A method of predicting driver intent at an intersection, comprising the steps of: Receiving intersection situation information, Identifying the intersection situation based on the received intersection situation information, Receiving at least one sensor signal, Providing at least one driver intent indicator based on the at least one received sensor signal, Providing the distance of the driver's vehicle to the intersection based on the at least one received sensor signal, wherein the driver's intention at the intersection is predicted based on the identified intersection situation, the at least one driver intent indicator, and the distance of the driver's vehicle to the intersection.

Images