JP3860392B2 - Route search device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a route search device , and more particularly to a route search device that searches for a route to a destination by sequentially extending search branches from a search start point.
[0002]
[Prior art]
The in-vehicle navigation device searches for a route with the lowest cost from the departure point to the destination point, for example, a route that follows the shortest distance, and displays the route on the screen to guide the driver (route guidance function). . According to this route guidance function, during actual driving, the guidance route is displayed in a specific color so as to be distinguishable from other roads so that the driver can easily reach the destination. . The Dijkstra method is known as a method for retrieving a route on which a vehicle travels from road map data. The Dijkstra method searches for the shortest route from the departure point to the destination in consideration of all intersections existing in a region having a radius of a straight line connecting the departure point and the destination, or a region larger than the region.
[0003]
FIG. 7 is a schematic explanatory diagram of the Dijkstra method. The road is a straight line and the intersection is a straight line intersection. The distance between the intersections is known, Ps is the starting point (vehicle position), and Pd is the target. It is the earth. In the Dijkstra method, primary intersections A _{1 to} A ₄ adjacent to the departure point Ps are obtained, and zeroth-order intersections (departure points) and distances from the departure points are stored in correspondence with the primary intersections A _{1 to} A _4. To do. Next, a secondary intersection Bij is obtained for each of the primary intersections A _{1 to} A ₄ , and a distance from the primary intersection and the departure point is obtained in correspondence with each secondary intersection. For example, for the primary intersection A ₂ , three secondary intersections B ₁₁ , B ₁₂ , B ₁₃ are obtained, and corresponding to each secondary intersection B ₁₁ , B ₁₂ , B ₁₃ ,
B ₁₁ : Distance from primary intersection A ₂ and departure point d ₁₁ B ₁₂ : Distance from primary intersection A ₂ and departure point d ₁₂ B ₁₃ : Distance from primary intersection A ₂ and departure point d ₁₃ .. (a)
Remember.
[0004]
In addition, _three secondary intersections B ₂₁ , B ₂₂ , B ₂₃ are obtained for the primary intersection A ₃ , and corresponding to each secondary intersection B ₂₁ , B ₂₂ , B ₂₃ ,
B ₂₁ : Distance from primary intersection A ₃ and departure point d ₂₁ .. (b)
B ₂₂ : Distance from primary intersection A ₃ and departure point d ₂₂ B ₂₃ : Stores primary intersection A ₃ and distance d ₂₃ from departure point, and similarly for other primary intersections A ₁ and A ₄ A secondary intersection is obtained and predetermined data is stored. By the way, the intersections B ₁₃ and B ₂₁ are the same intersection. In this way, when the intersection where data should be stored overlaps, the distances d ₁₃ and d ₂₁ from the starting point are compared, and only the smaller data is stored. For example, if d ₁₃ > d ₂₁ , the data of (b) is finally stored as the data of the intersection B ₁₃ (= B ₂₁ ).
Thereafter, similarly, a tertiary intersection Cij is obtained for each secondary intersection, and the distance from the secondary intersection and the starting point is obtained and stored in correspondence with each tertiary intersection. +1) If the next intersection is obtained, and the accumulated distance from the i-th intersection and the starting point is obtained and stored in correspondence with each (i + 1) -th intersection, the destination point Pd is finally reached. . The search is performed by Dijkstra method from both the starting point and the destination point, the search is stopped when the search paths from both sides are linked to each other, and both search paths are connected to form a path from the starting point to the destination point. You can also.
[0005]
The above is a case where a route search is performed using map data having no hierarchical structure by the Dijkstra method, but a route search can also be performed using map data having a hierarchical structure. Referring to FIG. 8, first, using (1) the lower intersection netlist, the optimum first route PT1 connecting the starting point Ps to the upper starting point Ps 'and the destination point Pd to the upper end point Pd'. The optimal second route PT2 connecting the upper layer starting point Ps 'and the upper layer end point Pd' using the upper layer intersection netlist (2) is then obtained by the Dijkstra method. (3) These first route → third route → second route are combined in order to obtain an optimum route from the starting point to the destination point.
[0006]
[Problems to be solved by the invention]
When a route search is performed using a hierarchical road network as described above, first, an upper layer level where search branches extending from both points Ps and Pd are connected is determined, and an intersection from the lowest layer including both points to an upper layer is determined. The search branch is extended to the top, then the search branch is extended from both sides in the upper layer, and the arrangement of links from the contacted search branch to both points is set as the optimum route. In such a route search, (1) in the search from the lowest layer to the upper layer, it is important to search for an effective upper layer by extending the search branch over a wide range with a small amount of calculation, and (2) In the search in the upper layer, the important point is to contact the search branches from both points with a small amount of calculation.
However, when the search branch is extended by the conventional Dijkstra method, due to the nature of the algorithm, the area closer to the search start point is calculated more densely and the further away it is, the more sparsely calculated, the search range starts the search It gradually spreads slowly in a gradation around the point. That is, there is a problem that the above point cannot be solved by the route search by the conventional pure Dijkstra method.
Accordingly, an object of the present invention is to provide a route search apparatus that can reduce the amount of calculation required for route search and shorten the search time.
[0007]
[Means for Solving the Problems]
In the present invention, (1) A is the straight line distance from the search start point to the target point or its neighboring points, D is the cumulative distance from the search start point to the adjacent intersection through the target intersection, and the search start point to the adjacent point When the straight line distance to the intersection is B, the cost C to the adjacent intersection is expressed by the following formula C = D + k (A−B)
(K is a coefficient), (2) Save the route with the lowest cost among the routes to the adjacent intersection and the cost corresponding to the adjacent intersection, and similarly correspond to the other intersection adjacent to the target intersection Save the route and cost. (3) After that, use each saved neighboring intersection as the target intersection and extend the search branch to search the route to the destination point.
[0008]
DETAILED DESCRIPTION OF THE INVENTION
(A) Principle of the Present Invention FIG. 1 is an explanatory diagram of the principle of the present invention. A is the straight line distance from the search start point Ps to the target point Pd or its neighbors, D is the cumulative distance from the search start point Ps to the adjacent intersection Pn ₁ through the target intersection Pi, and the search start point Ps to the adjacent point. Assuming that the straight line distance to the intersection Pn ₁ is B, the cost C to the adjacent intersection Pn ₁ is expressed by the following formula C = D + k (AB)
(K is a coefficient). Next, the route with the lowest cost among the routes up to the adjacent intersection Pn ₁ and the cost are stored in correspondence with the adjacent intersection Pn ₁ , and the route and cost are similarly stored in correspondence with the other adjacent intersections Pn ₂ ,. save. Thereafter, the stored adjacent intersections Pn ₁ , Pn ₂ ,... Are used as intersections of interest, the search branch is extended, and a route to the destination point is searched.
By doing so, k (AB) decreases as the intersection of interest moves away from the search start point and approaches the circumference of the search range circle, so that the search branch extends in the circumferential direction. Search speed is improved.
[0009]
(B) Configuration of Navigation Device FIG. 2 is a configuration diagram of a navigation device for realizing the present invention, and 11 is a DVD (map data storage unit) for storing a road map. The map data includes (1) a road layer composed of a node table NDTB, a road list RDLT, an intersection configuration node list CRLT, and an intersection netlist CRNL, and (2) a background layer for displaying objects on the map, (3) Consists of character layers for displaying municipality names and national highway names. The road layer included in the map data has a data structure shown in FIG. Of these, the road list RDLT gives attribute information of the corresponding road link. The total number of nodes on the link, each node constituting the link, and the type of road (national road, highway, width, etc.) ) And the like. The intersection node configuration list CRLT shows, for each intersection on the map, a set of other end nodes (referred to as intersection configuration nodes) of each link connected to the intersection. The node table NDTB is a list of all the nodes on the map. For each node, (1) location information (longitude, latitude), and (2) whether the node is an intersection. Intersection identification flag, (3) If an intersection, indicates intersection data, if not an intersection, a pointer indicating a road link to which the node belongs, (4) Map number / unit code of the map including the intersection It is composed of the intersection sequential number in the unit. The information of (4) becomes a pointer to the intersection netlist CRNL.
[0010]
As shown in Figure 4, the intersection netlist CRNL
(1) Map number of the map containing the intersection
(2) Unit code
(3) Intersection ID greater than or equal to the intersection sequential number in the unit
(4) Number of intersection nodes
(5) Upper-layer node presence flag / upper-layer intersection sequence number,
(6) Sequential number of each adjacent intersection
(7) Distance to each adjacent intersection
(8) Road attributes to each adjacent intersection (road type, width)
Etc., and is used in the route search process. The intersection netlist has an upper layer and a lower layer.
[0011]
As shown in FIG. 5, the upper-layer node presence flag (5) in the intersection netlist CRNL indicates that the road RD where the intersection CP _B1 exists in the lower layer is also represented by road data in the upper layer, and the intersection CP _B1 is the upper layer. in "1" is raised when that is registered as the intersection CP _a, otherwise (if the CP _B2, CP _B3 in FIG. 5) to be "0". The upper-layer intersection sequential number indicates the intersection sequential number in the upper layer when an intersection netlist is created in the upper layer for the same intersection as the intersection in the lower layer.
[0012]
Returning to FIG. 2, 12 is a display device (CRT) that draws a map, a vehicle position mark, a guidance route, etc. according to the current location of the vehicle, and 13 calculates the current location based on the travel distance and direction of the traveling vehicle. This is a current location detection unit. Although not shown, the current location detection unit 13 is an orientation sensor that detects the traveling direction of the vehicle, a vehicle speed sensor that detects the travel distance, and a CPU for position calculation that calculates the current location (longitude, latitude) of the vehicle based on the orientation and travel distance. have.
14 is an operation unit such as a remote controller having a map search key, a map scroll key, a route search key, a menu key, and the like, and 15 is a navigation control device, which generates a map image around the current vehicle location based on map data, A guide route is searched, and a guide route image and a car position mark are generated.
In the navigation control device 15, 15 a reads out map data (for example, map data for 9 screens) in a range wider than the screen display range around the current location from the DVD 11 based on the current location data, and generates a dot image based on the map data. It is a map image drawing part which generates a map image.
[0013]
15b is a guide route drawing unit for generating a guide route image based on guide route data from the starting point to the destination obtained by the guide route search process, and 15c is a video RAM for storing the map image and the guide route image. The map image drawing unit 15a rewrites the video RAM 15c at any time according to the driving of the vehicle so that the display range of the display screen does not exceed the image range of the video RAM 15c, and the guide route drawing unit 15b also responds to the driving of the vehicle. Thus, a guide route image is generated and stored in the video RAM 15c.
[0014]
A reading control unit 15d reads a map image for one screen from the video RAM 15c so that the current vehicle position is located at a fixed point on the display screen. The reading position is instructed from the map image drawing unit 15a. 15e is a vehicle position mark generating unit for displaying a vehicle position mark at a fixed point on the display screen, and 15f is a route search unit. When a departure point and a destination are input during a route search, map data is displayed. Based on the included road layer information, the shortest route from the departure point to the destination is calculated as the guidance route. A guidance route storage unit 15g sequentially stores nodes constituting a guidance route from the departure point to the destination. Reference numeral 15h denotes a synthesizing unit, which appropriately synthesizes the map image, the guide route image, the car position mark image, and other menu images read from the video RAM 15c and the car position mark generating part 15e, respectively, and outputs them to the CRT 12. indicate.
[0015]
(C) Route Search Processing FIG. 6 is a flow of route search processing according to the present invention. The present invention is a modified version of the search procedure of the Dijkstra method, and focuses on widening the search branch over a wide range by weakening (abandoning) the function for obtaining a reliable shortest path of the Dijkstra method. A circle centering on the search start point (search range circle) is set in advance (step 100). However, the radius A of the search range circle is a linear distance connecting the search start point Ps (see FIG. 1) and the target point Pd or its neighboring points. The destination point is, for example, an intersection that connects the lower layer and the upper layer.
[0016]
Next, the route search unit 15f sets the search order i to 0 (0 → i, step 101), and checks whether there is an intersection adjacent to the i-th intersection with reference to the lower-level intersection netlist CRNL (FIG. 4). (Step 102). The zeroth intersection is the starting point Ps.
If there is an intersection adjacent to the i-th intersection Pi, the cumulative distance D from the starting point Ps to the adjacent intersection Pn ₁ via the i-th intersection Pi is calculated (step 103). As will be described later, the distance d ₁ from the starting point to the i-th intersection is stored in the storage unit 15f-1 in correspondence with the i-th intersection, and from the i-th intersection Pi to the adjacent intersection Pn ₁ since the distance d ₂ are stored in the intersection network list, the following equation d ₁ + d ₂ → D
Thus, the cumulative distance D from the starting point Ps to the adjacent intersection Pn ₁ is obtained.
[0017]
Next, a straight line distance B from the starting point Ps to the adjacent intersection Pn ₁ is calculated (step 104), and the following formula C = D + k (AB)
The cost C is calculated based on (k is a coefficient) (step 105). k is a coefficient for adjusting the influence (effect) on the cost of (AB), and the effect of (AB) can be influenced by the magnitude of k, and the calculated quality of the search route and the search time are calculated. Can be adjusted.
Next, it is checked whether the search order of the adjacent intersection Pn ₁ is (i + 1) (step 106). If it is not (i + 1), the following data is associated with the adjacent intersection.
(1) Sequential number of the i-th intersection that is currently focused on,
(2) Total cost C from the starting point to the adjacent intersection,
(3) Cumulative distance D from the starting point to the adjacent intersection,
(4) (i + 1) as the search order of the adjacent intersection
Is stored in the storage unit 15f-1 (step 107), and thereafter, the process returns to step 102 to check whether or not the intersection adjacent to the focused i-th intersection still remains and the subsequent processing is repeated.
[0018]
On the other hand, if the degree of the adjacent intersection is (i + 1) in step 106, in other words, if the adjacent intersection is referred to as an intersection adjacent to another i-th intersection (step 107 has already If the data of (1) to (4) is stored), the total cost C ′ from the starting point stored corresponding to the adjacent intersection is compared with the cost C obtained in step 105. (Step 108).
If C <C ′, the sequential number of the i-th intersection that is currently focused on the sequential number of the i-th intersection stored in the storage unit 15f-1 corresponding to the adjacent intersection Pn ₁ is stored. In addition to the replacement with the sequential number, the cost C ′ and the cumulative distance D ′ are rewritten with C and D (C → C ′, D → D ′, step 109). Thereafter, the process returns to step 102 to focus on the i-th order. It is checked whether an intersection adjacent to the intersection still remains, and the subsequent processing is repeated. If C ≧ C ′, the process returns to step 102 without changing the contents stored in the storage unit 15f-1 corresponding to the adjacent intersection.
[0019]
In step 102, if there is no intersection adjacent to the i-th intersection of interest, it is checked whether another i-th intersection exists (step 110). If it exists, the other i-th intersection is determined. As a new i-th intersection (step 111), the processing after step 102 is repeated.
In step 110, if another i-th intersection does not exist, it is checked whether the destination point Pd has been reached, in other words, whether the destination point is included in the (i + 1) -th intersection (step 112) If it is not included, the search order is incremented by 1 (i + 1 → i, step 113), and the processing after step 102 is repeated. However, when the destination point is reached, (1) the destination point Pd → the destination point (assumed to be the m-th order intersection) and (m−1) the next intersection point → 3) the (m− 1) Stored in correspondence with the next intersection (m-2) Next intersection →... → (4) Primary intersection stored in correspondence with the secondary intersection → (5) The 1 The zeroth-order intersections (departure points) stored in correspondence with the next intersection are sequentially connected to determine a route with a low cost (step 114), and the route processing is terminated.
[0020]
According to the present invention, k (AB) decreases as the intersection of interest moves away from the search start point and approaches the circumference of the search range circle, so that the search branch extends in the circumferential direction. Increases speed.
The present invention has been described with reference to the embodiments. However, the present invention can be variously modified in accordance with the gist of the present invention described in the claims, and the present invention does not exclude these.
[0021]
【The invention's effect】
As described above, according to the present invention, since k (AB) is included in the cost, k (AB) becomes smaller as the intersection of interest moves away from the search start point and approaches the circumference of the search range circle. The search branches are greatly extended in the circumferential direction, the amount of calculation required for the route search is reduced, and the route search time is shortened.
[Brief description of the drawings]
FIG. 1 is a diagram illustrating the principle of route search according to the present invention.
FIG. 2 is a configuration diagram of a navigation device for realizing the present invention.
FIG. 3 is an explanatory diagram of a road layer.
FIG. 4 is an explanatory diagram of an intersection netlist.
FIG. 5 is an explanatory diagram of an upper layer node presence flag and upper layer road identification data.
FIG. 6 shows route search processing according to the present invention.
FIG. 7 is an explanatory diagram of the Dijkstra method.
FIG. 8 is an explanatory diagram of a route search method using map data having a hierarchical structure.
[Explanation of symbols]
Ps ··· Search start point Pd · · Destination point Pi · · Intersection of interest Pn ₁ , Pn ₂ · · Adjacent intersection

Claims (1)

In a route search device that searches for a route from a search start point to a target point by sequentially extending search branches,
A storage unit that stores a route with the lowest cost among the routes from the target intersection to the adjacent intersection, and stores the cost in association with the adjacent intersection;
A is the linear distance from the search start point to the target point or its neighboring points, A is the cumulative distance from the search start point to the adjacent intersection through the target intersection, and B is the linear distance from the search start point to the adjacent intersection. , The cost C to the adjacent intersection is expressed by the following equation: C = D + k (AB) (k is a coefficient)
And the route with the lowest cost among the routes to the adjacent intersection and the cost are stored in the storage unit in association with the adjacent intersection, and similarly the route in correspondence with another intersection adjacent to the target intersection And a cost search unit that saves the cost in the storage unit, and searches for a route to the destination point by extending the search branch using each of the stored adjacent intersections as a target intersection.
A route search device comprising:

Images