WO2012056526A1

WO2012056526A1 - Navigation device

Info

Publication number: WO2012056526A1
Application number: PCT/JP2010/069059
Authority: WO
Inventors: 沙耶香吉津; 直樹井原; 好紀横山
Original assignee: トヨタ自動車株式会社
Priority date: 2010-10-27
Filing date: 2010-10-27
Publication date: 2012-05-03

Abstract

Provided is a navigation device (1) which predicts an upcoming route of passage from a destination and/or a present location. The navigation device (1) comprises: a storage means (21) for storing a vehicle's past travel route history; a travel route acquisition means (10) for acquiring a route which the vehicle has traveled; and a comparison means (26) for comparing the past travel route history of a prescribed zone which is stored in the storage means (21) with the travel route which is acquired with the travel route acquisition means (10). With said configuration, it is possible to predict a destination or a route of passage with high precision by demarcating the past travel route history in the prescribed zone and comparing in the prescribed zone on a per route basis.

Description

Navigation device

The present invention relates to a navigation apparatus for predicting a passing route ahead from a destination or / and a current position.

Even when destinations and routes are not set, technologies for predicting destinations and passing routes have been developed for use in various applications. For example, Patent Document 1 discloses a technique for obtaining the probability of a route that will proceed from the past movement information history and the current movement route and predicting the destination of the vehicle. In this way, by predicting the destination and route, it is possible to update the data in advance by restricting the range around the destination and route from the enormous amount of data in automatic update of map data and content data. It is possible to provide the user with only the traffic jam information related to the destination and the route through the traffic jam information received from the infrastructure.

JP 2004-45413 A JP 2010-145115 A

In the prediction method described in Patent Document 1 above, nodes are used as past and present information, and processing is performed in units of nodes. Therefore, when there are a plurality of different paths between the nodes, it is impossible to determine which path is normally used by the user, and thus the prediction accuracy is lowered.

Therefore, an object of the present invention is to provide a navigation device that predicts a destination and a passage route with high accuracy.

A navigation device according to the present invention is a navigation device that predicts a passage route ahead from a destination or / and a current position, and stores a storage means for storing a past travel route history of the vehicle and a route on which the vehicle has traveled. And a comparison unit that compares the past travel route history of the predetermined section stored in the storage unit with the travel route acquired by the travel route acquisition unit.

This navigation device is provided with a storage means, and the storage means stores a past travel route history on which the vehicle has traveled. In the navigation device, when the route traveled by the vehicle is acquired by the travel route acquisition means, the travel route of the predetermined section of the past travel route history stored in the storage means by the comparison means and the travel route acquired by the travel route acquisition means And compare. This comparison is performed when learning the route on which the vehicle normally travels or when predicting the destination or passing route of the vehicle using the learned route. The predetermined section is a section in which the vehicle has traveled in the past (particularly, a section having a high frequency of travel is preferable). For example, a point that is the basis of life (such as a home) and a frequent destination (such as a workplace) ). In this way, the navigation device can predict the destination and the passing route with high accuracy by dividing the past travel route history into predetermined sections and comparing the predetermined sections in units of routes.

In the navigation device of the present invention, it is preferable that the travel route is a link array. In this way, in the navigation device, by setting the travel route as a link arrangement, the travel routes can be compared in units of links, and learning and prediction can be performed by the route by the link itself.

In the above navigation device of the present invention, the comparing means compares the link array constituting the travel route history of the predetermined section stored in the storage means with the link array constituting the travel route acquired by the travel route acquiring means. It is preferable to calculate the similarity between the travel routes and determine that the travel routes with high similarity are the same travel route. In this way, the navigation device compares the travel routes composed of link arrays and obtains the similarity to the travel route of the past predetermined section for the travel route acquired this time, so that the current travel route is the past predetermined section. Whether or not the route is the same as the travel route can be determined with high accuracy.

In the above navigation device of the present invention, the past travel route history stored in the storage means is a link array including the travel frequency of each link for each predetermined section, and the comparison means is similar based on the travel frequency of each link. It is preferable to calculate the degree. As described above, in the navigation device, the similarity can be obtained with high accuracy by obtaining the similarity with respect to the travel route of the past predetermined section for the current travel route using the travel frequency of each link in the link array. .

According to the present invention, it is possible to predict the destination and the passing route with high accuracy by dividing the past travel route history into predetermined sections and comparing the predetermined sections in units of routes.

It is a block diagram of the navigation apparatus concerning this Embodiment. It is an example of the same path | route determination which concerns on this Embodiment. It is a flowchart which shows the flow of a process in the navigation apparatus which concerns on this Embodiment.

Hereinafter, an embodiment of a navigation device according to the present invention will be described with reference to the drawings. In addition, the same code | symbol is attached | subjected about the element which is the same or it corresponds in each figure, and the overlapping description is abbreviate | omitted.

In the present embodiment, the present invention is applied to a navigation device configured in a center capable of wireless communication with a vehicle. The center according to the present embodiment is a center that provides various services to vehicles, for example, a center operated by an automobile company. The center according to the present embodiment collects travel data for each vehicle to be serviced, learns a route along which the vehicle travels using the collected travel data, and uses the learned route data to It has a function as a navigation device that predicts a passing route and provides information on the destination and the passing route to the vehicle. In the present embodiment, the navigation device in this center will be described in detail.

The center has a wireless communication function with each vehicle and is equipped with a wireless communication device. This communication device is a communication device for performing wireless communication with each vehicle by a wireless communication method such as a cellular phone. Each vehicle that receives service from the center has a wireless communication function with the center, and includes a wireless communication device. This communication device is a communication device for performing wireless communication with the center using the same wireless communication method as the center, for example, DCM [DateＤＣＭCommunication Module].

In the navigation device of each vehicle, map data (for example, link data, node data), GPS signals received by a GPS [Global Positioning System] receiver, sensor values detected by various sensors (for example, traveling direction, vehicle speed (unit time) Based on the winning travel distance)), the current position is detected by map matching by the hybrid method. In this current position detection, the road on which the vehicle is traveling is specified by map matching, and the link ID of the road is acquired. In each vehicle, every time the current position is detected, the link ID of the road that is running and the time stamp when the link ID is detected are transmitted to the center. In each vehicle, the time stamp when the ACC [ACCessory] switch is turned on by the driver and the position information (for example, node ID) of the place (boarding point) are transmitted to the center, and the ACC switch is turned off. The time stamp of the time and the location information of the place (alighting point) are transmitted to the center. In addition, when a vehicle automatically updates map data, etc., requests for predictions of destinations and future routes are requested according to requests from various applications, such as when restricting the data update area or specifying the area where traffic information is provided. Is sent to the center.

The navigation apparatus 1 according to the present embodiment will be described with reference to FIG. 1 and FIG. FIG. 1 is a configuration diagram of the navigation device according to the present embodiment. FIG. 2 is an example of the same route determination according to the present embodiment.

The navigation device 1 learns the route of the vehicle OD [OriginestDestination] unit (1 trip) for each service-provided vehicle, and predicts the passing route ahead from the destination and current position of the vehicle. In particular, in the navigation device 1, in the learning process / prediction process, data including a link array including the traveling frequency of each link in OD units is applied as route data, and the learned route data using the traveling frequency of each link is used. The similarity with the traveling data to be compared is calculated, and when the similarity is high, it is determined that the route is the same.

For this purpose, the navigation device 1 includes a vehicle travel data collection device 10 and a computer 20 (storage device 21, same point determination 22, life point determination 23, primary destination determination 24, trip determination 25, same route determination 26, destination / Route prediction 27). In the present embodiment, the vehicle travel data collection device 10 corresponds to the travel route acquisition unit described in the claims, the storage device 21 corresponds to the storage unit described in the claims, and the same route determination 26 is performed. This corresponds to the comparing means described in the claims.

The vehicle travel data collection device 10 is a device that collects data necessary for the learning process / prediction process. The vehicle travel data collection device 10 collects data by receiving a link ID of a road on which the vehicle is traveling and a time stamp when the link passes, for each vehicle, by the wireless communication function of the center. Further, the vehicle travel data collection device 10 collects data by receiving the time stamp when the ACC switch is turned on and off and the location information of the location for each vehicle by the center wireless communication function. Each time the data is collected, the vehicle travel data collection device 10 stores the travel history or the most recent trip for each vehicle in the storage device 21 of the computer 20.

The computer 20 includes a storage device 21 for storing various data used in the learning process / prediction process. Further, the computer 20 executes an application program for learning processing / prediction processing to thereby execute the same point determination 22, living base point determination 23, primary destination determination 24, trip determination 25, same route determination 26, destination / route. Prediction 27 is performed. In the learning process, the same point determination 22, the living base point determination 23, the primary destination determination 24, the trip determination 25, and the same route determination 26 are performed, and the same point determination 22, the living base point determination 23, and the primary destination determination 24 are performed. The point table is generated, and the route table is generated by the trip determination 25 and the same route determination 26. In the prediction process, the same route determination 26 and the destination / route prediction 27 are performed, and the destination / route prediction is performed by the same route determination 26 and the destination / route prediction 27.

The storage device 21 will be described. The storage device 21 is a storage device for storing a travel history, a point table, a route table, and the latest trip for each vehicle. The travel history is a history of vehicle travel data collected by the vehicle travel data collection device 10. This travel data includes a link ID, a time stamp at the time of passing each link, ACC ON / OFF position information (for example, node ID) for each travel (for example, a section from ACC ON to OFF). That time stamp. In particular, the most recent trip is travel data for the most recent trip. The point table is a table that stores data on stay points determined by the same point determination 22 in the learning process. This stay point data includes stay point ID, position information (for example, node ID), type (for example, life base point, primary destination, secondary destination), visit frequency, stay time for each visit (OFF of ACC) To the ON time). The route table is a table of data of routes determined as the same route having the same OD (the same departure point and arrival point) determined in the same route determination 26 in the learning process. This route data is position information (for example, node ID) of the departure point and the arrival point, and a link array from the departure point to the arrival point. This link array is composed of a link ID and the frequency of traveling each link. In addition, in order to cope with the noise etc. by GPS, this link arrangement | sequence may also contain the information of the frequency of not passing a link when there is no repetition of a link. The link array may also include information on the frequency of a boarding / alighting point (secondary destination) between the departure point and the arrival point.

The same point determination 22 will be described. In the same point determination 22, each time the position information of the point where the ACC is turned off is collected by the vehicle travel data collection device 10, each stay point of the collected position information and the point table stored in the storage device 21. Are sequentially compared with each other, and it is determined whether or not it is the same point as the stay point registered in the point table. In this determination, an error in position detection using GPS is taken into account, for example, whether or not the position of the ACC OFF point is within a predetermined range (for example, within 50 m or 100 m) centered on the position of the staying point. It is determined whether or not they are at the same point. When it is determined that a certain staying point is the same point, in the

same point determination

22, 1 is added to the visit frequency of the staying point determined to be the same, and the staying point registered in the point table of the storage device 21 Update the data. When it is determined that there is no same stay point in the point table, in the same point determination 22, a new stay point ID is assigned to the OFF point of the ACC and 1 is set as the visit frequency. It is registered as data of a new staying point in the point table. Note that the determination may be performed using the position information of the boarding point where the ACC is turned on instead of using the position information of the getting-off point where the ACC is turned off.

The life base point determination 23 will be described. In the life base point determination 23, it is determined whether or not each stay point registered in the point table of the storage device 21 is a point (for example, a home) serving as a base of life. In this determination, for example, the number of visits at a staying point is high (for example, the number of visits per week is a predetermined number or more) and the total time of staying at the staying point is long (for example, the staying time per week) Judgment is made based on whether or not a predetermined time or more Places that become the foundation of life such as homes are visited every day, and the visit time per visit also becomes longer. When it is determined that the staying point is a point serving as a base of life, in the living base point determination 23, the type of the staying point determined as the point serving as a base of life is set as the base point of life and registered in the point table of the storage device 21. Update the data for that stay.

The primary destination determination 24 will be described. In the primary destination determination 24, it is determined whether or not each staying point other than the living base point registered in the point table of the storage device 21 is a point (for example, a workplace) serving as the primary destination. In this determination, for example, the determination is made based on whether or not the frequency of visits to the staying point is high (for example, the number of visits per week is equal to or greater than a predetermined number of times (the number of times that is less than the number of criteria for determining the living base point)). The primary destination is a place where the number of visits is less than the base point of life, but is visited frequently, such as at work. When it is determined that the staying point is a point to be the primary destination, the primary destination determination 24 sets the type of the staying point determined as the primary destination as the primary destination and registers it in the point table of the storage device 21. Update the data for that staying point.

It should be noted that a living destination registered in the point table and a visiting point other than the primary destination are secondary destinations. The secondary destination is a place to visit occasionally. For the secondary destination, the type of the stay point is set as the secondary destination, and the data of the stay point registered in the point table of the storage device 21 is updated.

The trip determination 25 will be described. In trip determination 25, from the travel history of the storage device 21, the section between the living base point and the primary destination is taken as one trip and is broken down into OD units (units where the living base point and the primary destination are the starting point and the arriving point). To do. Here, the position information of the departure point (a point where ACC is ON), the position information of the arrival point (a point where ACC is OFF) and the type of each stay point in the point table in the travel history of the storage device 21 are shown. Using this, the travel data in which the departure point is the life base point and the arrival point is the primary destination is extracted, and the travel data in which the departure point is the primary destination and the arrival point is the life base point is extracted. . One trip is composed of one traveling data when no one stops on the way, but one trip is composed of plural traveling data when one stops on the way. In trip determination 25, the travel data divided into OD units is stored in the storage device 21.

The same route determination 26 will be described. The same route determination 26 is performed in both the learning process and the prediction process. In the same route determination 26, the comparison target travel data is selected from the same OD (the same departure point and arrival point) registered in the route table of the storage device 21 or the route data of each trip having the same departure point. Extract the same route (high similarity). In particular, in the case of learning processing, the traveling data to be compared is traveling data for one trip that has been traveled with the same OD, and in the case of prediction processing, the traveling data to be compared has the same starting point. This is travel data of trips from the starting point during travel to the current position. The route data is a link array (link ID and frequency of traveling the link in the past) from the departure point to the arrival point (or the current position). The travel data to be compared is a link array (link ID only) from the departure point to the arrival point (or the current position).

In this determination, first, route data having the same departure point and arrival point (or departure point) registered in the route table based on the departure point and arrival point (or departure point) of the travel data to be compared are obtained. Extract sequentially. Then, the alignment of the link array is executed on the travel data to be compared and the route data extracted from the route table using dynamic programming. Using the frequency of each link ID of the link array that is representative of the route data after alignment as a weight, calculate the score of the representative link array of the route data after alignment and the link sequence of the travel data to be compared, from the score Calculate similarity. Furthermore, it is determined whether the similarity exceeds a threshold value. If the similarity exceeds the threshold, it is determined that the route is the same. If the similarity is less than the threshold, it is determined that the routes are not the same.

In the case of the learning process, if it is determined that the route is the same, the frequency of each link of the route data determined to be the same is rewritten based on the link arrangement of the comparison target travel data and registered in the route table of the storage device 21. Update its route data. On the other hand, if it is determined that there is no identical route in the route table, the frequency of all link IDs included in the link array of the comparison target travel data is set to 1, and the link array (link ID and the link traveling) is set. Frequency) is registered in the route table of the storage device 21 as new route data. Note that the route data registered in the route table is statistical information including routes (link arrangement) such as noise caused by GPS and getting-on / off points along the route.

In the case of prediction processing, since the comparison travel data is a link array from the departure point to the current position, when comparing with the route data, the link array to the current position is compared, and the link array to the current position is compared. Score and similarity are calculated. Therefore, strictly speaking, it is not determined whether or not the route data is the same, but whether or not the route data has a similar portion.

Referring to FIG. 2, a specific example of the same route determination method during the learning process will be described. Here, “P0” is the departure point, “P2” is the arrival point, “L0”, “L1”, etc. indicated as symbols are link IDs, and “100” is indicated as frequency. , “99” or the like is the frequency of traveling on each link, and “T1” is the boarding / exiting point (secondary destination) where the trip stopped.

Route data L ₁₀₀ is an example of route data from the start point P0 extracted from the routing table to the arrival point P2, past running 100 times for the path of the route data. The route data L ₁₀₀ starts from the departure point P0, passes through the link L0 100 times in 100 times (frequency: 100), and then passes through the link L1 99 times in 100 times (frequency: 99). Passes through the link L8 only once (frequency: 1) (passing this link L8 is considered to be noise detected by GPS (position misplacement detection)), and then passes through the link L2 100 times out of 100 times. (Frequency: 100) Next, the link L3 is passed 97 times out of 100 times (Frequency: 97), but gets on and off at the point T1 only 3 times (Frequency: 3), and then the link L4 is moved 100 times. Pass 100 times (frequency: 100), then pass the link L5 100 times in 100 times (frequency: 100), and then 98 times out of 100 times (frequency: 98), but 100 times Link L4 and link L5 again only twice Pass (frequency: 2) (repeating this link L4 and link L5 is considered to be noise detected by GPS (repetitive misdetection of position)), and then the link L6 passes 100 times out of 100 (frequency: 100) And the data reaching the arrival point P2.

The alignment with the path data _{L 100,} representative of the link sequence _{L 100} 'is, links L0, the link L1, the link L2, a link L3, a link L4, link L5, non-passed (without repetition of the link L4), non-passed (repeat No link L5) and a link array consisting of links L6. In order to compare with this representative link array L ₁₀₀ ′, the link array D of travel data is linked to the link data L 0, link L 1, link L 9, link L 3, link L 4, link L 5, un- This is a link array consisting of passing, non-passing, and link L6.

When the representative link array L ₁₀₀ ′ of the route data is compared with the link array D of the travel data, the link array D of the travel data passes through the link L0, passes through the link L1, and does not pass through the link L2, but passes through the link L9. Passes, passes through link L3, passes through link L4, does not pass through link L4 and link 5, and passes through link L6. When calculating the score based on this comparison result, link L0 has passed 100 times in the past 100 times, so it is set to 1.00 (100%), and link L1 has been passed 99 times in the past 100 times. Therefore, it is set to 0.99 (99%), link L9 has not passed once in 100 times, so 0.00 (0%), and link L3 has passed 97 times in 100 times in the past. Therefore, it was set to 0.97 (97%), the link L4 has passed 100 times in the past 100 times, so it is set to 1.00 (100%), and the link L5 has been passed 100 times in the past 100 times. Therefore, it is set to 1.00 (100%), and the non-passing with respect to the repeated link L4 is set to 0.98 (98%) because it has not passed 98 times out of 100 times in the past. As for non-passing, since it has not passed 98 times out of 100 in the past, it is set to 0.98 (98%), and for link L6 it has passed 100 times in the past 100 times, so it is set to 1.00 (100%). The score is 7.92, which is the sum of all values. When calculating the similarity from this score, dividing the score of 7.92 by 9 to calculate the average value gives 0.88 (88%), and 88% is the similarity. For example, when the similarity determination threshold is 80%, the similarity is 88%, and therefore, the link array L ₁₀₀ of the route data and the link array D of the travel data are determined to be the same route.

If the link sequence D of the travel data and link sequence L ₁₀₀ routes data is determined as the same route, route data link array L ₁₀₀ of the route data is rewritten on the basis of the link sequence D of the travel data, rewriting The route data L ₁₀₁ thus registered is registered in the route table. The new route data L ₁₀₁ departs from the departure point P0, passes through the link L0 101 times out of 101 times (frequency: 101), and then passes through the link L1 100 times out of 101 times (frequency: 100). However, it passes through the link L8 only once (frequency: 1), then passes through the link L2 100 times out of 101 times (frequency: 100), but passes the link L9 only once this time ( Frequency: 1) Next, the link L3 passes 98 times out of 101 times (frequency: 98), but gets on and off at the point T1 only 3 times (frequency: 3). Next, the link L4 is 101 times out of 101 times. Passes 101 times out of 101 times (frequency: 101) and then 99 times out of 101 times (frequency: 99) but 2 out of 101 times Pass through link L4 and link L5 again (frequency: 2), Next, the data passes through the link L6 101 times out of 101 times (frequency: 101) and reaches the arrival point P2.

The destination / route prediction 27 will be described. In the destination / route prediction 27, from the route data of each trip having the same departure point registered in the route table of the storage device 21 by the same route determination 26 to the current position from the departure point in the middle of the comparison target. If you extract route data that has a similar part to the trip data of the trip, the route data that is likely to drive after the current position, the stop point (entrance / exit point) along the route, and the purpose from the extracted route data The ground (arrival point) is extracted and output as a prediction result.

Referring to FIG. 1, the operation of the navigation device 1 will be described along the flowchart of FIG. FIG. 3 is a flowchart showing the flow of processing in the navigation device according to the present embodiment.

In each vehicle, when the driver turns on the ACC switch, data related to ACC ON (time stamp, position information of the boarding point) is transmitted to the center. Further, in each vehicle, when the driver turns off the ACC switch, data related to ACC OFF (time stamp, position information of the unloading point) is transmitted to the center. While traveling, each vehicle detects the current position using map data, GPS, etc. at regular intervals, and transmits data (link ID, time stamp) relating to the current position to the center.

The vehicle travel data collection device 10 of the center navigation device 1 receives (collects) the vehicle travel data from each vehicle (S1), and travels the collected vehicle travel data to a storage area for each vehicle in the storage device 21. Stored as history (last trip) (S2). The navigation device 1 determines whether or not there is a route prediction request from each vehicle (S3). For each vehicle, if a destination or route is not set on the vehicle side, a route prediction request is sent to the center when a future passage route or destination is required for updating map data, providing traffic information, etc. ing.

When it is determined in S3 that there is no route prediction request, in the same point determination 22 of the center, when the position information of the getting-off point where the ACC is turned off is collected for each vehicle, the collected position information and storage device The position information of each staying point in the point table stored in 21 is sequentially compared, and it is determined whether or not it is the same point as the staying point registered in the point table (S4). When it is determined that a certain stay point is the same point, in the same point determination 22, the visit frequency of the stay point is updated, and the data of the stay point is registered in the point table of the storage device 21 (S4). When it is determined that there is no same point in the point table, in the same point determination 22, a new stay point ID is assigned to the getting-off point and is registered in the point table of the storage device 21 as new stay point data. (S4).

In the living base point determination 23, it is determined whether or not the staying point registered in the point table of the storage device 21 is a living point (S5). This determination is performed, for example, when data on a staying point is newly registered in the point table. When it is determined as a living point, in the living base point determination 23, the type of the stay point determined as the living base point is set as the living base point, and the data of the staying point is registered in the point table of the storage device 21 (S5).

In the primary destination determination 24, it is determined whether or not the staying point other than the living base point registered in the point table of the storage device 21 is the primary destination (S6). This determination is performed, for example, when the staying point is determined not to be a living base point in the determination of the living base point determination 23. If it is determined as the primary destination, the primary destination determination 24 sets the type of the stay point determined as the primary destination as the primary destination, and registers the data of the stay point in the point table of the storage device 21. (S6).

In trip determination 25, based on the travel history of the storage device 21, it is determined whether or not vehicle travel data for one trip (OD unit) has been newly prepared (S7). If it is determined in S7 that the vehicle travel data for one trip is not yet complete, the process returns to S1 and the collection of the vehicle travel data is continued.

When it is determined in S7 that the vehicle travel data for one trip has been prepared, in the same route determination 26, the same OD registered in the route table of the storage device 21 (the same as the current vehicle travel data for one trip). It is determined whether or not there is trip route data similar to the current vehicle travel data for one trip among route data of all trips having a departure point and an arrival point (S8). If it is determined in S8 that there is no similar trip route data in the route table, the same route determination 26 sets a frequency 1 for each link ID of the current vehicle travel data for one trip, The new trip route data is registered in the route table of the storage device 21 (S9). On the other hand, if it is determined in S8 that there is route data of a similar trip in the route table, the same route determination 26 updates the frequency of each link ID of the route data of the similar trip, respectively. The route data is registered in the route table of the storage device 21 (S10).

When it is determined in S3 that there is a route prediction request, in the same route determination 26, vehicle travel data from the departure point to the current position is acquired from the storage area of the latest trip of the storage device 21 for the vehicle that has transmitted the route prediction request. (S11) Then, route data including a portion similar to the vehicle travel data up to the current position is extracted from the route data of all trips having the same departure point registered in the route table of the storage device 21 (S11). S12). Then, in the destination / route prediction 27, a route that is likely to travel after the current position, a stop point (entrance / exit point) in the middle of the route, a destination (arrival point), and the like are extracted from the extracted route data. The result is output as a prediction result (S13). The center transmits this prediction result to the vehicle that has transmitted the route prediction request.

According to this navigation device 1, by comparing the OD unit route data stored in the route table with the vehicle travel data collected from the vehicle in the link unit, the OD unit route data traveled by the vehicle is highly accurate. The destination and the passage route can be predicted with high accuracy using the learning data. Since learning processing and prediction processing are performed in units of links, even when there are a plurality of routes between nodes, it is possible to accurately determine the route through which the vehicle actually passes.

By predicting such destinations and passage routes, it is possible to update data in advance by restricting the range around the destination or passage route from a huge amount of data in automatic update of map data etc. From the traffic jam information received from the infrastructure, only traffic jam information related to the destination and route can be provided to the user, and when the user searches for facilities, etc., not only around the current position but also around the destination and route You can search for facilities and provide them to users.

Moreover, in the navigation apparatus 1, by using a link array composed of the link ID passed in the OD section and the frequency of traveling each link as the route data used for learning processing and prediction processing, the routes can be compared in units of links. The similarity can be obtained with high accuracy using the frequency of. Further, by including data such as link ID replacement due to GPS noise and link ID repetition in this link array, learning and prediction can be performed with high accuracy in consideration of GPS noise and the like. Since the route data includes data including the frequency of each link, the resistance to such noise becomes stronger by continuing learning for a long period of time and accumulating the frequency of many travelings.

Further, in the navigation device 1, the route data for performing the learning process and the prediction process is set as an OD unit (life base point and primary destination), so that the section between the life base point and the primary destination is a vehicle. Since the driver frequently travels, a lot of data can be accumulated as the vehicle travel data, and the more data can be accumulated, the more accurate learning can be performed.

Further, by configuring the navigation device 1 in the center, it is not necessary to perform learning processing and prediction processing in each vehicle, so that the processing load on the computer and the storage capacity of the storage device can be reduced.

As mentioned above, although embodiment which concerns on this invention was described, this invention is implemented in various forms, without being limited to the said embodiment.

For example, in the present embodiment, the learning process and the prediction process are performed by the navigation device configured in the center capable of wireless communication with the vehicle, but the navigation device mounted on the vehicle, the portable terminal capable of wireless communication with the vehicle, or the like The learning process and the prediction process may be performed. Moreover, although it was set as the structure which performs both a learning process and a prediction process, it is good also as a structure which performs only one process.

In this embodiment, the learning process and the prediction process are switched according to the presence / absence of a route prediction request. However, the timing for performing the learning process and the prediction process may be other forms. For example, the prediction process May be performed online, and the learning process may be performed offline.

In the present embodiment, the frequency of traveling each link included in the route data is accumulated, the score is calculated using the frequency, and the similarity is obtained from the score to determine whether the route is the same route. Other methods may be used for the same route determination (similarity determination) in units of links.

Further, in the present embodiment, the route data for performing the learning process and the prediction process is an OD unit (life base point and primary destination), but another route may be used as the route data.

DESCRIPTION OF SYMBOLS 1 ... Navigation apparatus, 10 ... Vehicle driving | running | working data collection apparatus, 20 ... Computer, 21 ... Memory | storage device, 22 ... Same point determination, 23 ... Living origin determination, 24 ... Primary destination determination, 25 ... Trip determination, 26 ... Same Route determination, 27 ... Destination / route prediction.

Claims

A navigation device for predicting a passing route ahead from a destination or / and a current position,
Storage means for storing a past travel route history of the vehicle;
Travel route acquisition means for acquiring a route traveled by the vehicle;
Comparison means for comparing the past travel route history of the predetermined section stored in the storage means with the travel route acquired by the travel route acquisition means;
A navigation device comprising:
The navigation device according to claim 1, wherein the travel route is a link array.
The comparison means compares the link array constituting the travel route history of the predetermined section stored in the storage means with the link array constituting the travel route acquired by the travel route acquisition means, and The navigation apparatus according to claim 2, wherein a similarity is calculated, and a travel route having a high similarity is determined as the same travel route.
The past travel route history stored in the storage means is a link array including the travel frequency of each link for each predetermined section,
The navigation device according to claim 3, wherein the comparison unit calculates a similarity based on a running frequency of each link.