WO2008015966A1

WO2008015966A1 - Collision information indicator and its method

Info

Publication number: WO2008015966A1
Application number: PCT/JP2007/064738
Authority: WO
Inventors: Mototaka Yoshioka; Jun Ozawa
Original assignee: Panasonic Corporation
Priority date: 2006-08-04
Filing date: 2007-07-27
Publication date: 2008-02-07

Abstract

A collision information indicator equipped in a car for indicating a risk of collision with another car or a pedestrian when the another car or the pedestrian may has a misunderstanding that the car having this indicator acts in a way different from that in which the driver of the car intends to act. An intention information detecting section (104) detects intention information on the future action of a mobile object (100). A multiple action judging section (106) judges if the driver may perform multiple actions on the basis of the detected intention information. A collision risk judging section (107) predicts an area where there is a risk of collision if the multiple action judging section (106) judges that the driver may perform multiple actions, and judges if there is a risk of collision with another moving object. A collision information indicating section (108) indicates information on the risk of collision with a person or a moving object with reference to the intention information if the collision risk judging section (107) judges that there is a risk of collision.

Description

Specification

Collision information notification device and method

Technical field

[0001] The present invention relates to a collision information notification device that notifies information related to the risk of collision, and is particularly applicable to a mobile terminal device such as a car navigation system (hereinafter referred to as a car navigation system).

Background art

Conventionally, a destination is detected based on information such as a current position of a moving body and a blinker by inter-vehicle communication that communicates between vehicles using a communication unit, and a dangerous area is determined based on the detected destination. There are devices that identify and determine the risk of collision (see Patent Document 1 and Patent Document 2). Fig. 1 shows an example of the configuration of the system disclosed in Patent Document 1.

In FIG. 1, an information acquisition unit 11 is a means for acquiring the position of a vehicle and the like. Based on the position acquired by the information acquisition unit 11 and the like, the situation identification unit 12 is a means for identifying a dangerous area surrounded by a predetermined area centered on the position of the vehicle, for example. The danger judging unit 13 is a means for judging the possibility of the intersection of the identified dangerous areas. The vehicle control unit 14 is a means for controlling the vehicle on the assumption that there is a risk of collision when it is determined that the vehicles are mixed. FIG. 2 is a diagram illustrating an example of a dangerous area when a conventional vehicle control device performs a risk determination or the like. For example, in Fig. 2, vehicle 1 is located at the location shown in the figure and is going straight ahead. On the other hand, vehicle 2 is about to turn left from the right side of the intersection. A dangerous area is identified for each vehicle, and the possibility of crossing is determined. If it is determined that they will cross each other, control is performed assuming that the vehicle has a risk of collision.

Patent Document 1: JP 2005-56372 A

Patent Document 2: JP 2002-260192 A

Disclosure of the invention

Problems to be solved by the invention

[0004] The invention disclosed in Patent Document 1 is calculated around the position of the vehicle as described above. The possibility of collision is judged based on the specified dangerous area. Further, the invention disclosed in Patent Document 2 calculates dangerous areas in all the routes that the vehicle can take, and determines the possibility of collision based on the dangerous areas. However, the danger of a collision cannot always be avoided by the crossing of dangerous areas identified above! In particular, as a scene where a collision occurs, there is a force S that occurs when there is a difference between the action that one side intends and the action that the other side recognizes. Furthermore, the scene where a collision occurs due to such a difference is not limited to a predetermined location. Hereinafter, a specific example will be described.

[0005] FIG. 3 is a diagram illustrating an example of a scene that creates a risk of collision. In Fig. 3, vehicle 1 blinks the left turn blinker and displays the intention to turn left. In the above prior art, the danger area is identified in the left turn direction based on the blinker information and the like, and the possibility of crossing is judged. For example, a dangerous area is identified along route A. On the other hand, vehicle 2 is going to turn right on route B. Therefore, in the above prior art, a dangerous area is specified along the route B, and it is determined that the dangerous areas of both vehicles do not cross each other, that is, do not collide. However, the vehicle 1 does not necessarily turn left on the route A and may turn left on the route B. For example, when making a right or left turn, it is common to leave the blinker from about 30m before, but if the distance between intersections is short, for example, it is not possible to know which route to turn right or left even if the blinker is taken. There are many things that are not. On the other hand, the user of vehicle 2 thinks that vehicle 1 turns left on route A when viewed from vehicle 2, which is an oncoming vehicle, and tries to turn right as it is. Here, if vehicle 1 actually turns to the left on route B, the two will cross each other. In other words, a collision accident is not judged by the possibility of crossing the surrounding area from the positional relationship of the vehicle, but the behavior that one side intends and the behavior that the other recognizes. There is often a danger if there is a difference between! Moreover, such a collision is not limited to a vehicle, but is the same for a pedestrian. For example, pedestrian 3 in Fig. 3 thinks that vehicle 1 will turn left onto route A, and when trying to cross the road, there is a risk of actually colliding with vehicle 1 turning left on route B. is there.

[0006] Therefore, in the present invention, there is a risk of collision with another vehicle or pedestrian when there is a possibility that the other vehicle or pedestrian will misunderstand that the behavior intended by the own vehicle will be different. It is an object of the present invention to provide a collision information notification device that can notify that there is a problem. Means for solving the problem

[0007] In order to solve the above-described problem, a collision information notification device according to the present invention includes a current position acquisition unit that acquires a current position on a map of a moving object, and an intention display that indicates a future action of the moving object. Intention information detection means for detecting intention information issued for the purpose, the detected intention information, the current position when the intention information is detected, and the intention information within a predetermined range from the current position The movement history accumulating means for accumulating the movement history of the moving object including the selected action, and the intention information, when the intention information is detected, the same intention information within a predetermined range from the current position at that time Duplicate action determination means for determining whether or not a different action has been selected for the same intention information as the ability to select a different action for the same intention information. The intention determination means If different behavior in response to distribution has been determined to be selected, characterized in that it comprises a collision information notifying means for notifying the information about the risk of collision with a person or other moving object.

[0008] It should be noted that the present invention can be realized as a method in which the processing means constituting the device can be realized as a step, or can be realized as a program for causing a computer to execute the step, or the program can be recorded. It can be realized as a computer-readable recording medium such as a CD ROM, or as information, data or a signal indicating the program. These programs, information, data and signals may be distributed via a communication network such as the Internet.

The invention's effect

[0009] The collision information notification device according to the present invention can avoid the risk of a collision by notifying information on the risk of a collision.

Brief Description of Drawings

FIG. 1 shows an example of the configuration of a system disclosed in Patent Document 1.

[FIG. 2] FIG. 2 is a diagram showing an example of a danger area when a conventional vehicle control device performs danger judgment or the like.

[FIG. 3] FIG. 3 is a diagram showing an example of a scene in which a risk of collision occurs. FIG. 4 is a block diagram showing a configuration of a collision information notification system of the present embodiment. 5] FIG. 5 is a diagram showing an example of position information detected at a predetermined interval.

6] Fig. 6 shows an example of a road network represented by the map information stored in the map information storage unit.

7] Figure 7 shows an example of map information (road network information) stored in the map information storage unit.

8] Fig. 8 is a diagram showing an example in which position information detected based on map information is converted into a node series.

9] Fig. 9 is a diagram showing an example of a mobile object's intention display and future behavior.

10] FIG. 10 is a diagram showing an example of action rules stored in the action rule storage unit.

FIG. 11 is a diagram showing an example of the current position of the user and the user's intention information detected at that position.

12] Fig. 12 is a diagram illustrating an area where there is a risk of collision with a third party for the user's future behavior.

[FIG. 13] FIG. 13 shows an example of a display in a collision information display section which is a display screen for car navigation or the like.

14] FIG. 14 is a flowchart showing an example of operations until the user's future behavior is estimated and the collision information is displayed in the collision information notification system of the present embodiment.

[FIG. 15] FIG. 15 is a flowchart showing an example of a specific operation in estimation of future behavior.

16] Fig. 16 is a diagram showing an example of user's intention information and future actions that can be taken in response to intersections with complex branch paths.

17] FIG. 17 shows an example of future actions that can be taken when vehicle 1 makes a right turn as intention information at the same intersection as FIG.

18] FIG. 18 is a diagram showing an example of point information showing information on a plurality of actions corresponding to intention information for each point.

19] FIG. 19 is a diagram showing an example of a display for notifying the vehicle when there is a risk of collision due to overlapping actions. FIG. 20 is a block diagram showing an example of the configuration of a collision information notification device for notifying only other vehicles of the risk of collision, but not only the own vehicle.

FIG. 21 is a block diagram showing an example of a hardware configuration that implements the functional configuration shown in FIG.

22] FIG. 22 is a diagram for explaining an example of the estimated behavior estimated for the intention information at the time of stopping.

En 23] FIG. 23 is a diagram showing an example of the estimated behavior estimated for the intention information at the time of stopping shown in FIG. 22 by a node and a link.

[24] Fig. 24 shows the details of the node information accumulated for the area including the current position of the moving object.

25] FIG. 25 is a diagram for explaining a search for a place where the vehicle can be stopped, as in FIG.

[FIG. 26] FIG. 26 is a diagram showing an example of a route to a stop possible point estimated by the behavior estimation unit as a future behavior.

27] Fig. 27 is a diagram showing an example in which overlapping actions are estimated for the same intention information.

28] FIG. 28 is a diagram showing another example in which overlapping actions are estimated for the same intention information.

[FIG. 29] FIG. 29 is a diagram for explaining calculation of a stop point and a left turn point.

30] FIG. 30 is a diagram showing information relating to bus stops accumulated as one of the points where parking is possible.

FIG. 31 is a block diagram showing an example of a system configuration of the present invention in PtoP inter-vehicle communication in which position information and the like are communicated with each other.

Fig. 32 is a system configuration diagram in the case where a collision information conversion means is provided in order to make the collision area usable by a third party mobile body.

[FIG. 33] FIG. 33 is a diagram showing collision information obtained by converting a collision area from latitude information into latitude and longitude information that can be commonly recognized by any mobile body.

34] FIG. 34 is a diagram showing another example of an area where there is a risk of collision with a third party for the user's future behavior. 35A] FIG. 35A is a diagram showing an example of information when collision information is transmitted to the pedestrian 3.

FIG. 35B is a diagram showing an example of information when collision information is transmitted to the pedestrian 4.

[FIG. 36] FIG. 36 is a diagram showing an example of informing the user of the risk of a collision by estimating the action in the own vehicle based on the current position sent from the oncoming vehicle and the intention information of the left turn. It is.

FIG. 37 is a diagram for explaining an example of behavior estimation in a parking lot.

FIG. 38 is a block diagram showing a configuration of a collision information notification system according to the second embodiment.

FIG. 39 is a diagram showing the movement path of the user shown in FIG. 5 by the link and node series shown in FIG.

En 40] FIG. 40 is a diagram showing the movement history stored in the movement history storage unit, which is represented by an ID sequence of nodes and links.

[FIG. 41] FIG. 41 is a diagram showing a situation in which the user is going straight on Hana 1 Kyoto Street and now has a left turn blinker.

[FIG. 42A] FIG. 42A is a diagram showing the current travel, which is the travel route of the user to the present time, as a sequence of link IDs and node IDs.

[FIG. 42B] FIG. 42B is a diagram showing an example of extracting from the travel history the history of passing the route consistent with the current travel and turning left at Hanamachi 3 intersection.

[FIG. 42C] FIG. 42C is a diagram showing an example of extracting from the travel history the history of passing the route that coincides with the current travel and turning left at the back crossing 1 intersection.

42D] Figure 42D shows the calculation of the possibility of each future action when there are multiple actions that can be taken as future actions.

FIG. 43 is a diagram showing an example of a display screen that displays a notification of collision information accompanied by the movement probability of an oncoming vehicle.

FIG. 44 is a flowchart showing the operation of the collision information notification device in the collision information notification system of the second embodiment.

[FIG. 45] FIG. 45 shows the collision information in the destination prediction process shown in step S310 of FIG. It is a flowchart which shows detailed operation | movement of a knowledge apparatus.

46] FIG. 46 is a diagram showing an example of the behavior of the moving object when the history is accumulated only when the duplicate action occurs.

47] FIG. 47 is a block diagram showing a configuration example of a collision information notification system that implements a modification of the second embodiment.

48] FIG. 48 is a diagram for explaining an example of estimating future actions from intention information based on the accumulated movement history.

FIG. 49A is a diagram showing an example of current traveling.

49B] FIG. 49B is a diagram showing intention information generated by the moving object.

49C] FIG. 49C is a diagram showing the future behavior estimated by the behavior estimation unit.

49D] FIG. 49D is a diagram showing an example of the current running when the behavior estimation unit takes a behavior different from the behavior estimated by the behavior estimation unit.

FIG. 49E is a diagram showing an example in which the current travel shown in FIG. 49D is accumulated as a movement history.

FIG. 50 is a block diagram showing an example of the configuration of a collision information notification system that predicts the destination of another vehicle and determines the danger.

51] FIG. 51 is a diagram for explaining notification control using the destination of another moving object.

[FIG. 52] FIG. 52 is a block diagram showing an example of the configuration of a server when point information is automatically generated using a movement history and collision is avoided using the automatically generated point information. 53] FIG. 53 is a diagram showing an example of a case where a server installed at a specific point generates point information from intention information and movement history of a moving body moving at the point.

54] FIG. 54 is a diagram showing another example of the case where the server installed at a specific point generates point information from the intention information and the movement history of the moving body moving at the point.

55] FIG. 55 is a view showing the appearance of a car navigation device equipped with the collision information notification device of the present invention.

Explanation of symbols

80 cars

81 Car navigation system 100, 200 mobile

101 Location information detector

102 Node series converter

103 Map information storage

104 Intention information detector

105 Behavior estimation unit

106 Duplicate action determination unit

107 Collision risk judgment unit

108 Collision information notification section

109 Collision information receiver

110 Collision information display

111 Action Rules Accumulation Department

112 Notification partner identification part

113 Location information transceiver

114 Second location information transmitter / receiver

115 Second position information detector

116 Collision information converter

117 Movement history storage

118 Destination prediction unit

119 Second destination prediction unit

120 Second destination transmitter

121 Second movement history storage

122 Duplicate action extractor

123 point information generator

124 Collision risk calculator

131 CPU

132 Primary memory

133 In-vehicle information detector 134 External memory

300 servers

BEST MODE FOR CARRYING OUT THE INVENTION

Hereinafter, a collision information notification device according to the present invention will be described with reference to the drawings.

[0013] (Embodiment 1)

FIG. 4 is a block diagram showing the configuration of the collision information notification system of the present embodiment. Hereinafter, each component will be described first, and then the operation flow of the present invention will be described.

[0014] As shown in the figure, the moving body 100 of the present embodiment estimates behavior other than the intention of the own vehicle from the viewpoint of other moving bodies, with respect to the intention information related to the movement of the moving body 100. Mobile device equipped with a collision information notification device for notifying the danger of collision in the event that the position information detection unit 101, node sequence conversion unit 102, map information storage unit 103, intention information detection unit 104, behavior estimation Unit 105, overlapping behavior determination unit 106, collision risk determination unit 107, collision information notification unit 108, and behavior rule accumulation unit 111. The moving body 200 includes a collision information receiving unit 109 and a collision information display unit 110.

[0015] Further, in the mobile object 100, the intention information indicates an intention of a future action of right turn, left turn, and stop, and the action rule accumulating unit 111 indicates that the action indicated by the intention information and the action are performed. This is an example of an action rule accumulating means for accumulating action rules indicating the range on the movement route that can be broken, and the action estimation unit 105 and the duplicate action determination unit 106 refer to the accumulated action rules as follows. This is an example of the “duplicate action determination unit” that determines whether or not a plurality of actions corresponding to the detected intention information can be performed within the range on the movement route.

Furthermore, the position information detection unit 101 is an example of “a current position acquisition unit that acquires a current position on the map of the mobile object”, and the map information storage unit 103 is “map information for storing map information”. The behavior estimation unit 105 and the duplicate behavior determination unit 106 refer to the map information and the behavior rules, and start from the current position acquired when the intention information is detected. As an example of the “duplicate action determination unit” that determines whether or not a plurality of actions indicated by the action rule can be performed within the range indicated by the action rule.

[0017] Further, the behavior estimation unit 105 and the duplicate behavior determination unit 106 indicate that "the intention information has been detected. There are multiple right turn points, left turn points, and stop points indicated by the intention information within the range on the movement route indicated in the action rule, starting from the current position acquired from time to time. The overlapping action that determines whether the moving body can take a plurality of actions with respect to the intention information when there are a plurality of any of the points in the range on the movement route. It is an example of “determination means”.

[0018] In the moving body 100, the map information storage unit 103 stores "map information and point information indicating a moving route that can be selected as an action corresponding to the intention information for each point on the map. The position information detection unit 101 is an example of “current position acquisition unit for acquiring the current position on the map of the moving object”, and includes an action estimation unit 105 and an overlapping action determination unit 106. Refers to the point information accumulated for the point immediately in front of the current position acquired when the intention information is detected, and there are a plurality of movement routes that can be selected for the detected intention information. The collision information notification unit 108 is an example of the “duplicate action determination means for determining whether or not there is”, and the collision information notification unit 108 determines that “when the duplicate action determination unit determines that there are a plurality of movement paths, Information about the risk of collision It is an example of the said collision information notification means to notify.

The position information detection unit 101 in the moving body 100 is a means for detecting the current position of the moving body 100. For example, in the case of a car navigation system, a GPS for detecting the current position of the user is provided, and latitude and longitude information is detected at a predetermined interval such as an interval of about 1 second. In this embodiment, it is assumed that the position information detection unit 101 is configured by GPS or the like, and detects latitude and longitude information as position information along with the movement of the user at a predetermined interval. FIG. 5 is a diagram showing an example of position information detected at a predetermined interval. In Figure 5, the user turns right at Hanamachi 1 intersection and goes straight on Hana 1 Kyoto Street. White circles indicate the detected position information, and are detected at predetermined intervals as the user moves.

[0020] The map information storage unit 103 is means for storing map information. In addition to information about facilities and landmarks, the map information stored in car navigation systems has road network information that includes, for example, intersections and facilities as one node, and links are structured between the nodes. It is common. FIG. 6 is a diagram illustrating an example of a road network that is map information stored in the map information storage unit 103. Figure 6 shows the first intersection of Hanamachi The corresponding “N (node) 100” and “N101” which is the second intersection of Hanamachi are shown. In addition, Hana 1 Kyoto Street, which is the route connecting Hanamachi 1 intersection and Hanamachi 2 intersection, is “L (link) 204”, “N10 0” is “L204”, “L201”, “L202”, “L203” Connection information of nodes and links. Furthermore, it also has link length information such as the distance between “Hanamachi 1 intersection” and “Hanamachi 2 intersection”, that is, the distance of “L 204” is 400 meters. FIG. 7 is a diagram showing an example of road network information stored in the map information storage unit 103. A map is generally divided into predetermined areas and stores information about the nodes that exist in each area, such as facilities and intersections, and the links that connect them. In FIG. 7, the area ID “E01” is the area shown in FIG. As node information existing in the area ID “E01”, “N100”, “N101”, “N102” are stored. In addition, information on each node and link information connected to that node is stored. "N100" is the location "Statistics 135 degrees 34 minutes, North latitude 34 degrees 33 minutes", the name "Hanamachi 1 intersection", and the links to connect, such as "L201", "L201" length 300 meters, etc. Is attached.

The node series conversion unit 102 is a means for converting latitude and longitude, which is position information detected as a user moves based on map information, into a node or link ID series. In general, in car navigation systems, map matching is performed on the route based on the latitude and longitude detected by GPS, that is, to the nodes and links indicated by the road network described above. This is because the detected latitude and longitude are matched to the road network, and route search and guidance are performed using this road network. Similarly, the node series conversion unit 102 in the present embodiment converts the detected position information into an ID series such as a node based on the map information.

FIG. 8 is a diagram illustrating an example in which position information detected based on map information is converted into a node series. Like FIG. 5, in FIG. 8, the user turns right at Hanamachi 1 intersection and goes straight on Hana 1 Kyoto Street. Along with this movement, latitude and longitude indicated by white circles are detected as position information. The node series conversion unit 102 converts the latitude / longitude series into an ID series such as a node with reference to the map information. For example, in the case of FIG. 8, the “L203 → N100 → L204” force S movement sequence is obtained. In this embodiment, the user's moving direction is automatically calculated by the conversion to the ID series. For example, in Fig. 8, The users who are passing through the N100. Therefore, L204 is traveling in the N100 direction force, the next N101 direction, that is, from left to right (eastward), and the moving direction can be known from the ID series of nodes and links in this way. Note that the moving direction may be calculated from the detected latitude and longitude instead of necessarily calculated from these ID sequences, or may be calculated using a gyro. In addition, the method of converting latitude and longitude into a road network is known in conventional map matching techniques such as vertical projection, and is not a problem here.

[0023] The intention information detection unit 104 is means for detecting intention information indicating a user's intention to act against others. For example, in the case of a vehicle, winker information corresponds to intention information. For example, a right turn signal means a right turn, a left turn signal means a left turn or a stop on the shoulder. Moreover, not giving a winker at the intersection also means a straight forward intention display. Furthermore, although it does not exist as a road rule, it may be used to flash the headlamps and give way, or to turn on the hazard lamps and call attention to the stop or the rear vehicle. In addition, not only intention indications that exist as road rules but also intention information that is generally well known to the driver.

[0024] The behavior estimation unit 105 detects the location information detected by the location information detection unit 101, that is, the current travel route of the user converted by the node sequence conversion unit 102 and the intention information detection unit 104 in this embodiment. It is a means to estimate future behavior based on the information about the user's intention and the map information stored in the map information storage means. Hereinafter, a specific example will be described.

[0025] FIG. 9 is a diagram showing an example of intention display and future behavior of a mobile object. In Fig. 9, the user goes straight on Hana 1 Kyoto Street (L205) toward the back Hana 1 intersection. And later, let's assume that the left wing force is one. The intention information detection unit 104 detects the blinker information and detects the intention to turn left. On the other hand, the behavior estimation unit 105 estimates the future behavior from the map information where the left is going to turn. In the case of this example, it is thought that there is an Urahana 1 intersection (N102) 10 meters ahead, and turn left to Urahua Street (L210). On the other hand, there is the Hanamachi 3 intersection (N103) 20 meters ahead straight ahead on L206, and it is thought that turning left at Hanamachi 3 intersection (N103) toward Dahua Street (L209). general The driver is expected to give a blinker from about 30 meters before the intersection where the vehicle turns right or left. On the other hand, the driver may issue a blinker just before the point where the vehicle turns or stops. Therefore, for example, when intersections are close to each other, a plurality of actions may be considered without knowing which intersection is the curvature. For example, in the case of this example, according to the principle of 30 meters before, it would turn left at the 3rd intersection of Hanamachi. It would be possible to turn left at one intersection. In such a case, there is a high possibility that a third party who thinks to take one action will collide with a user who actually takes the other action, causing an accident. Therefore, the behavior estimation unit 105 estimates such a plurality of future behaviors based on the map information from the user's travel! / Route and intention information.

For example, in the present embodiment, the behavior is estimated with reference to the behavior rules accumulated in the behavior rule accumulation unit 111. FIG. 10 shows an example of behavior rules stored in the behavior rule storage unit 111. In the action rules, it is assumed that rules regarding intention information, actions to be performed in response to the intention information, and the scope of the actions to be performed are accumulated. For example, the action rule “001” stores the intention information “left winker”, the action “left turn”, and the range “within 50 m (metre) in the moving direction”. This means that if you take the left turn signal, you will make a left turn at a point within 50 meters in the direction of movement. Note that setting 50 meters instead of 30 meters may be given in advance depending on the user, and the allowable range is taken into consideration in order to ensure higher safety from the viewpoint of fail-safety.

[0027] First, the behavior estimation unit 105 refers to a behavior rule based on the current position detected by the location information detection unit 101 and the intention information detected by the intention information detection unit 104, and responds to the intention. Search for points where action is possible. FIG. 11 is a diagram showing an example of the current position of the user and the intention information of the user detected at that position. In FIG. 11, it is detected that the user has issued a left blinker, which is intention information, at a point of “136 degrees 0 1 minute east longitude, 34 degrees 55 minutes north latitude”. It should be noted that the user's direction and direction of movement are calculated from the node ID system 歹 IJ, gyroscope and position information differences, etc. as described above to indicate that they are heading east on Hua 1 Kyo Street (L205). Yes. Referring to the rules of action, the left turn signal is marked as “turn left” within the range “50 meters in the direction of movement”. So within 50 meters from Hana 1kyo Street, turn left Search for possible locations. As shown in Fig. 7, latitude and longitude information of each node such as an intersection is accumulated in the map information. Based on this latitude and longitude information, a search is made for a point where a left turn is possible within 50 meters of the moving direction. Become.

[0028] First, Hana 1 Kyoto Street (L205) is located on the east from the Urahana 1 intersection (N102), and further to the east, the Hanamachi 3 intersection (N103). If you refer, the location of Urahana 1 intersection (N102) is “136 degrees 02 minutes east longitude, 34 degrees 55 minutes north latitude”, and the location of Hanamachi 3 intersection (N103) is “136 degrees 04 minutes east longitude 34 degrees 55 minutes north latitude”. It has become. The difference in distance between the current position of 136 degrees 01 minutes east longitude 34 degrees 55 minutes north latitude and the intersection of Urahana 1 intersection 136 degrees 02 minutes east longitude 34 degrees 55 minutes north latitude is 10 meters and The difference in distance to “East 136 degrees 04 minutes, North latitude 34 degrees 55 minutes” is 30 meters, and therefore, they are included in the range of action rules, “within 50 meters in the direction of movement”, so these can act. It is searched as a special point. Furthermore, the left turn for the eastward movement of Hua 1 Kyo Street (L205), which the user is currently traveling, means Ura Hua Street (L210) and Hua 3 Intersection (N103 ) Is Dahua Street (L209), and these are calculated as future actions. Therefore, the action estimation unit turns the left side of Urahana 1 intersection into the future action “001” and proceeds along Urahana Street, that is, “N102 → (from) L210”, and the future action “002”, the Hana 3 intersection. Turn left on the road and proceed through Dahua Street, that is, “N102 → L206 → N103 → L209” is estimated as the future action.

[0029] The duplicate action determination unit 106 is a means for determining whether or not future actions estimated by the action estimation unit 105 are duplicated! /. Duplicate behavior refers to multiple actions that can be performed on the same intention information indicated by the user. In other words, when there is only one action estimated from the intention information indicated by the user, the third party is able to take an avoidance action against that action S. As described above, a collision accident is generally indicated by the intention information indicated by the user. This often occurs when there are multiple actions estimated from the above. Therefore, in order to prevent such a collision, the duplicate action determination unit 106 determines whether or not future actions estimated by the action estimation unit 105 are duplicated! /. For example, when a plurality of actions are estimated by the action estimating unit 105, it is determined that the actions overlap.

[0030] The collision risk determination unit 107 determines that there is an overlapping action in the overlapping action determination unit 106. In this case, it is a means for determining the risk of collision with a third party. For example, it is assumed here that an area where there is a risk of collision is determined. This will be described below with reference to the drawings.

FIG. 12 is a diagram for explaining an area where there is a risk of collision with a third party with respect to the future behavior of the user. As in Fig. 9, the user (in this case, vehicle 1) shows his intention to turn left. On the other hand, as described above, vehicle 1 is estimated to have two actions: a left turn at the back 1 intersection and a left turn at the Hua 3 intersection! In such a case, there is a possibility of collision with a third party. For example, in FIG. 12, the user of oncoming vehicle 2 thinks that vehicle 1 turns left at Urahana 1 intersection, and he may turn right at Hanamachi 3 intersection and proceed to Dahua Street. On the other hand, if the vehicle 1 actually turns left at the Hana 3 intersection instead of the Hana 1 intersection, the two may collide at the Hanamachi 3 intersection (N103). In addition, for vehicle 2, vehicle 1 assumes that it will turn left at the back 1 intersection, so it will be delayed to notice and the risk of a collision may be further increased. Also, the danger of collision is not limited to oncoming vehicles. For example, in the case of pedestrian 4 as well, assuming that vehicle 1 turns left at Urahana 1 intersection, for example, if you try to cross Hana 1 Kyoto Street, it is actually behind car 1 that went straight to turn left at Hanamachi 3 intersection. There may be a collision at L206 between 1 intersection and Hanamachi 3 intersection. Or, from the same assumption, there may be a collision with vehicle L1 that crosses Dahua Street and turns left at Hanamachi 3 intersection. On the other hand, there is a reverse case. For example, from the viewpoint of pedestrian 3, if vehicle 1 thinks that it will turn left at Hua 3 intersection instead of Urahana 1 intersection, it will cross Urahana street, and if vehicle 1 actually turns left at Urahana 1 intersection, it will collide at L210 There is a possibility that it will.

[0032] As described above, a collision accident often occurs in a situation where there are a plurality of possible actions. In order to avoid such a collision accident, the collision risk judgment unit 107 causes duplicate actions. If this is the case, determine the area at risk of collision. For example, when there is an overlap in behavior, a point where both behaviors are different is determined as an area where danger occurs. For example, in the case of Fig. 12, two actions are estimated as future actions. Specifically, “N102 → L210” is assumed as the future action “001”, and “N102 → L206 → N103 → L209” is assumed as the four derived fi movement “002”. The different areas of the two actions are “L2 10” when turning left at Urahana 1 intersection and “L206 → N103 → L209” when going straight ahead and turning left at Hanamachi 3 intersection. Therefore, these are determined as collision areas where there is a possibility of collision.

The collision information notification unit 108 is means for notifying information on the risk of collision determined by the collision risk determination unit 107. Conventionally, a technique related to so-called inter-vehicle communication has been disclosed in which communication between vehicles using millimeter waves or microwaves is performed to support safe driving. In this inter-vehicle communication, the vehicle ID, current position, speed, etc. are generally transmitted, and used to prevent encounter collisions at intersections, for example. In addition, the communication method is a broadcast type in which the vehicle transmits the ID, position, and speed of the vehicle to an unspecified number of people around the vehicle (for example, several tens of meters). Display or warn accordingly. On the other hand, there is also a PtoP type communication method in which predetermined vehicles communicate with each other in terms of position, speed, etc. and perform cooperative driving. There is also a hybrid system that switches them according to the situation. Here, a description will be given with a so-called broadcast type embodiment.

For example, in the system configuration diagram of FIG. 4, the collision information notification unit 108 of the moving body 100 notifies information related to the risk of collision within a predetermined range (for example, within several tens of meters). In vehicle-to-vehicle communication, it is common to transmit the ID, speed, position, etc. of the vehicle of the host vehicle as described above, but in addition to these, the collision information notification unit in this embodiment determines the collision risk. The collision area determined by the unit 107 is notified. On the other hand, the moving body 200 is a receiving-side moving body, which is received by the collision information receiving unit 109 and displays information on the danger of collision by the collision information display unit 110. Specific examples will be described below with reference to the drawings.

FIG. 13 shows an example of display on the collision information display unit 110 which is a display screen of, for example, car navigation (car navigation). The situation is similar to that shown in Figure 12.

[0036] Specifically, the vehicle 1 is currently traveling straight on Kyoto 1 on Hana 1 Kyoto Street, and further, the vehicle 1 has a left turn intention. Based on this intention, the collision area is determined in the vehicle 1, The notified information is received by the vehicle 2 that is the oncoming vehicle, and the car navigation screen of the vehicle 2 is shown.

[0037] Now, Vehicle 1 is going straight on Hana 1 Kyoto Street. Note that this is the vehicle ID, position and speed. This is possible with conventional inter-vehicle communication technology that provides notification. In addition, the force that vehicle 1 shows the intention to turn left Based on this intention, the collision area is determined in vehicle 1, and the area is indicated by, for example, an arrow based on the notified information. Please indicate ", there is a possibility of a collision. From vehicle 2, which is an oncoming vehicle, it is assumed that a left turn of vehicle 1 alone would make a left turn to Urahana 1 intersection, and if it was going to turn right at Hanamachi 3, there was a risk of collision. As shown in Fig. 5, it is possible to drive safely by displaying the collision area determined from the overlapping behaviors.

FIG. 14 is a flowchart showing an example of an operation until the user's future action is estimated and the collision information is displayed in the collision information notification system of the present embodiment. FIG. 15 is a flowchart showing an example of a specific operation in estimating future actions. The operation of the present invention will be described below with reference to the flowcharts of FIGS.

First, the position information detecting unit 101 detects position information (step S 101). Then, the map information stored in the map information storage unit 103 is referred to (step S102), and for example, position information detected in the latitude / longitude information is converted into a node series (step S103). Meanwhile, the intention information is detected by the intention information detection unit 104. For example, it is detected whether or not intention information such as a right / left turn blinker is detected (step S 105) .If it is detected, the process proceeds to step S 105, and if not detected, the process returns to step S 101, and the loop of this operation is performed. repeat. Position information and intention information are detected as the user travels.

[0040] When intention information is detected in step S105 (Yes in step S105), the behavior rule stored in the behavior rule accumulation unit 111 is referred to (step S106). Then, the map information stored in the map information storage unit 103 is referred to (step S107), and the behavior estimation unit 105 estimates the future behavior (step S108).

FIG. 15 shows an operation flow for estimating future actions. To estimate the future behavior, first, the behavior corresponding to the detected intention information is referred to in the action rule (step S201), and the range of the behavior is referred to (step S202). For example, as shown in Fig. 10, action rules are accumulated. For example, when a left winker is detected, a left turn action is taken within 50 meters. On the other hand, the current position of the user detected by the position information detection unit 101 is referred to (step S203), and the map information stored in the map information storage unit 103 is referred to (step S203). S204), and calculates the point where the action is possible (step S205). Then, it is determined whether or not the corresponding point exists (step S206). If there is a possible point (Yes in step S206), the process proceeds to step S207. If not (No in step S206), the estimation of the future action is terminated. For example, in the case of Fig. 9, it is estimated that two points, the Urahana intersection (N102) and the Hanamachi 3 intersection (N103), can be turned left.

[0042] If it is determined that the corresponding point exists (Yes in step S206), then the behavior of transition to each point is estimated. First, a flag is set at one point (step S207), and the behavior of the point shift is estimated (step S208). For example, in the case of Fig. 9, the left turn at the Urahana intersection (N102) is first made, and the route forwarded to Urahua Street (L210) is estimated as the future action. Next, in order to estimate the behavior at the remaining points, it is determined whether or not there is a point where the flag is not set (step S209). If there is a point, the process returns to step S207 and the above operation is repeated. In the case of Fig. 9, the route to the other Hanamachi 3 intersection (N103) and the left turn to Dahua Street (L209) are estimated as future actions. When future behavior is estimated for all points (No in step S209), the estimation of future behavior is terminated.

Next, it is determined whether or not there are a plurality of future actions in the overlapping action determination unit 106 (step S 109). If there are a plurality (Yes in Step S109), the process proceeds to Step S210. If there is not a plurality (No in Step S109), the process returns to Step S101.

[0044] If there are multiple items (Yes in step S109), the collision risk determination unit 107 determines the risk of collision. For example, an area where there is a risk of collision is determined (step S110). For example, a different point in the estimated future behavior is set as the region. For example, since L206 N103 L210 L209, which is a different area as shown in FIG. 12, may collide with a third party, it is possible to avoid danger by using these areas as collision areas.

Next, the collision information notification unit 108 is notified of the collision information (step S 111). On the other hand, on the third party side, the collision information receiving unit 109 receives the collision information (step S112), and the collision information display unit 110 displays the collision information (step S113). Figure 13 shows an example of the collision information display. Since it can be estimated that the oncoming vehicle trying to make a left turn may take multiple actions as indicated by the arrows, the driver should be alerted. Yes. this Thus, it is possible to avoid the risk of collision and to travel more safely.

In the present embodiment, the future behavior estimated based on the intention information is estimated based on the behavior rule stored in the behavior rule storage unit 111. Specifically, as shown in Fig. 10, when a left winker is issued, based on the behavior rule of turning left within a range of 50 meters, the map information ability is calculated by calculating multiple points where left turn is possible and estimating the behavior. I was doing. However, the estimation of future behavior is not limited to this. Hereinafter, explanation will be given using a specific example.

[0047] FIG. 16 is a diagram showing an example of user intention information at an intersection having a complicated branch route and future actions that can be taken in response thereto. In Fig. 16, Hanshin 1 Street, Kyoto 2 Street, Nara 3 Street, Urayama 5 Street are five-way intersections that cross at Raku 4 intersection. The route is not necessarily a four-way intersection where the streets intersect at right angles, but there are various road shapes depending on the location. In addition, the more complicated the route, the more likely the user's intended behavior and the third party's recognition will be, and the more they will collide. For example, in Fig. 16, it is assumed that vehicle 2 is traveling straight on Hanshin 1 street in the direction of Raku 4 intersection. On the other hand, suppose that vehicle 1 is going straight on Nara 3 streets in the direction of 4 easy intersections, and is now making a left turn blinker and indicating the intention to turn left. Here, the user of the vehicle 2 thinks that the vehicle 1 turns left at the Raku 4 intersection and goes straight on Hanshin 1 street, and goes straight ahead. On the other hand, the intention to turn left for the user of vehicle 1 means the way to 5 back hills, and may not be the force along the route to 5 back hills. And in such a case, both sides are going to the Raku 4 intersection to Urayama 5 streets, and there is a possibility that they will collide.

FIG. 17 is the same route as FIG. 16, and is an example that may occur when the vehicle 1 makes a right turn display. For example, for the user of vehicle 1, the right turn means a route that goes from the Raku 4 intersection to Kyoto 2 streets. On the other hand, for pedestrian 3, the right turn may be assumed to be a route to Raku 4 intersection to Hanshin 1st Street (L213) and cross 2 Kyoto streets. In such a case, there is a possibility that both will collide. In this way, collisions often occur when there is a difference between the intention information of one side and the actual behavior of the other side and the behavior that the other side thinks. Overlapping behavior also has various patterns. Therefore, not only the action rules shown in Fig. 10, Information on multiple actions that can be taken in response to information (hereinafter referred to as point information) is accumulated, and point information is referenced according to detected intention information and position, and future actions are estimated. Also good. The point information can be accumulated as, for example, a road network that is one of the map information accumulated in the map information accumulation unit 103.

[0049] FIG. 18 is a diagram showing an example of point information indicating information regarding a plurality of actions corresponding to intention information for each point. The location information corresponds to the “location” information of the user, the “intention information” issued at that location, and the “estimated behavior” information indicating the behavior estimated for the intention information issued at that location. It has been accumulated. For example, the spot information “001” is information relating to the spot shown in FIG. The point information “001” is information when the user is “L216” as the current location and “N106 → (from) N105” as the direction, that is, when Nara 3 streets in FIG. . At this time, if “Left winker” is issued as intention information, one of the presumed actions is “N105 → L212 → N104 → L215”, that is, the action of going through the back five ways toward the Raku 4 intersection, One is “N105 → L212 → N104 → L211”, that is, the action of passing through the Hanshin 1 street from the Raku 4 intersection, and the fact that these two can be estimated. Therefore, for example, in FIG. 16, a vehicle 1 traveling straight on Nara 3 streets and taking a left turn winker refers to the location information stored in the map information storage unit 103 and the like as “N105 → L212 → N104 → L215 ”and“ N10 5 → L212 → N104 → L211 ”can be estimated.

[0050] In addition, the point information “002” is also the information when the position “L216, direction N106 → (from) N105”, that is, the direction of Nara 3 street in FIG. When “right blinker” is issued as intention information, two presumed actions “N105 → L213” and “N 105 → L214” are described as presumed actions. Therefore, for example, vehicle 1 traveling straight on Nara 3 streets and taking a right turn winker in FIG. 17 estimates “N105 → L214” and “N105 → L213” as future actions by referring to this point information. It becomes possible. Based on the estimated future behavior, it is possible to avoid a collision with a third party by determining the dangerous area as shown in this method and notifying the collision information.

[0051] Further, although not shown in the drawing, the estimated future behavior is recorded in the point information, for example, as the point information "003" even in a state where the winker is not taken out as intention information. The That is, when “winker” is not given as intention information, two presumed actions “N105 → L214” and “N105 → L212 → N104 → L215” are described as presumed actions.

[0052] As described above, the intention information indicated by the user and the action actually taken by the user are not necessarily one, and there may be a plurality of actions, and the action varies greatly depending on the shape of the point or the route. And even if you are unfamiliar with or familiar with the point, you may assume that the other person takes a certain action, resulting in a collision. Therefore, as shown in this example, for example, intention information for each point and point information estimated as a user's action in the case of the intention information are accumulated, and by referring to this point information, the future can be appropriately It is also possible to estimate actions and prevent collisions.

[0053] In the present embodiment, the collision information in the collision information notification unit 108 is a force that was for a third party. This is not limited to this, and can be displayed by a similar method. is there. Furthermore, notifications may be given not only to third parties but also to the vehicle. For example, FIG. 13 shows collision information of vehicle 1 against vehicle 2 that is an oncoming vehicle. Specifically, vehicle 1 shows the intention to turn left. Ura Hana 1 Indicates that it is possible to turn left at the 1st intersection, and vice versa. It was a warning about the danger of oncoming vehicles. However, such collision information may be for not only the oncoming vehicle but also the own vehicle.

FIG. 19 is a diagram showing an example of a display for notifying the own vehicle when there is a risk of collision due to overlapping actions. For example, a user of vehicle 1 is now going straight on Hana 1 Kyoto Street, turning left turn signal and turning left on Dahua Street. On the other hand, if a left turn blinker is issued at the current position, as described above, the behavior of turning left at Urahana 1 intersection and the two overlapping actions of turning left at Hanamachi 3 intersection are calculated, resulting in a collision with a third party. The area that causes the risk is calculated. Then, as shown in Fig. 19, the area causing the risk of collision with a third party is displayed on the display screen of the vehicle to call attention.

[0055] Suppose that you are going to turn left at Hanamachi 3 intersection, and that you are relieved that it is okay because you are taking the left turn signal. However, there is a possibility that a third party thinks that he will turn left at the 1st intersection in front of the Hanamachi 3 intersection. The oncoming car may go unnoticed and collide with the car. Thus, by notifying the vehicle of the collision information and raising the danger in this way, an extraordinary effect of avoiding the collision by trying to drive more safely is produced.

Therefore, for example, in the system configuration diagram of FIG. 4, a collision information display unit or the like may be further provided in the moving body 100, and the information notified by the collision information notification unit 108 may be alerted to the own vehicle! /. Therefore, the moving body 200 is not necessarily an essential component in the present invention! / ヽ. Further, the notification in the collision information notification unit 108 can be notified by a voice or a warning lamp that is not necessarily limited to the screen display.

Therefore, the invention described in this embodiment can be realized by the components shown in FIG. FIG. 20 is a block diagram showing an example of the configuration of the collision information notification device when notifying other vehicles only of the risk of collision, but notifying only the own vehicle. That is, the minimum configuration for carrying out the present invention includes a position information detection unit 101 that detects current position information of a moving object, and an intention information detection unit 104 that detects action intention information related to the future action intention of the moving object. , A map information storage unit 103 that stores map information related to the positional relationship of points, and an action estimation unit that estimates future actions related to a future turn or stop from the position information, the action intention information, and the map information. A duplicate action determination unit 106 for determining whether or not the future action estimated by the action estimation unit 105 is duplicated, and a future action for which the duplicate is calculated calculated by the duplicate action determination unit 106. In addition, a collision risk determination unit 107 that determines the risk of a collision with another moving body, and a collision information notification unit 108 that notifies information on the collision risk determined by the collision risk determination unit 107. Constitution Is done.

FIG. 21 is a block diagram showing an example of a hardware configuration that implements the functional configuration of the collision information notification device shown in FIG. The mobile unit 100 according to the present embodiment is a car navigation device, and includes a position information detection unit 101, a CPU 131 that performs calculation processing, a primary memory 132 that stores the calculated information, and a collision information notification unit 108. , An in-vehicle information detection unit 133, an external memory 134, and a program storage unit 135. The in-vehicle information detection unit 133 corresponds to the intention information detection unit 104 in the present invention, and is means for detecting in-vehicle information such as a blinker as intention information. The external memory 134 is realized by a hard disk etc. The map information storage unit 103 according to the present invention is a means for storing map information. The CPU 131 is a means for processing the detected information and the accumulated information through the primary memory 132. By executing the program stored in the program storage unit 135, the CPU 131 in the present invention. This is means for performing calculations performed by the behavior estimation unit 105, the duplicate behavior determination unit 106, and the collision risk determination unit 107. In addition, with the hardware configuration shown in the figure, the program stored in the program storage unit 135 and the data stored in the node disk can be changed, and FIGS. 4, 17, 28, 29, 35, 44, 47 and 49 It is possible to realize a configuration represented by all functional block diagrams.

[0059] In the present embodiment, an explanation has been given by taking an example of overlapping behavior at the time of a left or right turn at an intersection. However, it is not limited to a right or left turn at the time when the estimated behavior is duplicated. May also occur. For example, FIG. 22 is a diagram for explaining an example of estimated behavior estimated for intention information at the time of stopping.

In FIG. 22, truck 1 is traveling on Hanshin 7 street. Suppose now that the stop intention information of truck 1 is detected. In general, there are cases where the hazard lamp blinks to indicate the intention of stopping and the vehicle stops on the shoulder or enters the parking lot. Therefore, the intention information detection unit 104 in this example detects the hazard lamp as an intention to stop in this way. In addition, as a conventional inter-vehicle communication technology, not only right / left turn information but also a signal for stopping is prepared in advance, and there is a technology for communicating with each other, and this signal for stopping may be detected. On the other hand, there are multiple future actions that can be considered only by the intention of stopping, and there is a risk of collision with a third party. For example, it is estimated that the vehicle 2 in the rear will stop at the parking lot of the store B by looking at the truck 1 indicating the intention to stop. And sometimes you want to overtake track 1 from the right and speed up. On the other hand, if the user of truck 1 tries to stop at the parking lot of store A, not store B, the two may collide. In particular, trucks and business operators may take actions that are different from the user's thoughts, such as swinging their own vehicle to the right and stopping to the left or entering the parking lot in the back. Errors often occur in speculation. Similarly, the user of opposite car 3 also thinks that it will stop at the parking lot of store B, and if it tries to stop at the parking lot of store A at the SB store, it will still collide. Become . In this way, even if the intention to stop is shown, the estimated future behavior is not necessarily one, and there may be a plurality of future actions, and if there is an error in recognition between the vehicle and the third party, I will end up. Therefore, as shown in this embodiment, it is possible to avoid collisions by using this method that estimates future behavior and provides information on the danger of collision when duplicate behavior is determined. Become. The following describes the operation when the vehicle is stopped.

FIG. 23 is a diagram showing an example of an estimated action estimated for the intention information at the time of stopping shown in FIG. 22 by a node and a link. First, FIG. 23 shows the same situation as FIG. 22 and shows an example of a road network stored in the map information storage unit 103. In a road network, nodes are generally assigned not only to intersections but also to branch points of routes and landmarks such as facilities. Furthermore, a network is constructed by links connecting nodes. In the case of FIG. 23, the seven Hanne-Sen are represented by “: L221”, “N121”, “: L222”, “N122”, “L223”. In addition, “N124” is assigned to the parking lot of store B, “N123” is assigned to the parking lot of store A, and links “L225” and “L224” are connected. Figure 24 shows the details of the node information accumulated for the area including the current location of the mobile object. For example, the area shown in FIG. 23 has area ID “E11”. In FIG. 24, “N121”, “N122”, “N123”, “N124” and the like are shown as the nodes existing in the area ID “E11”. Further, detailed information of each node is accumulated. For example, “N123” is a node assigned to the parking lot of store A, where the location is “136 degrees 04 minutes east longitude, 34 degrees 52 minutes north latitude”, the name “store A parking lot”, the surrounding link “L224 (30 m)”, The node type “parking lot” and the stop information “possible”, which is information indicating whether or not the location can be stopped, are stored. Or “N124” is a node assigned to the parking lot of store B, where the location is “136 degrees 02 minutes east longitude, 34 degrees 57 minutes north latitude”, the name “A store parking lot”, the surrounding link “L225 (20m)”, The node type “parking lot” and stop information “possible” are stored.

[0062] Similar to the left-right turn behavior estimation described above, the behavior estimation unit 105 is based on the current position detected by the location information detection unit 101 and the intention information detected by the intention information detection unit 104. The point where the action corresponding to the intention is possible is searched with reference to the action rule. FIG. 25 shows the same situation as FIG. 23 and the like, and is a diagram for explaining a search for a place where the vehicle can be stopped. In Figure 25 Assume that a hazard, which is intention information for stopping, is detected at a point “136 degrees 01 minutes east longitude, 34 degrees 55 minutes north latitude”. Note that, as described above, the moving direction of the user is calculated from, for example, the node ID series and the like, and for example, it is assumed that Hanshin 7 Street is heading east. Referring to the action rule “010” shown in FIG. 10, in the case of a hazard, it is stated that the vehicle “stops” within “100 meters in the moving direction and 30 meters around”. Therefore, a point within the area where the stop intention information is detected is searched for in the region.

[0063] Calculate an area of 100 meters in the east direction and 30 meters around it based on “136 degrees 01 minutes east longitude, 34 degrees 55 minutes north latitude” where the intention information of the stop was detected For example, it becomes a rectangular area filled with diagonal lines 34 degrees 52 minutes north latitude 34 degrees 58 minutes north latitude 136 degrees 01 minutes east longitude 136 degrees 11 minutes east longitude 11 minutes. On the other hand, the map information stores the latitude and longitude information of the parking lot expressed as a node as shown in FIG. 24. Based on this latitude and longitude information, the map information is searched for points where parking is possible. It will be. In the case of this example, the store A parking lot (N123) and the store B parking lot (N124) are within the calculated areas and are searched for points that can be stopped.

[0064] FIG. 26 is a diagram illustrating an example of a route to a stop possible point that the behavior estimation unit 105 estimates as a future behavior. Next, the behavior estimation unit 105 searches for a route to a point where the vehicle can be stopped, and estimates it as a future behavior. First, the route “N121 → L225 → N124” from Hanshin 7 Street (L221), which is the current location, to the B store parking lot (N124) as the future action “001”, The route “N121 → L225 → N121 → L222 → N122 → L22 4 → N123” is estimated as a future action. Then, from these overlapping actions, it is possible to identify an area where there is a risk of collision by using the method shown in the present invention, notify the oncoming vehicle or the rear vehicle, and avoid the collision.

[0065] Furthermore, the present invention can be applied not only to stopping and turning left and right, but also to cases where both of these actions can be considered. This will be described below with reference to FIG.

[0066] FIG. 27 is a diagram showing an example of a case where overlapping actions are estimated for the same intention information. Now, it is assumed that bus 1 takes out the left winker and the intention information detection unit 104 detects this left winker. On the other hand, from the rear vehicle 2 and the oncoming vehicle 3, you may not know the power of bus 1 trying to stop at the bus stop or whether you are going to turn left at the intersection. . For example, the rear vehicle 2 thinks that the bus 1 is about to stop at the bus stop, and tries to overtake it, or the oncoming vehicle 3 tries to turn right. On the other hand, if the driver of bus 1 leaves the winker to the left for a left turn at an intersection that does not stop at the bus stop, they will collide.

FIG. 28 is a diagram showing another example in a case where overlapping actions are estimated for the same intention information. In addition to buses, as shown in Fig. 28, for example, if an oncoming vehicle is making a right / left turn blinker at an intersection, etc. In many cases, it is likely to collide if you do. In this way, the winker is generally used not only for indicating intention of turning right and left but also for indicating intention to stop in that direction. Thus, it is possible to avoid a collision by estimating a plurality of actions from the intention information shown in the present invention. The following is an explanation using the bus example shown in Fig. 27.

In the present embodiment, for example, as the action rule “011” shown in FIG. 10, when the left winker is detected, the fact that the vehicle “stops” within the area “moving direction 100 meters and left 30 meters” is accumulated. is doing. Then, similarly to the method shown in the present embodiment described above, the behavior estimation unit 105 calculates a left-turnable point and a stopable point from the map information, and estimates future behavior.

FIG. 29 shows the same situation as FIG. 27, and illustrates the calculation of the stop point and the left turn point. Suppose that the bus currently leaves the left winker at the point of 136 degrees 40 minutes east longitude, 35 degrees 50 minutes north latitude, indicated by a white circle. From the action rules shown in Fig. 10, the left turn signal is "turn left in the direction of travel (50 meters)" (action rule "001") and "stop" in the direction of travel 100 meters and the left side 30 meters ( It is accumulated as action rule “011”). Therefore, referring to the map information, search for a stopable point and a left-turnable point, and estimate future behavior. In Figure 29, first, the Naramachi intersection (N132) is located at 136 degrees 43 minutes east longitude, 34 degrees 51 minutes north latitude, and this Naramachi intersection is located within 50 meters of travel direction. It is searched as a possible point. On the other hand, there is a possibility of stopping based on the current location, within a 100 meter travel direction and 30 meters to the left, a bus stop at the position of 136 degrees 41 minutes east longitude 34 degrees 52 minutes north latitude “Heianji-mae” Is done. FIG. 30, like FIG. 24, shows information about bus stops accumulated as one of the points where parking is possible. Node ID “131” location “Eastern longitude 136.41 north latitude 34.52”, name “Heianji”, node type “bus stop”, information that can be stopped, etc. are stored as map information. The future behavior is estimated using these left turnable points and stopable points. As future action “001”, it is estimated that “N131 → L235 → N131” and the stop to the bus stop. On the other hand, it is estimated that the left turn at the intersection of “N131 → L232 → N132 → L234” as 4 derived fi motion “002”. In the same manner as described above, the risk of collision is determined from a plurality of calculated future actions, and notification is made. Collisions can be avoided by estimating multiple actions, such as turning left or stopping.

[0070] In the present embodiment, vehicle-to-vehicle communication between mobile unit 100 and mobile unit 200 has been described as a so-called broadcast type, but is not limited to this. For example, the PtoP method may be used in which predetermined vehicles communicate with each other information such as their own vehicle positions to collaborate and urge attention to prevent collisions! /.

FIG. 31 is a block diagram showing an example of a collision information notification system configuration of the present invention in PtoP inter-vehicle communication in which position information and the like are communicated with each other. In addition to the components of the collision information notification device shown in FIG. 4, the moving body 100 in FIG. 31 is newly provided with a position information transmission / reception unit 113 and a notification partner identification unit 112. Further, the mobile object 200 is newly provided with a second position information transmitting / receiving unit 114 and a second position information detecting unit 115. Here, the position information transmitting / receiving unit 113 is an example of “position information receiving means for receiving the current position of the moving body from another moving body”, and the collision risk determining section 107 is the “duplicate action determining means”. Therefore, it is an example of a “collision risk determination unit that determines an area where there is a risk of collision with another mobile object when it is determined that a plurality of actions can be taken. The collision information notifying unit 108 indicates that, based on the current position received from another moving body, the other moving body may exist in an area where there is a risk of collision in the future. It is an example of the “collision information notification means” for notifying a mobile object of information on the danger of collision. First, in the case of this example, the position information transmission / reception unit 113 transmits the position information detected by the position information detection unit 101 in the moving body 100. On the other hand, the second position information transmitting / receiving unit 114 in the moving body 200 is a means for receiving position information of other vehicles. Also The second position information detection unit 115 is configured by, for example, a GPS or the like that detects the position of the moving body in the same manner as the position information detection unit 101, and the second position information transmission / reception unit 114 transmits it to another vehicle. In FIG. 31, the force S, which shows only the moving body 100 and the moving body 200, and other plural moving bodies transmit / receive position information and the like, and predetermined moving bodies construct an ad hoc network. This is a so-called PtoP inter-vehicle communication system. Further, in FIG. 31, the system element of the present invention shown in FIG. 4 and the like is added to the moving body 100, and the present invention is realized.

[0072] Further, if communication with other mobile objects and the position of other mobile objects can be grasped in this way, a mobile object that is at risk of collision is not just broadcasted to the surroundings. It is also possible to notify the specified moving body. The notification partner specifying unit 112 is a means for specifying a partner to notify the collision information based on the position information of the other mobile body received by the position information transmitting / receiving unit 113. Hereinafter, description will be made with reference to FIG.

[0073] In FIG. 12, as described above, vehicle 1 is making a intention to turn left, and multiple actions are estimated when turning left at Urahana 1 intersection or turning left at Hanamachi 3 intersection. Dangerous areas have been identified. For example, if you are going to turn left at the Hanamachi 3 intersection, you may collide with an oncoming vehicle 2 or a pedestrian 4 that you are thinking of. Therefore, among the third parties received by the location information transmission / reception unit 113, a user who is located in an area where there is a possibility of a collision or a user who will be located in the future is specified as a notification partner and notified. May be. In vehicle-to-vehicle communication, many mobiles communicate with each other, and it is necessary to use limited communication capacity appropriately, especially at intersections. Therefore, by specifying the notification partner in this way, it is possible to use the minimum communication capacity and appropriately promote safe driving.

In the present embodiment, the information notified in collision information notifying section 108 is described by taking an example of an area with a possibility of collision determined by collision risk determining section 107! I have Specifically, in Fig. 12, of the future actions of vehicle 1, the different actions are `` L210 '' when turning left at Urahana 1 intersection and `` L206 → N103 → L209 '' when turning left at Hanamachi 3 intersection. These are determined as collision areas and notified as collision information. As a result, the third party who receives the notification knows the other behavior different from the behavior that he / she was aware of. It is possible to avoid collisions. Currently, map information is generated independently by each company, and road network structures such as node IDs and link IDs are generally unique to each company. On the other hand, there is a technology to unify these (Kiwi format). When the road network structure is common as described above, the notification is received by notifying the collision area by the ID series as shown in this embodiment. A third party can display information about the collision area on the vehicle. However, IDs are not always common. Therefore, in order to make the identified collision area available to a third-party mobile body, a conversion means is provided, and the converted collision information is notified.

FIG. 32 is a system configuration diagram in the case where conversion means is provided in order to make collision information available to a third party mobile body. In FIG. 32, a collision information converting unit 116 is added to the components shown in FIG. For example, even if “L210” and “L206 → N103 → L209”, which are the collision areas determined in FIG. 12, are notified, where the third party mobile body 200 indicates! There are cases where I can't figure out. Therefore, the collision information conversion unit 116 uses, for example, the map information stored in the map information storage unit 103 and converts it into latitude and longitude information that can be commonly recognized by any mobile body. In general, map information is stored in correspondence with the ID of each node or link and latitude and longitude information. Therefore, by referring to this map information, the ID of the area specified as the collision area can be converted into latitude and longitude information.

FIG. 33 is a diagram showing collision information obtained by converting a collision area from latitude / longitude information that can be commonly recognized by any mobile body from a series of IDs. “L210” identified as the collision area has been converted into a value indicated by latitude and longitude as “135.30 east longitude, 35.18 north latitude”. Similarly, the ID of each point identified as a collision area, such as “L20 6” and “Latitude 135.30, North latitude 35.18”, is converted into latitude and longitude values. As the collision information, for example, the “vehicle ID”, “behavior”, and “collision area” converted into the latitude / longitude are notified by the collision information notification unit 108. On the other hand, the third party who receives the collision information can grasp where the danger of the collision is by mapping the point indicated by the latitude and longitude values on the map of the own vehicle. .

[0077] In the example shown in Fig. 33, the power of converting the ID of a node or link into latitude and longitude information Furthermore, it may be converted into a series of latitude and longitude information in which nodes and links are complemented at a predetermined interval, and transmitted. Furthermore, taking into account the position of the third party that simply sends the collision area, it is necessary to specify the distance to the information | Also good. This will be described below with reference to the drawings.

FIG. 34 is a diagram showing another example of a region where there is a risk of collision with a third party with respect to the user's future behavior. FIG. 34 shows the same situation as FIG. Currently, it is estimated that there are several actions: vehicle 1 takes a left turn signal at Hana 1 Kyoto Street, turns left at Urahana 1 intersection, and turns left at Hanamachi 3 intersection. In Fig. 34, there are pedestrians 3 who are going to cross Hana 1 Kyoto Street and pedestrians 4 who are going to cross 3 Hanamachi Streets. For these pedestrians, it may be assumed that the vehicle 1 makes a left turn at an intersection at the back of the front as described above, and there is a risk of a collision. In this embodiment, therefore, the collision area where there is a risk of collision is calculated and notified to the pedestrian. On the other hand, it is not always necessary to send the same information to both. For example, pedestrian 3 can avoid the danger by notifying vehicle 1 that there is a risk that vehicle 1 will then go straight through Urahana 1 intersection and collide with him. It is not always necessary to go to the left after the intersection and go to Taihua Street. On the other hand, pedestrian 4 needs to be notified that vehicle 1 has a risk of colliding with himself who is going to cross Hwakacho after crossing Hanamachi 3 and going to Taihua Dori. Therefore, it is also possible to notify the sequence up to the point where there is a risk of collision in consideration of the current position of the third party or the future position of the third party rather than simply reporting the collision area. FIG. 35A is a diagram showing an example of information when collision information is transmitted to the pedestrian 3. FIG. 35B is a diagram illustrating an example of information when collision information is transmitted to the pedestrian 4. In FIG. 35, pedestrian 3 is notified of the collision area as “collision information” with the area up to the current location of pedestrian 3 as “135 degrees and 33 minutes east longitude, 35 degrees 20 minutes north latitude”. On the other hand, for pedestrian 4, the collision area is “135 degrees and 33 minutes east, 35 degrees 20 minutes north → 135 degrees 34 minutes east, 35 degrees 20 minutes north → 135 degrees 34 minutes east, 35 degrees 18 minutes north” The area up to the current location of pedestrian 4 is attached and notified as collision information. In such notification of urgent information, the amount of information to be transmitted becomes a problem. Therefore, the position information of the third party to be transmitted is detected and transmitted up to the detected position. As a result, the amount of information can be reduced, and information can be transmitted efficiently.

Furthermore, in the present embodiment, the position information and intention information of the own vehicle is detected in the moving body 100 to estimate a plurality of behaviors, and the determination of the risk of collision is performed. It is also possible to detect the position information and intention information of a third party that does not perform these processes, and to determine the risk of these third parties. For example, in the system configuration diagram shown in FIG. 20, the position information detection unit 101 detects the position information of a predetermined third party, and the intention information detection unit 104 detects the third party's intention information. It is.

[0080] In recent years, research on inter-vehicle communication has been conducted, and in this inter-vehicle communication, vehicle ID, speed, position, and intention information such as a blinker are transmitted in a unified format. Even if you receive intention information such as turning left, you may not know which route you actually turn left. Therefore, as shown in the present embodiment, it is possible to avoid danger by notifying the information related to the dangerous area, but on the other hand, it has a function to display information related to the collision area. It is not always true. Especially in inter-vehicle communication, not all vehicles have communication units with the same functional specifications. For example, only unified information such as location information and intention information is transmitted. In some cases, each processing is performed on the receiving side. Therefore, a plurality of actions are estimated from the position information and intention information of the third party detected by the position information detection unit 101 and intention information detection unit 104, and the collision information notification unit 108 notifies the vehicle of the danger. Good as.

[0081] FIG. 36 is a diagram showing an example in which an action is estimated by the own vehicle based on the current position sent from the oncoming vehicle and the intention information of the left turn, and the user is notified of the risk of collision. is there. The car navigation screen shows the location of the vehicle that is about to turn right at the intersection and the position of the oncoming vehicle. On the other hand, the oncoming vehicle has the power to indicate a left turn. On the other hand, it does not necessarily mean that it is going to turn left at the intersection. It may be the power to stop at the parking lot of “Family K”. It is done. Therefore, the dangerous area is identified from the estimated action, and the user is notified of this by an arrow. In particular, in car navigation systems, it is possible to alert the driver and provide more safety by notifying the danger and not just notifying the danger, but notifying where and how dangerous. it can. The oncoming vehicle on the sender side will also be unified. Even if the vehicle does not have a function to display information about the estimation means or the dangerous area, the partner can estimate the behavior of the vehicle. Being careful, it is possible to reduce the risk of collision.

[0082] Further, the plurality of behavior estimations in the present invention are not limited to behaviors on roads and intersections! /. For example, it can be used even in a parking lot.

FIG. 37 is a diagram for explaining an example of behavior estimation in a parking lot. In Fig. 37, suppose that vehicle 2 is looking for an empty parking lot. At this time, it is often difficult to know the power of the vehicle 1 trying to park, or whether it is going out. For example, the user of the vehicle 2 thinks that the vehicle 1 is about to come out, and if it proceeds as it is, there is a risk that both will collide. Therefore, a collision can be avoided by using the method shown in the present invention. For example, when vehicle 1 stops, that is, in the case of parking, and in the case of left turn, that is, in this example, the multiple actions that are about to take place are estimated as future actions, and the collision area is determined from the estimated future actions. A collision can be avoided by identifying and notifying the vehicle 2. Furthermore, in the case of this example, for example, the movement history of the vehicle 1 is accumulated, and if it has stayed for a predetermined time until now, it is going to come out. Predict 1 action and notify to that effect. For vehicle 2, if vehicle 1 is about to leave, enter it, while if vehicle 1 is about to park, give up and look for another parking space, avoiding collisions and more efficient parking management. Can also be performed.

[0084] It should be noted that the method of accumulating the movement history and predicting the destination of the vehicle will be described in detail later.

[0085] (Embodiment 2)

In the present embodiment, the detected position information is further accumulated, the user's destination is predicted from the accumulated position information, and a region that may collide is determined using the predicted destination. Will be described.

FIG. 38 is a block diagram showing a configuration of the collision information notification system according to the second embodiment.

In the present embodiment, the collision information notification device of the moving body 100 includes a position information detection unit 101, a node sequence conversion unit 102, a map information accumulation unit 103, a movement history accumulation unit 117, an intention information detection unit. It comprises an output unit 104, a behavior estimation unit 105, a destination prediction unit 118, a duplicate behavior determination unit 106, a collision risk determination unit 107, and a collision information notification unit 108. The same components as those in the first embodiment are given the same reference numerals. In the moving body 100 of the present embodiment, the position information detection unit 101 is an example of “current position acquisition means for acquiring the current position on the map of the moving body”, and the intention information detection unit 104 This is an example of “intention information detection means for detecting intention information issued for intention display indicating future behavior of the user”, and the movement history accumulation unit 117 detects “the detected intention information and the intention information are detected”. The movement position storage means for storing the movement history of the moving body including the current position at the time of the movement and the action selected with respect to the intention information within the predetermined range of the current position force. The determination unit 106 determines whether or not a different action has been selected for the same intention information within a predetermined range from the current position when the intention information is detected. Are different for the same intention information. “Duplicate action determination means for determining whether or not an action can be selected” and “the point information accumulated for a point within a predetermined range from the current position acquired when the intention information is detected”. Is an example of the duplicate action determining means for determining whether or not a different movement route can be selected for the detected intention information, and the collision information notification unit 108 is “by the duplicate action determining means, When it is determined that a different action can be selected corresponding to the intention information, a collision information notification means for notifying information on the risk of a collision with a person or another moving body and the duplicate action determination means Is an example of the above-mentioned collision information notification means for notifying the information on the risk of collision when it is determined that a different travel route can be selected for the intention information. . Further, the map information accumulating unit 103 “accumulates map information representing a map in advance, and stores the intention information and the intention for each point on the map from the movement history accumulated in the movement history accumulation unit. FIG. 6 is an example of a “map information storage unit that further stores point information indicating a travel route selected as an action corresponding to information”. “From the travel history stored in the travel history storage unit, Is a point information generating unit that generates the point information indicating the intention information and the movement route selected as an action corresponding to the intention information for each point. The movement is performed with respect to the intention information by the duplicate action determination means. When it is determined that the body can take a plurality of actions, it is accumulated in the movement history accumulation means! /, Based on the movement history! /, And a movement destination prediction means for predicting the movement destination ” It is an example.

The movement history accumulating unit 117 is means for accumulating position information detected as the user moves as a movement history. In the present embodiment, the node series conversion unit 102 converts the position information detected as the latitude and longitude values by the GPS into the ID series of nodes and links with reference to the map information, and moves this ID series. It will be accumulated as a history. Note that the node series conversion unit 102 may store values such as latitude and longitude that are not necessarily essential components. However, for example, the latitude and longitude detected by GPS in a car navigation system have an error, and since the latitude and longitude are detected at intervals of about 1 second, the amount of information can be enormous. On the other hand, in order to grasp the movement of the user, it is possible to grasp the approximate movement based on which route it is traveling on and which intersection it is passing through, etc., and in order to reduce the calculation cost in the matching described later, It is desirable to accumulate in series.

FIG. 39 is a diagram showing the movement route of the user shown in FIG. 5 by the link and node series shown in FIG. The user turns right at Hanamachi 1 intersection and goes straight on Hana 1 Kyoto Street. As the user moves, position information is detected at predetermined intervals in the position information detection unit 101 and is indicated by white circles. On the other hand, as shown in the first embodiment, the map information storage unit 103 stores a road network indicated by nodes and links as map information. The node series conversion unit 102 converts the detected position information into an ID series of nodes and links. For example, in the case of FIG. 39, the route is converted into an ID sequence such as “L203 → N100 → L204”.

The movement history accumulating unit 117 is means for accumulating the user's movement history converted into the ID series. FIG. 40 is a diagram showing the movement history stored in the movement history storage unit 117, which is represented by an ID sequence of nodes and links. The movement history is, for example, one movement from the starting point where the engine is started to the destination where the engine is stopped, and the ID sequence of the passing route is accumulated. For example, departure from “N201” as history ID “001” (for example, home), passed through “L202”, “N100”, “L204”, etc. and arrived at the destination “N208” (for example, company) The history is accumulated! [0090] Similar to the first embodiment, the behavior estimation unit 105 is a unit that estimates future behavior based on a behavior rule (not shown) based on a user's behavior intention such as a winker detected by the intention information detection unit 104. It is. Similarly to the first embodiment, the duplicate action determination unit 106 also determines whether or not a plurality of future actions are estimated, and the collision risk determination unit 107 determines a region where there is a risk of collision. . The collision information notification unit 108 notifies the collision information. This will be described below with reference to the drawings.

[0091] FIG. 41 is a diagram showing a situation in which the user is going straight on Hana 1 Kyoto Street and now has a left turn blinker, similar to FIG. 9 and the like shown in the first embodiment. For example, as shown in the above embodiment, a future action is estimated based on point information that is information related to points stored in the map information storage unit 103. In this situation, two actions are estimated as future actions: the left turn at Urahana 1 intersection and heading towards Urahua Street, and the left turn at Hanamachi 3 intersection and heading toward Dahua Street. In this way, when multiple actions are estimated, there may be a difference between the action intended by the user and the action recognized by the third party, resulting in a collision. Therefore, as shown in the first embodiment, the overlapping action determination unit 106 determines whether or not there are a plurality of future actions, and the collision risk determination unit 107 determines a region where there is a risk of collision. Become.

On the other hand, the movement destination prediction unit 118 is a means for predicting the movement destination of the user based on the movement history accumulated in the movement history accumulation unit 117. The movement history accumulating unit 117 accumulates a movement history detected in accordance with the user's normal behavior, and this movement history is generally information reflecting the user's behavior tendency. Users usually act in a certain pattern, such as commuting from home to the company, moving to a lesson, driving a restaurant, supermarket, etc., and there are generally patterns in their travel routes. By accumulating the movement history, it is possible to extract the daily patterns of these users, and it is possible to predict the future destination by using the extracted patterns. Therefore, in the present embodiment, the movement destination is predicted using the accumulated movement history. This will be described below with reference to the drawings.

FIG. 42 is a diagram for explaining the prediction of the movement destination. FIG. 42A is a diagram showing the current travel, which is the travel route up to the present time, of the user as a series of link IDs and node IDs. Figure 42B shows the current run FIG. 5 is a diagram showing an example of extracting from the movement history the history of passing the route that coincides with and turning left at Hanamachi 3 intersection. FIG. 42C is a diagram showing an example of extracting from the movement history a history of passing the route that coincides with the current travel and turning left at the back intersection 1 intersection. Fig. 42D is a diagram that calculates the possibility of each future action when there are multiple actions that can be taken as future actions. In FIG. 42A, the current travel “N201 → L202 → N100 → L204 → N101 → L205”, which is the travel route up to the present time, is displayed and displayed. This (departs from N 201 (for example, home) as shown in Fig. 41, turns left at Hana 1 intersection (N101), and now travels on Hana 1 Kyoto Street (L2 05)! Next, the history that may have passed the route that matches the current travel is extracted from the travel history.From the travel history shown in Fig. 40, the travel history that matches the current travel is the history ID “ Five of “001”, “003”, “005”, “007”, “009” are extracted.

On the other hand, the behavior estimation unit 105 estimates “N102 → L210” that makes a left turn at Urahana 1 intersection and “N102 → L206 → N103 → L209” that makes a left turn at Hanamachi 3 intersection. On the other hand, the movement history indicates that the history ID “001”, “003”, “005”, “007” has passed through the route “N102 → L206 → N103 → L209” four times in the past. On the other hand, the history ID “009” indicates that the vehicle has traveled the route “N102 → L210” once in the past. In other words, in this situation, there are two possible future actions for the user: the action of turning left at Urahana 1 intersection and the action of turning left at Urahana 1 intersection. The possibility of turning left at Urahana 1 intersection is 20% (1 ÷ 5), and the possibility of turning left at Hanamachi 3 intersection is 80% (4 + 5), so the future destination can be predicted. In the present embodiment, the current travel to be referred to is the route from the departure to the present, but the present invention is not limited to this. For example, it is possible to predict the future destination by referring to the past movement history only from the direction toward the current point. In addition, the user's behavior may depend on the day of the week. For example, the day of the week or the time of day is added as the movement history, and the destination is predicted with reference to these days.

The collision risk determination unit 107 is means for determining an area where there is a risk of collision using the obtained predicted movement destination. In the first embodiment, when the estimated future actions overlap, different areas of both actions are determined as areas where there is a risk of collision. For example In the case of 41, as in FIG. 12, it becomes the L210, L206, N103, L209 force S collision area, which is a different area of the estimated future action, and is notified. In this example, based on the destination predicted from the history, the risk of collision, for example, the user's power and probability in that area, is calculated as the risk. The collision area L210 has a movement probability of 20%. On the other hand, the collision areas L206, N103, and L209 have a movement probability of 80%. Therefore, not only the collision area is notified, but the degree of danger is attached and notified as collision information.

FIG. 43 is a diagram showing an example of a display screen that displays a notification of collision information accompanied by the movement probability of the oncoming vehicle. FIG. 43 is an example of a display screen of an oncoming vehicle that has been notified of collision information, as in FIG. 13, but is different in that it is displayed with the oncoming vehicle's movement probability. The oncoming car now leaves the left winker and goes straight through the Hana 1 intersection.However, when turning left at the Hana 1 intersection (movement probability 20%) and turning left at the Hanamachi 3 intersection (movement probability 80%) ) And urges the user to pay attention to this oncoming vehicle.

[0097] From the viewpoint of fail-safety, when there are multiple future actions, it is desirable to always notify and call attention. On the other hand, notifications that are not dangerous can lead to user distrust. For example, it is preferable to call attention, such as `` Be careful with oncoming vehicles '', even though the oncoming vehicle makes a left turn at the Urahana 1 intersection in Figure 43! . Therefore, the collision information notified by predicting the destination of the route using the movement history as shown in this example, calculating the risk from the predicted probability of moving to the destination, and notifying with the risk This will increase the reliability of the third party, which will allow the third party to pay attention and increase the effectiveness of collision avoidance.

FIG. 44 is a flowchart showing the operation of the collision information notification device in the collision information notification system of the second embodiment. FIG. 45 is a flowchart showing a detailed operation of the collision information notification device in the destination prediction process shown in step S310 of FIG. The operation of the present invention will be described below with reference to the flowcharts of FIGS. 44 and 45.

First, the position information detecting unit 101 detects position information (step S301). Then, the map information stored in the map information storage unit 103 is referred to (step S302), and for example, position information detected in the latitude / longitude information is converted into a node series (step S303). Then, it is accumulated as a movement history in the movement history accumulation unit 117 (step S304). In the meantime, intention information inspection The outgoing information 104 detects intention information (step S305). For example, whether or not intention information such as a right / left turn blinker is detected is detected (step S306) .If it is detected, the process proceeds to step S307, and if not detected, the process returns to step S301, and the loop of this operation is performed. repeat. As the user travels, position information is detected, movement history is accumulated, and intention information is detected.

[0100] When intention information is detected in step S306 (Yes in step S306), the point information stored in the map information storage unit 103 is referred to (step S307), and the behavior estimation unit 105 estimates future behavior. (Step S308). Next, it is determined in the overlapping action determination unit 106 whether or not there are a plurality of future actions (step S309). If there are multiple (Yes in step S309), the process proceeds to step S310, and if there are not multiple (No in step S309), the process returns to step S301.

[0101] When there are a plurality (Yes in Step S309), the movement destination prediction unit 118 predicts the movement destination (Step S310). In the present embodiment, the movement of the destination prediction is not limited to the force that is to be performed when future actions overlap. For example, a technique for predicting a destination and providing traffic information related to the destination is known. For example, every time position information is detected (step S101) or converted into a node series (step S303). You can do it every time you cross an intersection!

FIG. 45 shows an operation flow of movement destination prediction in the movement destination prediction unit 118.

In this embodiment, the node sequence is converted in the node sequence conversion unit 102 based on the position information detected as the user moves (step S301 to step S303). Reference (step S401). Then, the movement history stored in the movement history storage unit 117 is referred to (step S402). Then, a history that matches the current travel is extracted (step S403).

On the other hand, the behavior estimation unit 105 estimates future behavior (step S308), and refers to this future behavior (step S404). Then, it is determined whether there is a matching future action (step S405). If it exists, the process proceeds to step S406, and if it does not exist, the process ends. If it exists (Yes in step S405), the frequency of the history that matches each future action is calculated (step S406). Then calculate the movement probability of each future action (step S407). In the case of this example, as shown in Fig. 42, the behavior of turning left at Urahana 1 intersection is 20%, and the behavior of turning left at Hanamachi 3 intersection is 80%, and the destination is predicted with the estimated movement probability.

[0104] Next, the collision risk determination unit 107 determines the risk of collision! For example, different points of the estimated future behavior are determined as collision areas, and the risk (for example, the degree of risk) is determined based on the destination predicted by the destination prediction unit 118 and its movement probability (step S311). ). The collision information notification unit 108 notifies the determined collision information. Here, the behavior estimation unit 105 is an example of “the collision information notification unit that calculates the probability that the vehicle is heading for the destination based on the travel history stored in the travel history storage unit”. is there.

For example, FIG. 43 is an example in which notification is performed together with the collision area and its movement probability. The oncoming car leaves the left turn signal and goes straight at the Hana 1 intersection, but when turning left at the Hana 1 intersection (movement probability 20%) or turning left at the Hanamachi 3 intersection (movement probability 80%) Yes, the user is alerted to pay attention to this oncoming vehicle. Here, the collision information notifying unit 108 indicates that “when the duplicate action determining means determines that a plurality of actions can be taken, the risk of collision for each destination predicted by the destination predicting means. And the collision information notification means for notifying the risk level indicating the level of danger when heading to each destination. By providing notification along with the degree of danger, the reliability of the notified collision information can be improved, so that a third party can pay attention and increase the effectiveness of collision avoidance.

[0106] In the present embodiment, the movement history is a force that has always accumulated the node series detected along with the movement. For example, the movement history may be accumulated only when a duplicate action occurs. Good. FIG. 46 is a diagram showing an example of the operation of the moving body when the history is accumulated only when the duplicate action occurs. This will be described below with reference to FIG.

[0107] Fig. 46 is a diagram showing the Raku 4 intersection point shown in Fig. 16 and the like. This Raku 4 intersection is a five-way road that crosses Hanshin 1 street, Kyoto 2 street, Nara 3 street, and back 5 street.For example, even if the user makes the same right / left turn indicator, This is an example of a point that is difficult to pass or not. [0108] For example, suppose that vehicle 1 moves to Hanshin 1st street by taking a left turn winker at Nara 3rd street. The movement history stores the position information and intention information of the vehicle 1 as a movement history. Vehicle ID “001”, intention information “left turn blinker”, and movement history “L216 → N105 → L212 → N104 → L211” are detected and stored as the movement history in which the vehicle 1 force is also detected. On the other hand, suppose that the vehicle 1 also made a left turn winker at Nara 3 streets in the same way, and exited 5 streets on the other side. As the movement history detected from the vehicle 1, the intention information “left turn blinker” and the movement history “L 216 → N105 → L212 → N104 → L215” are detected and accumulated! /.

[0109] In such a complex point, different actions may be taken while showing the same intention information. Furthermore, because of such a point, the speculations of the two often collide differently. Therefore, in the present invention, the history may be accumulated only when duplicate behavior occurs from the history of different behaviors while showing the same intention information from the motion history detected by the mobile body.

FIG. 47 is a block diagram showing a configuration example of a collision information notification system that implements the present modification. In addition to the system components shown in FIG. 38, a duplicate action extraction unit 122 is provided. The duplicate action extraction unit 122 is a means for extracting, from the movement history of each vehicle accumulated in the movement history accumulation unit 117, the same intention information and a history of different movements. Then, the extracted movement history is stored in the movement history storage unit 117. For example, in the case of Fig. 46, vehicle 1 is taking the same left turn winker, and is taking a route different from the direction toward Hanshin 1 route, the route going to Hanshin street, and the route going to the back 5 streets, and this history is extracted. .

[0111] The movement history accumulated at a predetermined interval such as an interval of 1 second may have a huge amount of information. Therefore, a calculation time for searching for the duplicate action from the accumulated movement history is also required. Therefore, by accumulating only when duplicate actions occur in this way, the calculation speed for searching an important history in which the risk of a collision is increased, and the present invention for notifying information with danger in particular! / Demonstrate its effect.

[0112] Further, by accumulating the movement history, not only the points that are pre-stored in the map information as the point information that can be turned left or right or stopped, but also at points that do not include such information. User-specific behavior can be estimated, and collisions can be avoided. Hereinafter, a specific example will be described. FIG. 48 is a diagram for explaining an example of estimating future actions from intention information based on the accumulated movement history. In Fig. 48, it is assumed that truck 1 is traveling on Hanshin 7 street, and now the hazard is lit and the intention to stop is detected. Referring to the map information here, for example, there is information on only the parking lot of store B (N124) as a parking spot around the truck, and the behavior estimation unit 105 estimates the parking lot of store B as the stop location. Assume that the future action “N121 → L225 → N124” is estimated. On the other hand, suppose that you actually entered Shijo Corporation and stopped. In this case, for example, from the rear of the vehicle 2 as in FIG. 22, the truck 1 thinks that it will enter the parking lot of the SB store, trying to overtake the truck 1, and there is a risk of collision.

[0114] The user does not necessarily take the same action as the third person thinks, and thus may take a different action from the action that the third person thinks. In Embodiment 1 described above, the behavior that can be considered from the indicated intention information is estimated from the map information and the like, and the collision is notified when a plurality of behaviors are estimated. Accurate estimation is not always possible. For example, in the case of this example, there is a node that can stop at the parking lot of store B 1S. On the other hand, it is assumed that Shijo Co., Ltd., where truck 1 actually stops, does not have that information. Shijo Co., Ltd. shall not be registered as a node in the general map information that is commonly used for truck 1 for its own company. In this case, it cannot be estimated. Therefore, for example, when a behavior different from the behavior estimated by the behavior estimation unit 105 is taken, it is accumulated as a movement history, and this movement history is used for judging the risk of collision.

For example, the future behavior “N121 → L225 → N124” is estimated in the behavior estimation unit 105 now. It is assumed that truck 1 actually entered Shijo Corporation and stopped. Actions different from those estimated here are accumulated as movement history. In FIG. 48,! /, White circles, and circles are latitude / longitude information detected by the position information detection unit 101, indicating movements that entered Shijo Co., Ltd. and stopped! Then, an action different from the estimated future action is stored as a movement history.

FIG. 49 is a diagram showing a history of actions different from the estimated actions. FIG. 49A is a diagram showing an example of current running. FIG. 49B is a diagram showing intention information issued by the moving object. FIG. 49C is a diagram showing the future behavior estimated by the behavior estimation unit 105. Figure 49D FIG. 10 is a diagram showing an example of current running when an action different from the action estimated by the action estimating unit 105 is taken. FIG. 49E is a diagram showing an example in which the current travel shown in FIG. 49D is accumulated as a movement history. Since the current traveling “N141 → L221” is detected and the intention information “stop” is detected, the behavior estimating unit 105 detects the future behavior “N121 → L225 → N124”. However, the actions that the user actually took were “E (east longitude) 135.10, N (north latitude) 35.20”, “E135.12, N35.20”, “E135.14, N35.20”, “E135 .14, N35.24 "," E135.14, N35.28 "," E135.14, N35.31 "," E135.14, N35.33 " Has been detected. The detected movement is accumulated as a history. Stored as history ID “020”. Then, the risk of collision is judged using this accumulated information.

[0117] For example, in the same situation as in Fig. 48, suppose that truck 1 again intends to stop. Here, the behavior estimation unit 105 estimates the behavior with reference to the movement history as well as the map information. In the map information, the parking lot of store B (N124) is stored as a stoppable point, and “N121 → L225 → N124” is estimated as one of the future actions. Furthermore, the behavior indicated by history ID “020” from the movement history, that is, the behavior of stopping at Shijo Corporation indicated by a white circle in FIG. 48 is estimated as a future behavior. Since there are a plurality of estimated behaviors, the collision risk judgment unit 107 determines that both behaviors are different from each other as a collision region, and notifies them. For example, it is possible to know that the vehicle 2 behind may stop not only at the store B but also at Shijo Co., Ltd., thereby avoiding the danger of a collision.

[0118] Further, as shown in the above embodiment, this example may be not only a notification to a third party but also a notification to itself. In particular, for drivers who usually do the same, the behavior is a daily behavior, so it is assumed that the surrounding third parties may know their behavior. On the other hand, the behavior is different from the generally considered behavior, and the difference often creates a risk of collision. Therefore, as shown in this example, the movement history when the behavior different from the estimated behavior is accumulated, the collision risk is judged using the accumulated movement history, and notification is made. It is possible to provide safe driving support that takes into account the specific behavior of the vehicle. For example, if you take the action to enter Shijo Corporation every day and stop, The effect of avoiding the risk of collision by making it safer to drive by being notified that surrounding third parties will be aware that they will enter store B and that there is a risk of collision. Produce.

[0119] In the above example, the intention information is a force that has been described by taking a turn signal for turning right and left as an example. For example, it is possible to put it in the back gear and use the knock light as its intention information. In particular, as shown in Fig. 48, in a situation where a truck enters a factory or company, there may be a case where the vehicle turns back and parks just after showing a turn signal. Or, a third person often takes actions that cannot always be guessed, such as taking out the right turn signal, swinging it to the right, and then putting it in the back. Therefore, as shown in this example, as the movement history accumulation, for example, the movement history put in the back gear may be accumulated and used. It is an inconvenient force to accumulate the movement history at a predetermined interval, which is inconvenient because the data volume is enormous. Thus, the necessary history can be easily extracted by efficiently accumulating only the history in which the risk of collision occurs It becomes possible.

[0120] Especially in factories, construction sites, transportation companies, convenience stores, etc., each vehicle, especially large vehicles, often takes the same action, as shown in this example. The ability to realize safer driving by determining the risk of collision from actions that cannot be uniquely estimated (actions that cause ambiguity) based on S becomes Kanakura.

[0121] In the present embodiment, the destination of the own vehicle is predicted, and the power that has been determined to be dangerous by calculating the probability of movement is used to predict not only the own vehicle but also the destination of other vehicles. It is also possible to judge the risk.

FIG. 50 is a block diagram showing an example of the configuration of a collision information notification system that predicts the destination of another vehicle and determines the danger. First, in addition to the constituent elements shown in the present embodiment, the moving body 100 is added with the notification partner identification unit 112 and the collision risk degree calculation unit 124 shown in the first embodiment. The notification partner identification unit 112 in the above embodiment identifies, for example, a user in a collision area from the received position information of another moving body, and notifies that there is a danger of collision! /. In this example, the other party to be notified is identified using the predicted destination! On the other hand, the mobile body 200 includes a second position information detection unit 115, a second movement history. The storage unit 121, the second movement destination prediction unit 119, and the second movement destination transmission unit 120 are provided. Similar to the method described above, the second position information detection unit 115 detects the position information of the moving body 200 and stores it in the second movement history storage unit 121 as a movement history. Then, the second movement destination prediction unit 119 predicts the movement destination of the moving body 200 and transmits the movement destination predicted by the second movement destination transmission unit 120. This will be described below with reference to the drawings. In the moving body 100, the position information detection unit 101 is an example of “movement history receiving means for receiving the movement history of the moving body from another moving body”. When the determination means determines that the mobile body can take a plurality of actions with respect to the intention information, the movement history stored in the movement history storage means and the other mobile body The destination identification unit 112 is an example of a “destination prediction unit that predicts a destination of the vehicle and the other mobile object based on the movement history received from the vehicle”. Based on the above, it is an example of “the collision information notifying means for preferentially notifying information on the danger of collision to the other moving body having a high risk of collision with the own vehicle”. FIG. 51 is a diagram for explaining notification control using a destination of another moving object. In Fig. 51, Hana 1 vehicle 1 going straight on Kyoto Street is going to turn left, the probability of turning left at Urahana 1 intersection is 20%, and the probability of turning left at Hanamachi 3 intersection is 80%. The collision area is identified from these different actions. On the other hand, the oncoming vehicle 2 is calculated as a 90% probability of going straight at the Hanamachi 3 intersection as a destination and a 10% probability of turning right. Furthermore, the oncoming vehicle 3 has a 100% probability of turning right at the Hanamachi 3 intersection. Here, for example, in the collision risk calculation unit 124, the possibility of collision with each vehicle based on the product of each movement probability is 8% (10% X 80%) for vehicle 2 and 80% (100% X 80 for vehicle 3). %). Therefore, for example, notification control may be performed using the predicted destination of another vehicle, such as preferentially communicating with the vehicle 3 to avoid a collision. In this case, the collision risk determination unit 107 and the collision information notification unit 108 indicate that “the movement stored in the movement history storage unit when it is determined by the overlapping action determination unit that a plurality of actions can be taken. From the history, for each destination predicted by the destination prediction means, the host vehicle calculates the power, the probability of entering the destination as the risk, and from the movement history received from the other moving body, The probability that the other moving object is heading to the destination is calculated as the risk, and the information on the risk of collision is used as the information on the movement. It is an example of the “collision information notification means” for notifying the risk of the host vehicle and the other moving body heading forward.

[0124] In PtoP communication, in which vehicle-to-vehicle communication is performed one-to-one in inter-vehicle communication, the communication cost becomes a problem. In particular, it is important to specify which vehicle to communicate with at intersections where multiple vehicles are crowded. Therefore, as shown in this example, by identifying the collision area from the overlapping action estimated action, and further determining the risk of crossing from the predicted destination of the other vehicle to this collision area, The ability to avoid collisions while limiting communication costs is S Kanakura.

[0125] It should be noted that the moving body 100 of the present invention provides a means for confirming whether or not information on the danger of collision has been notified to other moving bodies, and notifies other moving bodies. There may be provided means for displaying what has been reported. Here, the “means for confirming whether or not the notification of the information on the danger of the collision has been performed” is “the collision that confirms whether or not the information on the danger of the collision is notified to the other mobile body” "Information notification confirmation means" is an example, and "means for displaying that notification has been given to another mobile body" is "confirmed that the notification has been made by the collision information notification confirmation means" In this case, it is an example of a notification confirmation display means for displaying that the notification has been made, and the collision information notification unit 108 is configured to display a message indicating that the notification has not been confirmed by the collision information notification confirmation means. In this case, it is an example of the “collision information notifying means for notifying the other mobile body of information relating to the risk of the collision again after changing the notification mode”. Specifically, for example, when information on the danger of collision is not given to other moving objects, a notification with a sound or a warning sound may be given. As a result, it is possible to increase the ability to call attention to the danger of collision to other mobile units that have not been notified of the danger of collision, and therefore can be effective. There is.

[0126] In the embodiment described above, for example, the point information accumulated in advance in the map information or the like is estimated as the behavior estimation from the detected intention information such as the blinker. At points where intersections are densely populated or where routes are complexly crossed, the behavior that the user performs while displaying the intention differs from the behavior that the third party recognizes, resulting in a situation where there is a collision. The Map the point information which is the information of the intention information detected there and the action actually performed It accumulates as information and refers to this point information to avoid collisions. However, adding point information to each point can be very costly. Therefore, this point information may be automatically generated using the movement history shown in the present embodiment, and collision may be avoided using the automatically generated point information. Hereinafter, description will be made with reference to the drawings.

FIG. 52 is a block diagram showing an example of the configuration of the server 300 when the point information is automatically generated using the moving history of the moving body and the collision is avoided using the automatically generated point information. This system includes a position information detection unit 101, intention information detection unit 104, behavior estimation unit 105, map information storage unit 103, overlapping behavior determination unit 106, collision risk determination unit 107, collision information notification unit 108, movement history storage unit 117, a duplicate action extraction unit 122, and a point information generation unit 123. The same reference numerals are given to the components shown in the first embodiment. In this server 300, the movement history accumulation unit 117 is an example of “movement history accumulation unit for accumulating the movement history of a moving object”, and the point information generation unit 123 is “stored in the movement history accumulation unit”. For each point on the map, point information indicating the intention information and the movement route selected as the action corresponding to the intention information is generated from the movement history, and the generated point information is stored in the map information storage unit. It is an example of “point information generating means”.

In addition, in the server 300, the point information generating unit 123 reads: “Furthermore, from the movement history accumulated in the movement history accumulating means, intention information indicating a stop for each point on the map, and the intention information Is an example of the point information generating means for generating point information indicating a parking position outside the road selected as an action corresponding to “.

[0129] In this server 300, the point information generating unit 123 reads: "Furthermore, from the movement history accumulated in the movement history accumulation means, for each point on the map, intention information indicating retreat, and the intention information Is an example of the point information generating means for generating point information indicating a parking position outside the road selected as an action corresponding to “.

Furthermore, in the server 300, the position information detection unit 101 is an example of “movement history receiving means for receiving the movement history of the moving object from a plurality of moving objects”. Further, from the received movement history, for each point on the map, the same intention information is selected! / And a different movement route is selected as an action corresponding to the intention information! The point information generating means for adding a moving route as new point information " It is an example.

[0131] In the embodiment and the like, the system is provided in a moving body such as a car navigation system, for example, a force that detects position information of the moving body to avoid a collision, and is not limited to this, for example, a server The position information etc. of each moving body may be detected. The position information detection unit 101 and the intention information detection unit 104 in the server 300 which is the present system are means for detecting the position information and intention information of each mobile object. Then, the position information and intention information detected along with the movement of each moving body are accumulated in the movement history accumulating unit 117.

FIG. 53 is a diagram showing an example of the case where the server 300 installed at a specific point generates point information from the intention information and the movement history of the moving body moving at the point. It is a figure which shows the easy 4 intersection point shown in FIG. This Raku 4 intersection is a five-way road where Hanshin 1 street, Kyoto 2 street, Nara 3 street, and back 5 street cross each other, for example, which route will actually take in the future even if the user makes a right / left turn signal This is an example of a point that is difficult to grasp.

[0133] For example, suppose that vehicle 1 has left turn signal at Nara 3 street and left Hanshin 1 street.

The movement history stores the position information and intention information of the vehicle 1 as a movement history. As the movement history detected from the vehicle 1, the vehicle ID “001”, the intention information “left turn blinker”, and the movement history “L216 → N105 → L212 → N104 → L211” are detected and stored. On the other hand, suppose that vehicle 2 has made a left turn winker in Nara 3 streets and exited 5 streets on the other side. As the movement history detected from the vehicle 2, the vehicle ID “002”, the intention information “left turn blinker”, and the movement history “L216 → N105 → L212 → N104 → L215” are detected and stored.

FIG. 54 is a diagram showing another example when the server 300 installed at a specific point generates the point information from the intention information and the movement history of the moving body that moves at the point. Similarly, Fig. 54 is a diagram showing 4 easy intersections. For example, suppose that vehicle 3 makes a right turn turn signal at Nara 3 Street and passes through Hanshin 1 Street. The movement history stores the position information and intention information of the vehicle 1 as a movement history. As the movement history detected from the vehicle 3, the vehicle ID “003”, the intention information “turn right turn signal”, and the movement history “L216 → N105 → L214” are detected. On the other hand, suppose that vehicle 4 has made a right turn turn signal at Nara 3 street and exited 5 streets on the other side. As the movement history detected from vehicle 2, vehicle ID “004”, intention information “right turn winker”, movement The history “L216 → N105 → L213” is detected.

[0135] In such a complex point, different actions may be taken while showing the same intention information. Furthermore, because of such a point, the speculations of the two often collide differently. Therefore, in the present invention, from the movement histories detected by the plurality of moving bodies, histories with different behaviors while showing the same intention information are extracted to generate point information.

The duplicate action extracting unit 122 is a means for extracting, from the movement history of each vehicle accumulated in the movement history accumulating unit 117, the same intention information and a history of different movements. For example, in the case of Fig. 53, vehicle 1 and vehicle 2 are taking the same left turn blinker, and are taking a route that goes to Hanshin 1 way, a reverse direction to 5 ways, and a different route from the other route. The

[0137] The point information generation unit 123 selects "the intention information and the action corresponding to the intention information for each point on the map from the movement history accumulated in the movement history accumulation means." This is an example of “point information generating means for generating point information indicating the travel route”, and means for generating information from the history extracted by the duplicate action extracting unit 122. For example, the point information includes “intention information” to be detected, “position” in which the intention information is indicated, and “estimated behavior” to be estimated in that case. In the case of FIG. 53, the location information “001” from the extracted history is the position “L216”, the intention information “left turn winker”, the estimated actions “N105 → L212 → N104 → L211”, “N105 → L212 → N104 → L215” "Ca is generated. In other words, if you make a left turn turn signal at Nara 3 Street (L216), information indicating that two routes are assumed as future actions: the route to Hanshin 1 through the Raku 4 intersection and the route to the back 5 Become! /

Similarly, in the case of FIG. 54, from the extracted movement history, the position information “002” is the position “L216”, the intention information “right turn winker”, the estimated actions are “N105 → L214”, and “N105 → L213”. Has been generated. In other words, if you make a right turn turn signal at Nara 3 (L216), there are two possible future actions: a route to turn right to Hanshin 1 through the Raku 4 intersection and a route to Kyoto 2 It becomes information indicating!

[0139] In such a complicated point, different actions may be taken while showing the same intention information, and the actions taken by the third party are often different from the actions recognized by a third party. In addition, at a point where multiple people detect the same intention information in the same way but take different actions. Confusion arises. Therefore, as shown in this example, the movement histories of multiple moving bodies are accumulated, and by generating point information from the accumulated histories, multiple future actions can be accurately estimated, and collisions can be avoided. Is possible.

[0140] Unlike noise information such as turning without turning a blinker by mistake, many such people often take such actions. Therefore, a threshold value may be provided to generate information on points where more users take different actions.

[0141] Further, in the above-described embodiments, the power described for an example of the collision information notification device and the collision information notification method corresponding to each case. They are implemented by freely combining the embodiments as long as they do not contradict each other. Anyway!

[0142] Fig. 55 is a diagram showing an external appearance of a car navigation device provided with the collision information notification device of the present invention. A car navigation device 81 equipped with the collision information notification device of the present invention is attached to the dashboard of the car 80. For example, when the screens shown in FIGS. 13, 19, 36 and 43 are displayed on the display screen of the car navigation device 81, the collision information notification device of the present invention can Information is notified to the user.

Note that each functional block in the block diagrams (FIGS. 4, 20, 21, 31, 31, 32, 38, 47, 50, 52, etc.) is typically an LSI that is an integrated circuit. As realized. These may be individually made into one chip, or may be made into one chip to include some or all of them.

For example, the functional blocks other than the memory may be integrated into one chip.

[0145] Here, it may be called IC, system LSI, super LSI, or mono-LSI, depending on the difference in power integration as LSI.

[0146] Further, the method of circuit integration is not limited to LSI's, and implementation using dedicated circuitry or general purpose processors is also possible. An FPGA (Field Programmable Gate Array) that can be programmed after manufacturing the LSI or a reconfigurable 'processor that can reconfigure the connection and settings of the circuit cells inside the LSI may be used.

[0147] Further, if integrated circuit technology comes out to replace LSI's as a result of the advancement of semiconductor technology or a derivative other technology, it is naturally also possible to carry out function block integration using this technology. There is a possibility of adaptation of nanotechnology. [0148] Further, among the functional blocks, only the means for storing the data to be encoded or decoded may be configured separately instead of being integrated into one chip.

Industrial applicability

[0149] The present invention is provided as a device for providing information related to a collision, for example, in a car navigation device or a portable terminal, and estimates future behavior based on detected position information and intention information. When an action overlaps, it can be used as a collision information notification device that can avoid a collision by generating and notifying information about the collision.

Claims

The scope of the claims

[1] Current position acquisition means for acquiring a current position on a map of a moving object;

Intention information detection means for detecting intention information generated by the mobile body for intention display indicating future behavior of the mobile body;

The movement history of the moving body including the detected intention information, a current position when the intention information is detected, and an action selected for the intention information within a predetermined range from the current position. Movement history storage means for storing;

When the intention information is detected, by determining whether or not a different action has been selected for the same intention information within a predetermined range of the current position force at the time of detection, Thereafter, a duplicate action determination means for determining whether or not different actions can be selected for the same intention information;

A collision information notifying means for notifying information on the danger of a collision with a person or other moving body when the duplicate action determining means determines that a different action can be selected corresponding to the intention information;

A collision information notification device comprising:

[2] The collision information notification device further includes:

Map information representing a map is stored in advance, and the intention information and an action corresponding to the intention information are selected for each point on the map from the movement history stored in the movement history storage unit. A map information storage means for further storing point information indicating the travel route is provided,

The duplicate action determination means refers to the point information stored for points within a predetermined range from the current position acquired when the intention information is detected, thereby detecting the intention The ability to select different travel routes for information,

The collision information notification means notifies the information related to the risk of collision when the duplicate action determination means determines that a different movement route can be selected for the intention information.

The collision information notification device according to claim 1, wherein:

[3] The intention information indicates a right turn, a left turn and a stop! /, An intention of a future action, or the collision information notification device further includes:

A behavior rule accumulating means for accumulating rules concerning the intention information, actions performed in response to the intention information, and a range in which the actions are performed;

The duplicate behavior determination means determines whether or not a plurality of behaviors corresponding to the detected intention information can be performed within the range on the movement route with reference to the stored behavior rules.

The collision information notification device according to claim 1, wherein:

[4] The collision information notification device further includes:

Current position acquisition means for acquiring a current position on a map of the mobile body;

Map information storage means for storing map information,

The duplicate action determining means refers to the map information and the action rule, and the action indicated in the action rule is defined as the action rule, starting from the current position acquired when the intention information is detected. 4. The collision information notification device according to claim 3, wherein it is determined whether or not a plurality of methods can be performed within the range indicated.

[5] The duplicate action determination means starts from the current position acquired when the intention information is detected, and includes a right turn point, a left turn point, and a turn point within a range on the movement route indicated in the action rule. It is determined whether or not there are a plurality of stop points indicated by the intention information, and when there are a plurality of any of the points within a range on the movement route, Judge that the moving body can take multiple actions

The collision information notification device according to claim 4, wherein:

[6] The collision information notification device further includes:

A position information receiving means for receiving the current position of the moving object from another moving object, and the collision information notifying means, when it is determined by the duplicate action determining means that a plurality of actions can be taken; It has a collision risk judgment unit that judges areas where there is a risk of collision with other moving objects,

The collision information notifying means is based on the current position received from another mobile unit, and when there is a possibility that the other mobile unit will exist in the area where there is a risk of collision in the future. Notify about the risk of a crash

6. The collision information notification device according to claim 5.

[7] The collision information notification device further includes:

A collision information notification confirming means for confirming whether or not information related to the risk of collision has been notified to the other moving body;

A notification confirmation display means for displaying that the notification has been made when the collision information notification confirmation means confirms that the notification has been made,

If the collision information notification confirmation unit does not confirm that the notification has been made, the collision information notification unit changes the notification mode and again sends information on the risk of the collision to the other moving body. Notify

The collision information notification device according to claim 6.

[8] The collision information notification means provides information on the risk of the collision to the other moving body that may exist in an area where there is a risk of collision. Notify the vehicle's intention information and the area of collision risk determined by the collision risk determination unit

The collision information notification device according to claim 6.

[9] The collision information notification device further includes:

Point information indicating intention information indicating stopping and a parking position outside the road selected as an action corresponding to the intention information for each point on the map from the movement history stored in the movement history storage means. Point information generating means for generating

The collision information notification device according to claim 1, wherein:

[10] The collision information notification device further includes:

Point information indicating intention information indicating withdrawal and a parking position outside the road selected as an action corresponding to the intention information for each point on the map from the movement history stored in the movement history storage means. Point information generating means for generating

The collision information notification device according to claim 1, wherein:

[11] The collision information notification device further includes:

Movement history receiving means for receiving movement histories of the moving body from a plurality of moving bodies; When a different movement route is selected as an action corresponding to the intention information for the same intention information for each point on the map from the received movement history, the different movement route is changed to a new point. Point information generating means to be added as information

The collision information notification device according to claim 1, wherein:

[12] The collision information notification device further includes:

When it is determined by the duplicate action determination means that the mobile body can take a plurality of actions for the intention information, based on the movement history stored in the movement history storage means, Destination prediction means for predicting

The collision information notification device according to claim 1, further comprising:

[13] The collision information notification means, when it is determined by the duplicate action determination means that a plurality of actions can be taken, for each destination predicted by the destination prediction means, the risk of collision And information on the above and the risk level indicating the degree of danger in case of force and inadvertentness.

13. The collision information notification device according to claim 12,

[14] The collision information notifying means calculates, as the degree of danger, the host vehicle's direction and likelihood of moving to the destination from the movement history stored in the movement history storage means.

The collision information notification device according to claim 13.

[15] The collision information notification device further includes:

When it is determined by the movement history receiving means that receives the movement history of the mobile body from another mobile body, and the duplicate action determination means that the mobile body can take multiple actions with respect to the intention information A destination prediction means for predicting the destination of the vehicle and the other mobile body based on the movement history stored in the movement history storage means and the movement history received from the other mobile body. Prepared,

The collision information notifying unit preferentially notifies the other mobile body of information on the risk of collision based on the prediction by the destination prediction unit.

The collision information notification device according to claim 1, wherein:

[16] The collision information notification means takes a plurality of actions by the duplicate action determination means. If it is determined that it is possible, for each destination predicted by the destination prediction means from the movement history stored in the movement history storage means, the probability that the vehicle will head for the destination is defined as the risk. And calculating the probability that the other mobile body is headed to the destination from the travel history received from the other mobile body as a risk level. Notify the danger level of cars and other moving objects

16. The collision information notification device according to claim 15, wherein

[17] The current position acquisition means for acquiring the current position on the map of the moving object is acquired by the moving object for displaying the intention indicating the future action of the moving object. An intention information detection step for detecting intention information;

The movement history storage means includes the detected intention information, a current position when the intention information is detected, and an action selected for the intention information within a predetermined range from the current position. The movement history accumulating step for accumulating the movement history of the moving body, and when the intention information is detected by the duplicate action determining means, the same intention information is obtained within a predetermined range from the current position at the time of detection. A duplicate action determination step for determining whether or not different actions can be selected for the same intention information by determining whether or not different actions have been selected.

A collision information notifying step for notifying information on the risk of a collision with a person or another mobile object when it is determined in the duplicate action determining step that a different action can be selected corresponding to the intention information;

A collision information notification method comprising:

[18] A program for a collision information notification device,

A current position acquisition step for acquiring a current position on a map of the mobile body, an intention information detection step for detecting intention information generated by the mobile body for intention display indicating a future action of the mobile body, And storing the movement history of the mobile body including the intention information, a current position when the intention information is detected, and an action selected for the intention information within a predetermined range from the current position. The movement history accumulation step and the intention Once the information is detected, it is determined whether or not a different action has been selected for the same intention information within a predetermined range from the current position at the time of detection. And determining whether or not different actions can be selected with respect to the intention information, and determining whether different actions corresponding to the intention information can be selected by the overlapping action determining step and the overlapping action determining step A program that executes a collision information notification step for notifying information on the risk of collision with a person or other moving object.

[19] A processor for executing a program stored in a storage means,

In the storage means,

A current position acquisition step for acquiring a current position on a map of the mobile body, an intention information detection step for detecting intention information generated by the mobile body for intention display indicating a future action of the mobile body, And storing the movement history of the mobile body including the intention information, a current position when the intention information is detected, and an action selected for the intention information within a predetermined range from the current position. When the movement history accumulation step and the intention information are detected, it is determined whether or not different actions have been selected for the same intention information within a predetermined range from the current position at the time of detection. By the determination, it is determined by the overlapping action determination step that determines whether or not different actions can be selected for the same intention information, and the overlapping action determination step. If different behaviors in response to the information is determined to be selected, and collision information notification step of notifying the information about the risk of collision with persons or other mobile

A processor storing a program for executing the above.

[20] GPS that obtains the current position of the moving object on the map,

The movement history of the moving body including the detected intention information, a current position when the intention information is detected, and an action selected for the intention information within a predetermined range from the current position, and A hard disk for storing data including map information; a intention information detecting step for detecting intention information issued by the mobile body for an intention display indicating a future action of the mobile body; and the intention information is detected. Different actions are selected for the same intention information within a predetermined range from the current position at the time of detection. By determining whether or not a different action can be selected for the same intention information by determining whether or not it has been performed, the overlapping action determining step and the overlapping action determining step A program for executing a collision information notifying step for notifying information on the danger of a collision with a person or another moving body when it is determined that a different action can be selected in response to the intention information; A stored memory, and a processor for executing a program stored in the memory! /

A car navigation device comprising: