JP4141225B2 - Patrol route guidance system - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a patrol route guidance system .
[0002]
[Prior art]
As a patrol route guidance system for guiding an optimum patrol route that patrols a plurality of points according to a user's request, the one described in Patent Document 1 is known.
[0003]
In this conventional example, when there is a guidance request from the user, the traffic information providing apparatus in which traffic information is stored in advance is accessed, and the optimum route extraction unit extracts the optimum route.
[0004]
However, in this conventional example, the extraction accuracy in the route extraction unit completely depends on the accuracy of the traffic information providing device, but it is very difficult to obtain accurate traffic information over the entire route to be searched. There is a problem that high extraction accuracy cannot be maintained and operation costs are high.
[0005]
In this regard, for example, the use of commercial traffic information such as VICS ( registered trademark: Vehicle Information & Communication System) provided by the Road Traffic Information Communication System Center as traffic information acquisition means is also described in Patent Document 2. However, since the range covered by these commercial traffic information is limited to the main trunk roads, there is a problem that it is difficult to guarantee the same level of accuracy as a whole.
[0006]
[Patent Document 1]
JP 2002-170197 A [Patent Document 2]
Japanese Patent Application Laid-Open No. 11-223531
[Problems to be solved by the invention]
The present invention has been made in order to eliminate the above drawbacks, and provides a cyclic route guidance system that can extract an optimum route with high extraction accuracy over a whole area with low operation cost. With the goal.
[0008]
[Means for Solving the Problems]
According to the present invention, the object is
A database unit 2 for storing traffic information acquired in the data acquisition unit 1 as link data for connecting nodes set on map data;
A traveling route guidance system having a computing unit 3 that computes an optimum traveling route for circulating between a plurality of nodes from link data in the database unit 2 in response to a user request,
The data acquisition unit 1 includes a commercial data acquisition unit 1a that acquires commercial traffic information;
An operation data acquisition unit 1b that acquires operation data of a vehicle that travels between arbitrary nodes including between lower nodes that are not covered by the commercial traffic information;
The calculation unit 3 performs the Dijkstra method up to the movement costs set in advance for all the search start points and all the search end points in a denser node density network, and thereby the coarser node density reached thereby. The Dijkstra method is performed in a network with a coarser node density, using nodes in a simple network as all new search start points and search end points, and the links are hierarchized into N layers by the density of nodes. This is achieved by providing a cyclic route guidance system that collectively calculates data.
[0009]
In the present invention, link data used when calculating a plurality of traveling routes is created based on both commercial traffic information and operation data from a vehicle that actually travels between arbitrary nodes. .
[0010]
By using operation data from actual vehicles for the creation of link data, it is possible to reliably acquire traffic information between lower-level roads that are not covered by commercial traffic information, for example, national roads or roads with a scale below the prefectural road. As a result, data collection is facilitated and guidance accuracy is improved.
[0011]
Vehicles for which operation data is to be obtained are based on the results of patrol route guidance by this method, such as trucks and other transportation companies that run on the link, transportation facilities such as taxis, and vehicles such as commercial vehicles. Vehicles can be included.
[0012]
As link data, it is possible to use raw data (current data) at the time of data acquisition, but it is possible to use predicted values obtained by performing statistical processing on actual data. In this case, after obtaining the predicted travel speed by statistical processing, it is desirable to use (inter-node distance) / (expected travel speed) = expected required time as link data.
[0013]
In addition, the predicted travel speed can be given as a travel speed prediction function obtained by multivariate analysis based on road characteristic factors for performance data.
[0014]
In this way, by statistically evaluating the influence of road characteristics on travel speed based on the overall link data, there is no actual operation or there are few actual populations, so there is less reliability between nodes. It is possible to ensure the reliability of the link data.
[0015]
As road characteristic factors, for example, in addition to static attributes such as road type, width, number of lanes, regulation speed, presence / absence of traffic lights, regional information such as national census results, commercial statistics, use areas, and whether it is a tourist destination Can be used.
[0016]
In addition, a plurality of link data is prepared for each driving condition characteristic,
When it is configured to use link data prepared for the traveling conditions at the time of departure at each traveling point when calculating the traveling route, it becomes possible to further improve the search accuracy. Since the degree is easily affected by the time zone, it is desirable to include the travel time zone in the running condition characteristics.
[0018]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 shows an embodiment of the present invention configured to search for an optimum route when a predetermined delivery destination such as a warehouse is visited from a predetermined starting point and returned to the original starting point. This system includes a data acquisition unit 1, a database unit 2 including expected movement cost information, and a calculation unit 3.
[0019]
As shown in FIG. 2, the data acquisition unit 1 transmits traffic information (hereinafter referred to as “VICS data” ) from the VICS (registered trademark) 5 operated by the Road Traffic Information Communication System Center via a communication network 4 such as the Internet. ”) And a commercial data acquisition unit 1b that acquires position information (hereinafter,“ operation data ”) transmitted by the car navigation device with a communication unit mounted on the traveling vehicle 6.
[0020]
The above VICS data is distributed on a dedicated line every predetermined time, for example every 5 minutes, corresponding to the link time (VICS link) allocated in advance by the data provider to the travel time in the link, traffic jam / regulation / accident information. On the other hand, the operation data is the position of the vehicle at the time of data acquisition, more precisely, longitude and latitude information, and the data contents are different. Therefore, in the commercial data acquisition unit 1a and the operation data acquisition unit 1b, The acquired data is converted into a predetermined format by the format conversion unit 1 c and then stored in the database unit 2.
[0021]
As shown in FIG. 3, the database unit 2 includes a travel database 7 and a road network database 8. In the road network database 8, the position, node number, and link number of the node set on the map data are displayed. Information related to links set between the first nodes, for example, static attributes of roads such as road type, number of lanes, width, and regulation speed are stored. The nodes of the road network database 8 are associated with the location information of the map database 9, and the regional information of the map database 9, for example, classification of commercial areas, residential areas, etc., classification of tourist areas, national census You can refer to various categories or commercial statistics.
[0022]
The travel database 7 includes a current database 7a and a storage database 7b, and each is associated with a link number of the road network database 8. The data format of the current database 7a is obtained by associating the travel time in the link with the link number, and the VICS data 10 in the present system is shown in FIG. 4A by using the VICS link as a key. The link numbers (L1, L2,...) Of the road network data 12 are searched and stored in the corresponding link in the current database 7a. The VICS data 10 includes traffic jams, regulations, accident information, and the like, which are also stored in the current database 7a together with the date and time of data acquisition.
[0023]
On the other hand, as shown by triangles in FIG. 4 (b), the operation data corresponds to the links (L10, L11...) On the road network data 12 with the transmitted positional information 11a, 11b. The transit time between them is stored as link data. In FIG. 4, nodes are indicated by white circles.
[0024]
As described above, the VICS data 10 and the operation data are registered in the current database 7a at the data acquisition timing, and then stored in the storage database 7b at a predetermined timing. This accumulation database 7b accumulates the time zone, the traveling speed for each day of the week, the traffic situation, the weather, the number of accidents, etc. for each link number.
[0025]
As shown in FIG. 3, the data in the accumulation database 7b is stored in the statistical processing unit 13 by static attributes of roads, that is, road type, link type, number of lanes, width, regulation speed, etc. Multivariate analysis based on geographical information such as altitude is performed, and a traveling speed prediction function is obtained. The travel speed prediction function is obtained for each of three types of time zones of morning, noon, and evening, and the estimated travel time within the link obtained by dividing the link distance by the estimated travel speed obtained from these is stored as the estimated travel cost 14. 7b.
[0026]
In this embodiment, the traveling speed prediction function is shown to be formed for each of three types of time zones, morning, noon, and evening. It is set appropriately.
[0027]
Further, in this embodiment, the estimated intra-link travel time 14 is obtained based on the travel speed prediction function for all links. However, the predicted link travel time 14 is acquired from a link that can acquire the VICS data 10 or at a high frequency. For the estimated travel time 14 of the possible operation data acquisition link, for example, the average value of the storage database 7b is used, and the travel speed prediction function is applied only when the number of operation data acquisition is small or not at all. It is also possible to do.
[0028]
Next, in FIG. 1, reference numeral 15 denotes an input unit, to which data related to a circulation destination such as an address and a search condition such as a tour date and time are input, and an operation for calculating an optimum circulation route based on the search condition from the input unit 15 The unit 3 includes an OD matrix generation unit 3a and a route search calculation unit 3b.
[0029]
As shown in FIG. 5, the OD matrix 3a ′ generated by the OD matrix generation unit 3a has a starting point (O) in the vertical direction and a destination (D) in the horizontal direction, and is the shortest at the vertical and horizontal intersections. The table data to which the movement cost (time) is allocated. In the storage database 7b, the N layer (in this embodiment, the lower layer network having the dense node density and the uppermost layer having the most dense node density) due to the density of the node density. A known algorithm for the shortest path problem such as the Dijkstra search method can be applied to the predicted intra-link travel time 14 that is hierarchized into the network and the intermediate layer of the middle layer of the node density.
[0030]
In general, in a three-layer hierarchy search with N destinations, in order to calculate the shortest cost of one route, 2 × N × (N−1) times (coefficient 2 depends on traffic restrictions, etc.) in the lower layer network and the middle layer network In order to solve one-way traffic with another route), N × (N−1) times of search is necessary in the upper layer network, and in order to obtain a combination of all routes, N × (N−1) times this In this embodiment, the OD matrix 3a ′ is obtained by calculating all points in the lower layer network, the middle layer network, and the upper layer network in a lump.
[0031]
The OD matrix 3a ′ obtained by this method is not an optimal solution for the shortest path problem but is an approximate solution, but has an advantage that the number of searches can be greatly reduced. In addition, the Dijkstra method or the like is performed for all the search start points up to a preset movement cost (search range restriction), and the search range restriction is similarly applied to all search end points from the search end point. To implement the Dijkstra method.
[0032]
In addition, in the calculation for the middle layer network, the search range is restricted from the search start point middle layer node group reached by the calculation in the lower layer network or the search end point middle layer node group for all search start point points and all search end point points. Until then, implement the Dijkstra method.
[0033]
Furthermore, in the calculation for the upper layer network, the search starting point upper layer node group reached by the calculation in the middle layer network to the search end point upper layer node group reached by the calculation in the middle layer network for all the search starting point points. Implement the Dijkstra method.
[0034]
In this way, after obtaining all routes of all layers, the route of the layer having the lowest travel cost among all layers is extracted as the shortest travel cost for each route.
[0035]
When the OD matrix 3a ′ is obtained by the algorithm as described above, the number of searches in the lower and middle layers is N × 2 (times), and the number of searches in the upper layer is N (times). The number of searches can be greatly reduced as compared with the case of obtaining an optimal solution.
[0036]
On the other hand, the route search calculation unit 3b applies the traveling salesman problem to the OD matrix 3a ′ generated by the OD matrix generation unit 3a, and obtains the optimum traveling route. As a solution to the traveling salesman problem, a known nearest neighbor method or a two-optimal method is adopted.
[0037]
In the above, an example has been given in which a time-first cyclic route search is performed by associating the travel time with the estimated travel cost 14, but a distance-priority search is performed by associating the distance with the travel cost. These priority modes can be selected when the search condition is input at the input unit 15.
[0038]
The route guidance from the output unit 16 that outputs the calculation result can include, for example, rearrangement of the delivery list in the order of circulation, delivery destination arrival time, distance, and map information regarding the route. Is notified to the user using a communication network such as the Internet.
[0039]
【The invention's effect】
As is clear from the above description, according to the present invention, it is possible to extract the optimum route with high extraction accuracy over the entire region.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an overall configuration of the present invention.
FIG. 2 is a block diagram of a data acquisition unit.
FIG. 3 is a block diagram of a database unit.
FIGS. 4A and 4B are explanatory diagrams showing data conversion, in which FIG. 4A shows a method for converting VICS data, and FIG. 4B shows a method for converting operation data;
FIG. 5 is a diagram showing an OD matrix.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 Data acquisition part 1a Commercial data acquisition part 1b Operation data acquisition part 1c Format conversion part 2 Database part 3 Calculation part

Claims (6)

A database unit for storing the traffic information acquired in the data acquisition unit as link data for connecting the nodes set on the map data;
A traveling route guidance system having a computing unit that computes an optimum traveling route that circulates between a plurality of nodes from link data in the database unit in response to a user request,
The data acquisition unit includes a commercial data acquisition unit that acquires commercial traffic information;
An operation data acquisition unit that acquires operation data of a vehicle that travels between any nodes including lower nodes that are not covered by the commercial traffic information;
The arithmetic unit performs the Dijkstra method up to the movement costs set in advance for all the search start points and all the search end points in a denser node density network, and the coarser node density reached thereby. The Dijkstra method is performed sequentially in a network with a coarser node density using all nodes in the network as new search start points and search end points, and link data hierarchized into N layers by node density coarse and dense. Traveling route guidance system that collectively calculates The calculation unit includes an OD matrix generation unit that generates an OD matrix including table data in which a starting point is arranged in the vertical direction, a destination is arranged in the horizontal direction, and the shortest movement cost is allocated to the vertical and horizontal intersections. The traveling route guidance system according to claim 1, wherein an OD matrix hierarchized into N layers by density density is generated. The traveling route guidance system according to claim 1, wherein the data acquisition unit includes a format conversion unit that converts data acquired by the data acquisition unit into a predetermined data format. The traveling route guidance system according to claim 1, 2 or 3, wherein time is used as the travel cost and transit time between nodes is used as the link data. The said route data are the cyclic | annular route guidance systems of Claim 4 derived | led-out from the estimated traveling speed obtained by performing a statistical process on performance data. 6. The traveling route guidance system according to claim 5, wherein the predicted traveling speed is given as a traveling speed prediction function obtained by multivariate analysis based on a road characteristic factor with respect to performance data.

Images