JP4421052B2 - Map data management method, route search device, and recording medium - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a patrol route selection method, and more particularly, to a method for efficiently patroling a plurality of points and selecting a patrol route from a departure point to a destination.
[0002]
[Prior art]
As is well known, social interest in traffic control technology and ITS-related technology has increased in recent years. Due to such social interests, a route search method for automatically selecting a practical route to a destination has been actively researched and developed, and has already been applied in various fields. As an application example thereof, there is an in-vehicle car navigation system or a center side device of a traffic information providing system such as ATIS. In these methods, the optimum route from one point (departure point) to another point (destination) is obtained using the technique disclosed in Japanese Patent Laid-Open No. 4-204881.
[0003]
In addition, for application to a delivery management system or the like, there is an increasing demand for a patrol route search technique for obtaining an optimum route that passes through all of a plurality of selected points. The traveling route search technique is mathematically called the “traveling salesman problem” and has been known as a difficult calculation problem with a large amount of calculation since ancient times. A method for obtaining a complete solution of the traveling salesman problem has not been established, and only a method for obtaining a practical approximate solution exists. Methods for obtaining an approximate solution are disclosed in "JP-A-3-191464" and "JP-A-4-57147".
Japanese Patent Laid-Open No. 3-191464 discloses a method in which an initial route is configured in the order of angles from the center of gravity of each point, and optimization is repeated within a certain number of points. Japanese Patent Application Laid-Open No. 4-57147 discloses a method in which an initial route is configured by sequentially adding points with the smallest increase cost, and the points are rearranged for optimization.
[0004]
First, an outline of a conventional cyclic route search method (hereinafter referred to as a first cyclic route search method) disclosed in “JP-A-3-191464” will be described. In the first patrol route search method, first, an angle formed between the center of gravity of each selected point and each point is obtained, and then an initial route is configured based on each obtained angle. Next, the configured initial route is divided for each fixed number of points. Next, a “depth-first search method” is applied to each divided area, and the points included in the area are rearranged so as to form a shorter route. Accordingly, the first cyclic route search method obtains the shortest cost route at each selected point. Furthermore, after the points included in each of the divided areas are exchanged, the above-described series of processing is repeated. Eventually, when the cost does not decrease, the processing of the first cyclic route search method ends. Details of the first cyclic route search method will be described below with reference to the flowchart of FIG. In particular, FIG. 37A is a main flowchart showing a processing procedure of the first cyclic route search method.
[0005]
In FIG. 37A, after the center coordinates of the set point group are determined (step S37a1), the points are connected in the order of angles connecting the center coordinates and the point coordinates (step S37a2). An initial route in the cyclic order is constructed. Next, an optimization loop process is performed based on the initial path (step S37a3), and a correction process to approximate the minimum cost path is performed. Here, FIG. 37B is a sub-flowchart showing a detailed processing procedure of step S37a3 or S37a6 of FIG.
In FIG. 37 (b), a start point for division is determined (step S37b1), and the set point group is set in an area for each fixed number of points based on the determined start point while following the order of the initial route. Divided. Next, for each divided area, a minimum cost route is obtained by a brute force method (step S37b2). Next, the divided areas are combined to form a one-round route (circular route) (step S37b3). Finally, if the cyclic route cost obtained in step S37b3 is smaller than the previous one (step S37b4), it is determined that the cyclic route cost may be smaller, and the process returns to step S37b1. Next time, the processing of steps S37b1 to S37b4 is repeated based on the cyclic route cost obtained in step S37b3.
On the other hand, in step S37b4, if the cyclic route cost obtained this time is not smaller than the previous one, it is determined that the cyclic route has almost the lowest cost. Then, the optimization loop process (step S37a3) is terminated assuming that the previous round route is optimal.
[0006]
Next, returning to FIG. 37 (a), after each point and each side are exchanged, m groups are switched in order of efficiency (step S37a4). Next, the first (m = 1) reconnection is performed (step S37a5). Furthermore, the same optimization loop process as step S37a3 is performed (step S37a6). Next, it is determined whether or not the cost of the cyclic route obtained in step S37a6 is the smallest so far (step S37a7). If the cost is the smallest, the cyclic route cost may be further reduced, and in step S37a6. The processes in steps S37a4 to S37a7 are repeated using the obtained traveling route cost as a reference.
On the other hand, if the cost of the cyclic route obtained in step S37a6 is not the smallest in step S37a7, step S37a8 is performed. Next, it is determined whether the cost of the cyclic route has been confirmed for all m groups (step S37a8). If the cost has been confirmed, the first cyclic route search method ends.
On the other hand, if all the m sets have not been confirmed in step S37a8, the cyclic route selected this time (that is, the cyclic route after reconnection has not been performed) does not have the minimum cost. It is returned to the one before the replacement (step S37a9). Further, after the next combination (m = m + 1) is switched (step S37a10), step S37a6 is executed again, and the processing of S37a6 to S37a7 is repeated.
[0007]
Next, an outline of a conventional cyclic route search method (hereinafter referred to as a second cyclic route search method) disclosed in Japanese Patent Laid-Open No. 4-57147 will be described. In the second cyclic route search method, first, an initial route is configured by sequentially adding points that form a low-cost triangle from among points to be visited. Thereafter, the process of reconfiguring the cyclic route by separating the points at random is repeated, whereby a cyclic route having the minimum cost is searched. Details of the second cyclic route search method will be described below with reference to the flowchart of FIG.
In FIG. 38, when the point data is input, a matrix in which the distance between the input points is recorded is created (step S3801). Next, after three points are selected from the input point group, the first triangle is created using the selected three points (step S3802). Next, one point constituting the optimum triangle is selected from the unselected point group, and a process of adding the selected point to the circulation route is performed. This process is repeated until there are no unselected point groups, thereby forming a circular route (step S3803). Next, the total distance of the configured cyclic route is calculated, and it is determined whether or not the total distance calculated this time is the shortest of the cyclic routes obtained so far. If the total distance is the shortest, the current tour route is adopted as the optimum one (step S3804). Next, it is determined whether or not the processing is completed based on the number of processing times and time (step S3805). If the predetermined number of processing times / predetermined time has been exceeded, the process ends. If the predetermined number of processing times / predetermined time has not been exceeded, step S3806 is performed. After the point is randomly separated from the configured tour route and the point is returned to the unselected state (step S3806), the process returns to step S3803 and the subsequent processes are repeated.
[0008]
In addition to the above methods, various methods using genetic algorithms and neural networks have been reported at academic societies and the like.
[0009]
[Problems to be solved by the invention]
However, the method using the latter genetic algorithm or neural network may fall into a minimal solution, and there is a problem in terms of path quality in the worst case.
Further, in the first cyclic route search method, it is determined whether to continue the correction of the cyclic route based on whether the cost of the cyclic route is the minimum. Therefore, there is a possibility that the cost of the circular route may fall into a minimum solution, and an optimal circular route may not be obtained. Furthermore, there is a possibility that the processing time until the cost of the cyclic route finally converges becomes long.
[0010]
Furthermore, the second route search method limits the process of randomly separating and reconfiguring the points by limiting the number of times, time, etc., so that the route cost (distance) is ultimately optimal. There is a problem that it is not constant whether it converges.
[0011]
Therefore, an object of the present invention is to provide a cyclic route selection method capable of selecting a more practical cyclic route in practical time.
[0012]
[Means for Solving the Problems and Effects of the Invention]
A first invention is a method for efficiently patroling a plurality of points and selecting a patrol route from the starting point to the destination, wherein the first point of selecting the starting point, the destination and the plurality of patrol points is selected. Obtained in the step, the second step of obtaining an evaluation value between two necessary points among the two points selectable from the starting point, the destination and the respective traveling points, and the second step. In addition, using each evaluation value, the third step of determining the order of visiting the plurality of tour points based on the depth-first search method, and the tour order determined in the third step, A fourth step of connecting a starting point, the plurality of patrol points, and the destination to form one patrol route from the starting point to the destination, and the third step includes: Survey order to be investigated A fifth step of calculating a total evaluation value until the middle (hereinafter referred to as a first total evaluation value) based on each evaluation value obtained in the second step, and a current investigation target When a certain cyclic order is configured, a total evaluation value of the configured cyclic order (hereinafter referred to as a second total evaluation value) is calculated based on each evaluation value obtained in the second step. And a seventh step of recording a minimum value of the second total evaluation value calculated in the sixth step, and the first sum calculated in the fifth step. If the evaluation value exceeds the minimum value of the second total evaluation value recorded in the seventh step, Subsequent investigation processing derived from the above-mentioned patrol route Will be terminated.
[0013]
According to the first invention, in the process of determining the order of patrol, during the cyclic route investigation process based on the depth-first search method, the first total evaluation value is the minimum value of the second total evaluation value ( That is, when the second total evaluation value, which has been the minimum in the past, is exceeded, the subsequent investigation process derived from the cyclic route is terminated. As a result, it is possible to omit the investigation regarding the combination of the tour points that cannot be the optimum tour route, and the selection time of the optimum tour route can be shortened.
[0014]
Book The invention is a method for efficiently patroling a plurality of points and selecting a patrol route from the starting point to the destination, wherein the first step of selecting the starting point, the destination and the plurality of patrol points;
Of the two points that can be selected from the starting point, the destination, and each patrol point, use the second step to obtain the necessary evaluation value between the two points and the evaluation values obtained in the second step. In accordance with the third step of determining the order in which the plurality of tour points are to be visited and the tour order determined in the third step, the departure place, the plurality of visit points and the destination are joined together, A fourth step of determining a single round trip route to the destination, and the third step is management information in which the round trip points are connected in a chain in order to manage a plurality of round trip points. And a sixth step of constructing a cyclic order based on the management information created in the fifth step.
[0015]
Book According to the invention, in the process of determining the order of patrol, management information in which unselected patrol points to be investigated are connected in a chain is applied. As a result, since an unselected patrol point can be efficiently selected, it is possible to shorten the selection time of the minimum cost patrol route.
[0016]
Book The invention is a method for efficiently patroling a plurality of points and selecting a patrol route from the starting point to the destination, wherein the first step of selecting the starting point, the destination and the plurality of patrol points;
Of the two points that can be selected from the starting point, the destination, and each patrol point, use the second step to obtain the necessary evaluation value between the two points and the evaluation values obtained in the second step. In accordance with the third step of determining the order in which the plurality of tour points are to be visited and the tour order determined in the third step, the departure place, the plurality of visit points and the destination are joined together, And a fourth step that constitutes one patrol route to the destination. The second step is a fifth step of reading map data in a range including at least a starting point or a destination and each patrol point. Step 6 is set in the sixth step by using the map data read in the fifth step and the sixth step for sequentially setting the starting point or destination and each patrol point as the search start point. Search start point The path between the point other than the search start point, and derived by Method a basic Dijkstra's algorithm, including a seventh step of calculating the respective evaluation values.
[0017]
Book According to the invention, in the process of obtaining the evaluation value between the points, the points are sequentially selected as the search start points, and the unidirectional search process based on the Dijkstra method is performed. As a result, the search process between each two points included in the m points can be reduced from m × (m−1) times to m times, so that it is possible to reduce the time required for selecting the optimum route. .
[0018]
Book The present invention is a method for efficiently patroling a plurality of selected patrol points and selecting a patrol route from a departure point to a destination, and candidate points that can be selected as the plural patrol points are determined in advance. In addition, a route between the candidate points is obtained, and route information including at least an evaluation value between the candidate points is recorded. From the starting point and the destination, and a plurality of candidate points. A first step of selecting a plurality of patrol points, a second step of acquiring an evaluation value between two necessary points among two points selectable from the starting point, the destination, and each patrol point; Using the evaluation values obtained in step 2, a third step for determining the order in which a plurality of patrol points are to be visited, and a departure place and a plurality of trips in accordance with the order in which the third step is determined. Point and destination And a fourth step that forms one circuit route from the starting point to the destination, and the second step is included in the plurality of circuit points selected in the first step. Each evaluation value between each two points is obtained by reading from pre-recorded route information.
[0019]
Book According to the invention, it is not necessary to calculate a route between each traveling point by determining a point that can be selected as a traveling point in advance, and obtaining a route between each point and recording the route information. It is possible to shorten the selection time of the optimum patrol route.
[0020]
Of the present invention The second step is a fifth step of reading map data in a range including the starting point and / or the destination and each traveling point when the starting point and / or the destination are not included in the candidate points. The sixth step for sequentially setting the starting point and the destination selected in the first step as the search start point and the map data read in the fifth step are set in the sixth step. A seventh step of obtaining each evaluation value by deriving a route between the search start point and each traveling point by a method based on the Dijkstra method.
[0021]
Book According to the invention, a search process based on the Dijkstra method is performed using an arbitrary starting point or destination as a search start point. As a result, the route from the departure point to each round trip point, or the route from each round trip point to the destination can be obtained in one search process, so the optimum round trip route from any departure point, or the optimum to any destination It is possible to shorten the time required for selecting a round route.
[0022]
Book invention Is Points on a plurality of roads that access the area including the candidate point are determined in advance as entry / exit points, and a route between the entry / exit point and each candidate point is obtained in advance, and at least the entry / exit point The route information including the evaluation value of the route between the point and each candidate point is recorded in advance, and the second step is that if the starting point or destination is outside the area including the candidate point, Select a temporary departure point or temporary destination from the list.
[0023]
Book According to the invention, a point on the road that can access the patrol area is determined in advance as an entrance / exit point. Further, routes between each entrance / exit point and each traveling point (candidate point) and between each traveling point (candidate point) and each entrance / exit point are also obtained in advance. These are recorded in advance as route information. Then, by selecting the entrance point to the patrol area as the temporary departure point, or the exit point from the patrol area as the temporary destination, the patrol route from the temporary departure point or the patrol route to the temporary destination Can be requested. Thus, for example, in an environment with a small amount of RAM such as an in-vehicle device, even if a route from the starting point to each patrol point or a route from each patrol point to the destination is not obtained, the selected point is patroled. It is possible to obtain a traveling route to be performed.
[0024]
Book invention Is As the entrance / exit point, a point on the main road that crosses the boundary of the area including the candidate point is selected in advance, and the second step is to use the entrance / exit point as the search start point or the search end point, and to set the departure point or destination. A temporary departure place or a temporary destination is selected by performing a bidirectional hierarchical search process on the place.
Book According to the invention, the minimum necessary entrance / exit point can be mechanically set by selecting a point on the main road that crosses the boundary of the circuit area as the entrance / exit point. As a result, the selection of entry / exit points can be simplified, and an appropriate temporary departure point or temporary destination can be automatically selected using the search process.
[0025]
Book invention Is As the entrance / exit points, points on the main road crossing the boundary of the area including the candidate point and points on the more main road passing through the area including the candidate point are selected in advance. In the second step, the temporary departure point is obtained by performing one-way hierarchical search processing with the entry / exit point as the search end point and the departure point or destination selected in the first step as the search start point. Or select a temporary destination.
[0026]
Book According to the invention, as the entrance / exit point, a point on the main road that is not included in the circuit area but passes through the periphery is selected. This makes it possible to select an entry / exit point that arrives at the lowest cost by a one-way hierarchical search method from the departure point or destination, so that the route selection process can be simplified and the selection time can be shortened.
[0027]
Book invention of The second step selects a temporary departure point and / or a temporary destination from the entry / exit points based on the user's selection.
[0028]
Book According to the invention, the temporary departure point or temporary destination is selected by the user from the entrance / exit points. As a result, the user can enter or exit the patrol area from any direction, so that the user can select an approach road to the patrol area or an escape road from the patrol area according to the user's preference.
[0029]
Book The present invention is a method for efficiently patroling a plurality of selected patrol points and selecting a patrol route from a departure point to a destination, and candidate points that can be selected as the plural patrol points are determined in advance. In addition, route information including at least an evaluation value between the candidate points is recorded, and a first step of selecting a plurality of traveling points from the starting point and the destination and the plurality of candidate points; The second step of obtaining the evaluation value between the two required points among the two points that can be selected from the departure point, the destination, and each patrol point, and the evaluation values obtained in the second step are used. Then, in accordance with the third step for determining the order in which the plurality of patrol points are to be visited, and one order from the departure point selected in the first step to the destination in accordance with the order of visits determined in the third step. The number that constitutes the patrol route And the fourth step is a fifth step of repeatedly obtaining a partial route from the starting point selected in the first step to the destination in accordance with the cyclic order determined in the third step. And a sixth step of connecting the partial routes obtained in the fifth step to form one cyclic route.
[0030]
Book According to the invention, the points (candidate points) that can be selected as the traveling points are determined in advance. Further, only the point-to-point route cost obtained based on the point-to-point route is recorded in advance. In this state, by finding the partial route of the cyclic route by repeatedly performing the point-to-point search process, it is not necessary to record the route data, so that the amount of data stored in the storage area can be reduced.
[0031]
Book invention Is Candidate points that can be selected as a plurality of traveling points are determined in advance, points on a plurality of roads that access an area including the candidate points are determined in advance as entry / exit points, and at least each of the candidates Route information including an evaluation value between points and an evaluation value between the entrance / exit point and each candidate point is recorded. In the second step, the departure point or destination selected in the first step is If it is outside the area including the candidate point, the temporary departure point or temporary destination is selected from the entrance / exit points, and the fifth step is the first patrol point from the departure point selected in the first step. When obtaining a partial route to the destination selected in the first step from the last patrol point to the destination selected in the first step. Destination It omitted to seek.
[0032]
Book According to the invention, the temporary departure point or temporary destination selected in the second step is skipped, and the partial route of the cyclic route from the departure point to the destination is obtained. As a result, the optimal tour route is selected from the tour point where the user really wants to stop, so the selection of the tour route that automatically passes through the temporary departure point or temporary destination selected automatically is prevented. Can do.
[0033]
Book invention Is Candidate points that can be selected as a plurality of patrol points are determined in advance, points on a plurality of roads that access an area including the candidate points are determined in advance as entry / exit points, and at least each of the candidate points The route information including the evaluation value between and the evaluation value between the entrance / exit point and each candidate point is recorded. In the second step, the starting point or destination selected in the first step is If it is outside the area including the candidate point, a temporary departure point or a temporary destination is selected from the entry / exit points based on the user's selection, and the fifth step is the departure selected in the first step. A partial route from the ground to the first patrol point via the temporary departure point selected in the second step, or the temporary purpose selected in the second step from the last patrol point Via the ground Determining a partial route to the destination selected in the first step.
[0034]
Book According to the invention, a partial route of a patrol route from the departure point to the destination is obtained without skipping the temporary departure point or temporary destination set by the user. As a result, a patrol route that passes through the temporary departure point or temporary destination set by the user is selected, so that the patrol route always enters or exits the patrol area through the access road intended by the user. Can be requested.
[0035]
Book The invention is a method for efficiently patroling a plurality of points and selecting a patrol route from the starting point to the destination, wherein the first step of selecting the starting point, the destination and the plurality of patrol points; Of the two points that can be selected from the starting point, the destination, and each patrol point, use the second step to obtain the necessary evaluation value between the two points and the evaluation values obtained in the second step. In accordance with the third step of determining the order in which the plurality of tour points are to be visited and the tour order determined in the third step, the departure place, the plurality of visit points and the destination are joined together, And a fourth step that constitutes one patrol route to the destination, and the third step is a fifth step of grouping adjacent ones of a plurality of patrol points into a hierarchical group; Summary in the fifth step Determine the cyclic order of groups in each layer, and finally, and a sixth step of determining the order of patrol multiple cyclic point.
[0036]
Book According to the invention, neighboring tour points are hierarchically grouped. A cyclic route is determined based on the hierarchical group. As a result, the number of combinations to be investigated decreases, so that the route selection time can be shortened especially when the number of points to be visited is large.
[0037]
Book invention of The fifth step collects a plurality of patrol points based on the spatial distance between the two points. According to the fourteenth aspect, it is possible to preferentially group into one group from those having a short spatial distance between two points. As a result, it is possible to group the neighboring tour points, so that the route quality can be improved with a relatively simple process.
[0038]
Book invention of The fifth step collects a plurality of tour points based on the route distance between the two points. According to the fifteenth aspect, it is possible to preferentially group into one group from a route whose distance between two points is short. As a result, it is possible to avoid a group of traveling points where a road between two points is detoured even though it is close in terms of spatial distance, and the route quality can be further improved.
[0039]
Book invention of The fifth step does not form an overly broad group compared to other groups. According to the sixteenth aspect, a group is not formed in a wider range than other groups. As a result, the size of each group can be averaged, and the selection of distorted cyclic routes can be prevented, so that the route quality can be further improved.
[0040]
Book invention of In the sixth step, the cyclic order is determined from the upper hierarchical group among the hierarchical groups. According to the seventeenth aspect, in the hierarchized group, by determining the cyclic order from the upper hierarchy group, the rough cyclic order is first determined in the upper hierarchy, and the overall calculation amount can be reduced. . As a result, it is possible to shorten the time for selecting the optimum route.
[0041]
Book The invention is a method for efficiently patroling a plurality of points and selecting a patrol route from the starting point to the destination, wherein the first step of selecting the starting point, the destination and the plurality of patrol points; Of the two points that can be selected from the starting point, the destination, and each patrol point, use the second step to obtain the necessary evaluation value between the two points and the evaluation values obtained in the second step. Then, according to the third step of determining the order of visiting the plurality of tour points based on the depth-first search method, and the tour order determined in the third step, the departure place, the plurality of visit points, and the destination And a fourth step of constructing one circular route from the departure place to the destination.
[0042]
Book According to the invention, the evaluation value obtained in the second step is for each two points. Therefore, the traveling order determined based on the evaluation value assumes a road on which the vehicle actually travels. As a result, it is possible to obtain an efficient patrol order that does not make the user feel uncomfortable.
[0043]
Of the present invention The second step is to obtain a route between the starting point and each patrol point or destination, between each two points included in a plurality of patrol points, and between each patrol point and the destination. Based on the evaluation value of the route, the evaluation value between the departure point and each patrol point or destination, each evaluation value between each of the two points included in the plurality of patrol points, and between each of the patrol points and the destination Get evaluation value between.
[0044]
Book According to the invention, in the second step, the evaluation value is obtained by actually obtaining a route between each two points. For this reason, it is possible to determine a more preferable circulation order.
[0045]
Of the present invention The second step uses the map information stored in advance as a search start point for each of the departure point, the destination, and each patrol point, and performs a departure side search process and / or a destination side search process. Find the route between the departure point and each of the circuit points or the destination, between each of the two points included in the plurality of circuit points, and between each of the circuit points and the destination. Based on the evaluation value, the evaluation value between the departure point and each patrol point or destination, each evaluation value between each of the two points included in the plurality of patrol points, and between each of the patrol points and the destination Get the evaluation value of.
[0046]
Book According to the invention, in the second step, the route between the two points is obtained by the start side search process and / or the destination side search process with each point as the search start point, so the number of search processes is minimized. To the limit. As a result, routes can be obtained for all necessary combinations in a short time.
[0047]
Of the present invention Map information , It has a hierarchical structure, and the lower layer map information represents a map in which a relatively narrow range is recorded in detail, and the upper layer map information represents a map in which a relatively wide range is recorded roughly. The second step selects the hierarchy of map information to be used, and uses the map information of the selected hierarchy as the starting point for each of the starting point, destination, and each patrol point. Perform side search processing and / or destination side search processing, between the departure point and each of the traveling points or the destination, between each of the two points included in the plurality of traveling points, and each of the traveling points A route to the destination is searched, and when the route information cannot be obtained with the map information of the selected hierarchy, the route is obtained using the map information of the upper hierarchy.
[0048]
Book According to the invention, in the second step, first, search processing from each search start point is started using the map information of the lower hierarchy. Then, when there is a route that cannot be obtained with the lower layer map information, the search process is continued using the upper layer map information. As a result, a search range suitable for the road network recorded in each hierarchy can be set, so that a route between two points with a distance can be obtained in a short time.
[0049]
The present invention Is A method for efficiently patroling a plurality of points and selecting a patrol route from the departure point to the destination, the first step of selecting the departure point, the destination, and the plurality of patrol points, and the departure point A second step of obtaining an evaluation value between two necessary points out of two points selectable from the destination and each of the patrol points, and using each evaluation value obtained in the second step Then, according to the third step of determining the order of visiting the plurality of tour points based on the depth-first search method, and the tour order determined in the third step, And a fourth step of connecting the patrol point and the destination to form one patrol route from the starting point to the destination, wherein the second step includes the starting point and each of the patrols. Point or with the destination The route between each two points included in the plurality of tour points and between the visit point and the destination is obtained, and the departure point and each visit point are determined based on the evaluation value of each found route. Or an evaluation value between the destination, each evaluation value between each two points included in the plurality of patrol points, and an evaluation value between each of the patrol points and the destination, In the fourth step, the point-to-point search is repeated by the number of combinations of two points selectable from the starting point, the destination, and each of the patrol points according to the patrol route determined in the third step. Thus, one cyclic route is configured.
[0050]
Book According to the invention, The evaluation value obtained in the second step is for each two points. Therefore, the traveling order determined based on the evaluation value assumes a road on which the vehicle actually travels. As a result, it is possible to obtain an efficient patrol order that does not make the user feel uncomfortable. In the second step, the evaluation value is obtained by actually obtaining a route between each two points. For this reason, it is possible to determine a more preferable circulation order. Also, By performing the point-to-point search again in the fourth step, it is possible to configure a cyclic route without leaving the route obtained in the second step. As a result, it is possible to reduce the RAM capacity required for selecting a circular route.
[0051]
The present invention Is A method for efficiently patroling a plurality of points and selecting a patrol route from the departure point to the destination, the first step of selecting the departure point, the destination, and the plurality of patrol points, and the departure point A second step of obtaining an evaluation value between two necessary points out of two points selectable from the destination and each of the patrol points, and using each evaluation value obtained in the second step Then, according to the third step of determining the order of visiting the plurality of tour points based on the depth-first search method, and the tour order determined in the third step, And a fourth step of connecting the patrol point and the destination to form one patrol route from the starting point to the destination, wherein the second step includes the starting point and each of the patrols. Point or with the destination The route between each two points included in the plurality of tour points and between the visit point and the destination is obtained, and the departure point and each visit point are determined based on the evaluation value of each found route. Or an evaluation value between the destination, each evaluation value between each two points included in the plurality of patrol points, and an evaluation value between each of the patrol points and the destination, The second step further records route data having a minimum evaluation value for every two required points, and the fourth step follows the second route according to the cyclic route determined in the third step. Based on the route data recorded in this step, one cyclic route is constructed.
[0052]
Book According to the invention, The evaluation value obtained in the second step is for each two points. Therefore, the traveling order determined based on the evaluation value assumes a road on which the vehicle actually travels. As a result, it is possible to obtain an efficient patrol order that does not make the user feel uncomfortable. In the second step, the evaluation value is obtained by actually obtaining a route between each two points. For this reason, it is possible to determine a more preferable circulation order. Also, Since the cyclic route is configured based on the route data of the minimum evaluation value recorded in the second step, it is not necessary to perform the search process again in the fourth step. As a result, it is possible to reduce the time required for selecting a circular route.
[0053]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a block diagram showing a configuration of a car navigation system in which the present route selection method is adopted. In FIG. 1, the car navigation system includes a locator 101, a data storage unit 102, an operation unit 103, a navigation device 104, and an information presentation unit 105.
[0054]
Locator 101 includes a GPS, a vehicle speed sensor, an angular velocity sensor, an absolute direction sensor, and the like, and collects data for calculating the current position of the vehicle.
The data storage unit 102 is usually a CD or DVD, and is necessary for navigation, such as intersection information, road connection status, coordinates / shape / attributes / regulation information, map information about road networks, index information for setting points, etc. Various types of information are stored.
The operation unit 103 includes a remote controller, a touch sensor, a keyboard, a mouse, and the like, and inputs point information and the like according to a user's operation, and specifies an item that the user wants to select.
The navigation device 104 includes a CPU, memory (program memory, working memory), and the like. The navigation device 104 basically includes a location function for identifying the current position based on vehicle movement information obtained from the locator 101, a route guidance function for automatically selecting and guiding a route to the destination, and index information. It has an information search function for searching and displaying a map of an arbitrary place.
The information presentation unit 105 includes a display device (liquid crystal display, CRT display, etc.), a speaker, etc., and presents the current vehicle position, guidance route guidance, and search information to the user.
[0055]
Next, each function of the car navigation system configured as shown in FIG. 1 will be described.
First, the location function will be outlined. The locator 101 collects information on the vehicle position that changes every moment (change angle of the vehicle direction, travel distance of the vehicle, absolute position, etc.). The navigation device 104 calculates the movement locus of the vehicle from the information regarding the vehicle position obtained by the locator 101. The navigation device 104 further compares the calculated vehicle movement trajectory with the road network on the surrounding map obtained from the data storage unit 102 using the map matching technique or the like, Identify the location. The information providing unit 105 displays the current position of the vehicle specified by the navigation device 104 on a map.
[0056]
Next, the information search function will be outlined. The operation unit 103 selects contents to be searched from various items displayed on the display of the information presentation unit 105. The selected item is searched from the recorded information in the data storage unit 102 and displayed on the display of the information presenting unit 105. If necessary, sound or the like is output from the speaker of the information presentation unit 105.
[0057]
Furthermore, the route guidance function according to the present invention will be outlined. First, the user operates the operation unit 103 to set a departure point and a destination. That is, the user operates the operation unit 103 to scroll the map image displayed on the information presenting unit 105, and inputs a desired point as a departure point and a destination. In addition, you may use the present position of the vehicle detected by the above-mentioned location function as a departure place. Next, the navigation device 104 starts searching for the closest point on the node or link on the map stored in the data storage unit 102 with respect to the position of the departure point and the destination set as described above. Adopt as point and search end point. Further, the navigation device 104 calculates the shortest cost route using a known Dijkstra method, etc., and converts the obtained route into a link row, a node row, or a coordinate row to obtain a guidance route. In the present embodiment, cost is adopted as an example of the “evaluation value” in the claims. In the following first to fourth cyclic route selection methods, the cost is regarded as travel time, and a route with a short travel time is selected. Finally, the information presenting unit 105 proceeds in which direction to proceed on the guidance route based on the guidance route obtained by the navigation device 104, the current vehicle position, and the map data stored in the data storage unit 102. The user is informed of whether it should be done by voice or display.
[0058]
In the navigation system, a plurality of patrol points can be set between the departure point and the destination. The user operates the operation unit 103 to set a plurality of patrol points in the same manner as the destination setting. At this time, the area to be visited may be selected before setting the place to be visited. Further, the navigation device 104 refers to the route cost between the departure point, the destination, and the patrol point, passes through each set patrol point, and further reaches the optimum patrol route from the departure point to the destination with a smaller route cost. Is elected. This patrol route selection method will be described in detail later using a flowchart.
[0059]
FIG. 2 is a block diagram showing a general configuration of a delivery vehicle management system to which the present route selection method can be applied. In FIG. 2, the delivery vehicle management system includes a wireless device 201, a data storage unit 202, an operation unit 203, a computer 204, and an information presentation unit 205.
The wireless device 201 includes a mobile phone, a dedicated radio, and the like, and collects the current position of each delivery vehicle from a location system mounted on each delivery vehicle.
A large capacity hard disk is usually used for the data storage unit 202. The data storage unit 202 is a delivery management service such as intersection information and / or road connection status, coordinates, shape, attributes and regulatory information, map information related to the road network, index information for setting points, delivery destination management information, etc. Various information necessary for the storage is stored.
The operation unit 203 includes a remote controller, a touch sensor, a keyboard, a mouse, and the like, and inputs point information and the like according to a user's operation and specifies an item that the user wants to select.
The computer 204 includes a CPU, memory (program memory, working memory), and the like. Basically, the computer 204 automatically selects an optimal delivery route by specifying a plurality of delivery destinations for each vehicle, a location management function for managing the current location of each delivery vehicle obtained from the wireless device 201. It has a route selection function and a delivery destination information management function for retrieving and inputting delivery destination information.
The information presentation unit 205 includes a display device (liquid crystal display, CRT display, etc.), a speaker, and the like, and presents the current position of each delivery vehicle, guidance of delivery routes, and delivery destination information to the user.
[0060]
Next, each function of the delivery vehicle management system configured as shown in FIG. 2 will be described.
First, the location management function will be outlined. The wireless device 201 collects position information of each delivery vehicle that changes every moment. The computer 204 obtains information regarding the position of each delivery vehicle from the wireless device 201 and displays the road network of the delivery area obtained from the map data stored in the data storage unit 202 on the display of the information presentation unit 205. . Further, the information presenting unit 205 displays the current position of each delivery vehicle in a superimposed manner on the displayed road network.
[0061]
Next, the delivery information management function will be outlined. The operation unit 203 can designate an arbitrary position by moving the cursor displayed on the map of the delivery area displayed on the display of the information presentation unit 205 to a desired position with a mouse. Attached information (ID, delivery destination name, contact information, supplementary information) regarding the delivery destination is input from the keyboard of the operation unit 203. In addition, the attached information regarding the delivery destination input in the past is stored in the data storage unit 202, and the information is called and selected by the mouse operation of the operation unit 203. Then, the attached information related to the delivery destination is corrected and deleted by operating the keyboard of the operation unit 203.
[0062]
Furthermore, the route selection function according to the present invention will be outlined. First, the user operates the operation unit 203 to set a delivery point. That is, by operating the mouse of the operation unit 203, the user refers to the map of the delivery area displayed on the display of the information presentation unit 205 and selects a delivery destination for actual delivery. In addition, the present location of the delivery vehicle grasped | ascertained by the delivery center or the above-mentioned position management function can be used for a departure place and a destination. Next, the computer 204 determines which area the delivery destination belongs to, and determines which delivery vehicle to allocate. If the delivery destinations are concentrated in a specific delivery vehicle, the delivery destinations may be assigned to other delivery vehicles that are in charge of the surrounding area and have few delivery destinations. Further, the computer 204 selects an optimum patrol route that starts from the departure point for each delivery vehicle, passes through all assigned delivery destinations, and reaches the destination. This patrol route selection method will be described in detail later using a flowchart.
[0063]
(1) First patrol route selection method
FIG. 3 is a main flowchart showing a processing procedure of the first cyclic route selection method. 3 is performed by the navigation device 104 of FIG. 1 or the computer 204 of FIG.
In FIG. 3, as described above, the user selects a departure point / destination and each traveling point (each delivery point) (steps S301 and S302). Next, a point-to-point route calculation process for obtaining a minimum cost route between the traveling points is performed (step S303).
Here, FIG. 4 is a flowchart showing a detailed processing procedure of step S303 in FIG. In FIG. 4, first, map data in a range including a departure point, a destination, and each patrol point is read in a batch (step S401). Next, one unselected point is selected from the departure point and each traveling point (step S402). The selected point is used in the subsequent processing as a search start point. Next, the search processing based on the known Dijkstra method is expanded over the entire area of the read map data, starting from the search start point (step S403). As a result, a route from the search start point to another point (each patrol point or destination) as the end point is configured, the configured route cost is recorded in the cost matrix, and further, route data indicating the route of the route Is recorded (step S404). Here, the cost matrix is a path cost from the start point to the end point recorded in a matrix having the start point as a row and the end point as a column. FIG. 5 shows an example of the configuration of the cost matrix. In the example of FIG. 5, the number of patrol points other than the starting point and the destination is n. Note that the route cost starting from the destination, the route cost starting from the departure point, and the route cost from the departure point to the destination are not necessary for the cyclic order determination process (step S304) performed later. It is described with a “−” mark.
[0064]
Next to step S404 in FIG. 4, it is determined whether or not an unselected point remains as a search start point (step S405). If there is an unselected point, the process returns to step S402, and one of the unselected points is selected as the next search start point. Thereafter, the series of processes (steps S403 to S405) are repeated. On the other hand, if there is no unselected point in step S405, the point-to-point route calculation process (step S303 in FIG. 3) ends. At the end of the point-to-point route calculation process, the cost matrix of FIG. 5 is completed and each column is filled.
In step S402 described above, the search start point is selected from the departure point and each traveling point, but the search start point may be selected from the destination and each traveling point. In this case, in step S403, the search process is expanded based on the search start point selected from the destination and each patrol point. As described above, the search process may be performed in consideration of the directionality as the search start point.
[0065]
Next, the process proceeds to step S304 in FIG. 3, and a traveling order determination process is performed to determine in what order it is optimal to pass each traveling point (step S304).
Here, FIG. 6 is a flowchart showing a detailed processing procedure of step S304 of FIG. In FIG. 6, after all the tour points are unselected, the tour points are selected one by one. At this time, the same patrol point must not be selected twice. The patrol points are registered in the management information of unselected patrol points in the selected order (step S601). This management information has a list structure illustrated in FIG. The list structure is a chain structure that points to the next point in sequence from the first point. In FIG. 7, n patrol points (N1 to Nn) are continuously arranged in ascending order of subscripts. <N1> is recorded as the first point in “Top”, and <N2> is recorded as the next point after “Top” in “First”. Thereafter, in the same manner, each traveling point is connected in the selected order. Furthermore, <Nn> is recorded as the last selected point in “(n−1) th”, and <End> mark is recorded as ID information indicating the end of the tour route in “nth”. Has been. By adopting the above list structure, it becomes possible to examine the next point in order from the “top”. As a result, in the cyclic recursion process (step S604) to be performed later, it is possible to efficiently find an unselected cyclic point Ni (i = 1, 2,..., N).
[0066]
Refer to FIG. 6 again. After step S601, the departure place is selected as the previous traveling destination PP that is required to calculate the traveling cost X (step S602). Further, the traveling cost X is initialized (step S603). That is, X = 0 is set. Thereafter, a cyclic recursion process is performed (step S604).
Here, FIG. 8 is a flowchart showing a detailed processing procedure of step S604 of FIG. First, the total traveling cost Y, the minimum total traveling cost Ymin, and the intermediate traveling cost X used in the processing procedure of FIG. 8 will be described. Now, it is assumed that n (n is a natural number) patrol points N1 to Nn are selected. In the present embodiment, as illustrated in FIGS. 9 (a) to 9 (c), the traveling route refers to a departure while passing through all the traveling points N1 to Nn without overlapping the same traveling point. It means the route from the ground to the destination.
First, the total traveling cost Y is the cost of one traveling route. Cost from starting point to each traveling point Ni (i is a natural number satisfying i ≦ n), cost between two traveling points Ni and Nj (j is a natural number satisfying i ≦ j and j ≦ n), each The cost from the traveling point Nj to the destination is described in the cost matrix of FIG. Therefore, if the order of patrol around the patrol points N1 to Nn is determined, each cost required for configuring one patrol route is specified. If the identified costs are totaled, the total traveling cost Y can be calculated.
[0067]
In the depth-first search method, after obtaining the total cyclic cost Y of a certain cyclic route, the process of calculating another total cyclic cost Y is repeated. The minimum total cyclic cost Ymin has the smallest value among the total cyclic costs Y that have already been calculated.
Furthermore, in the depth-first search method, when configuring one tour route, different tour points Ni are selected one by one. The traveling cost X in the middle is the cost of the route from the departure point to the currently selected traveling point Ni (the traveling route up to the middle). Similarly to the total traveling cost Y, the intermediate traveling cost X is also calculated with reference to the cost matrix of FIG.
[0068]
In step S801 in FIG. 8, the intermediate traveling cost X and the minimum total traveling cost Ymin are compared to determine whether or not X ≧ Ymin is satisfied (step S801).
When it is determined that X ≧ Ymin, there is no possibility that the currently-visited cyclic route has the minimum total cyclic cost Ymin. Therefore, the cyclic recursion process is aborted because the subsequent processes are useless.
If it is determined that X <Ymin, the current route to be investigated may have the minimum total cost Ymin, so step S802 is performed to perform the subsequent processing. It should be noted that if no cyclic route is obtained at the time of step S801, the minimum cyclic cost Ymin is regarded as infinite. Therefore, in this case as well, it is determined that X <Ymin.
[0069]
Next, the traveling point Ni set at the “top” of the management information in FIG. 7 is selected as the current traveling destination CP (step S802). As a result, the tour point Ni to be visited next to the departure point NSP is selected. Next, the previous circulation destination PP is selected as the starting point, and the current circulation destination CP is selected as the end point. Thereafter, the cost between the start point and the end point is read from the cost matrix (see FIG. 5). The read cost is added to the cyclic cost X (step S803). As a result, a traveling cost X in the middle of the traveling route currently being investigated is obtained.
[0070]
Next, with reference to the management information of FIG. 7, it is determined whether or not the “first” tour point Ni is the last visit destination LP (step S804). If there is an <end> mark immediately after this tour point Ni, the current visit destination CP becomes the last visit destination LP. In other words, since the route to be investigated is one route, step S811 is performed.
In this step, the current circulation destination CP (that is, the last circulation destination LP) is selected as the starting point, and the destination ND is selected as the ending point. Thereafter, the cost between the start point and the end point is read from the cost matrix. The read cost is added to the cyclic cost X (step S811). As a result, the total traveling cost Y (total) of the traveling route currently being investigated is calculated.
[0071]
Next, the total cyclic cost Y obtained this time is compared with the minimum total cyclic cost Ymin, and it is determined whether or not Y <Ymin is satisfied (step S812).
When it is determined that Y <Ymin, the total cyclic cost Y obtained this time is newly recorded as the minimum total cyclic cost Ymin (step S813). Furthermore, the traveling order of the traveling points N1, N2,. Note that if the minimum total cyclic cost Ymin is not obtained at the time of step S812, the minimum cyclic cost Ymin is regarded as infinite. Therefore, in this case as well, it is determined that Y <Ymin. After this step S812, the cyclic recursion process ends.
On the other hand, if it is determined that Y ≧ Ymin, the total recurring cost Y obtained this time is not recorded, and the recursive recursive process ends as it is.
[0072]
In step S804, if the “leading” tour point Ni is not the last tour destination LP, the survey target tour route is not completely configured, so the next tour point Nj (j ≠ i) is selected to go to the tour. In order to calculate the cost X, step S805 is performed.
In step S805, the current circulation destination CP is selected as the previous circulation destination PP. Further, the traveling point Ni selected as the current traveling destination CP is deleted from the management information of FIG. 7 (step S805), whereby only the unselected traveling point Nj (j ≠ i) is managed in FIG. It remains in the information. Furthermore, any of the unselected patrol points Nj (j ≠ i) is “the top”.
[0073]
In the next step S806, the cyclic recursion process of FIG. 8 is called again. At the time of shifting to this step (that is, in the initial state), the traveling cost X up to the middle of the traveling route currently being investigated is calculated. In the current cyclic recursion process, it is first determined whether or not the intermediate stage cyclic cost X ≧ minimum total cyclic cost Ymin (step S801). When X ≧ Ymin, the cyclic recursion process ends as described above. If X <Ymin, step S802 is performed as described above.
Next, the tour point Nj currently set at the “head” of the management information (see FIG. 7) is selected as the current visit destination CP (step S802). As a result, the tour point Nj to be visited next is selected. Thereafter, as described above, the cost between the starting point (= previous traveling destination PP) and the end point (= current traveling destination CP) is added to the traveling cost X (step S803). As a result, the traveling cost X from the departure point NSP to the currently selected traveling point Nj is obtained.
[0074]
In the management information of FIG. 7, if the “leading” tour point Nj is not the last visit destination LP (step S804), the current visit destination CP is selected as the previous visit destination PP. Further, the traveling point Nj selected as the current traveling destination CP is deleted from the management information of FIG. 7 (step S805), whereby only the unselected traveling point Nk (k ≠ i, k ≠ j) is stored. It remains in the management information of FIG. Furthermore, any one of the unselected patrol points Nk becomes “the top”.
[0075]
Thereafter, the cyclic recursion process of FIG. 8 is called again (step S806). As described above, the cyclic recursion process is repeatedly called until the cyclic point Ni disappears from the management information (that is, until one cyclic path is configured) unless it is terminated in step S801. In each cyclic recursion process, the current circulation destination CP is selected from the unselected circulation points Ni remaining in the management information (step S802), and then the current circulation destination CP is set as the previous circulation destination PP. At the same time, the current circulation destination CP is deleted from the management information. The management information from which the previous circulation destination PP and the current circulation destination CP are set is taken over by the next cyclic recursion process. In such a situation, the next cyclic recursion process is recursed.
[0076]
When the above step 806 is completed, the current circulation destination CP deleted in step S805 is returned to the management information of FIG. 7 (step S807). Further, the intermediate cost X is returned to the state before the addition in step S803 (step S808). In the next step S809, the previous tour destination PP is returned to the state before the change in step S805 in order to make the new tour route the subject of investigation. Further, in the management information of FIG. 7, the tour point Ni connected immediately after the current visit destination CP returned in step S807 is selected as the new current visit destination CP (step S809).
[0077]
Next, it is determined whether or not the <end> mark has been selected as the new current circulation destination CP (step S810). If this is the <End> mark, the recursive recursion process ends, assuming that all of the visiting locations of the Yayama residence have been investigated. On the other hand, if this is not the <end> mark, it is determined that all unselected patrol points have not been investigated, and the process returns to step S803 and is repeated in the same manner as described above.
In the cyclic recursion process of FIG. 8 described above, when the first called process is completed, all the cyclic paths are investigated, and the minimum cyclic cost Ymin is recorded. Therefore, the cyclic recursion process in step S604 in FIG. 6 ends, and the cyclic order determination process ends.
[0078]
Here, as a specific example of the cyclic order determination process (step S304), the case where n = 5, that is, the case where the cyclic order of the cyclic points N1 to N5 is determined will be described. In this case, the cost matrix is created as shown in FIG.
First, when step S304 is started, management information as shown in FIG. 11A is created (step S601). In FIG. 11, in the management information, five traveling points (N1 to N5) are arranged in order from the smallest subscript. The <end> mark is recorded at the last "fifth". As a result, the five tour points N1 to N5 to be selected are held.
[0079]
Next, the departure point NSP is selected as the previous traveling destination PP (step S602), and the traveling cost X is initialized to “0” (step S603). Thereafter, the first cyclic recursion process is called, and the total cyclic cost Y1 of the first cyclic route TR1 starts to be calculated. Here, FIG. 12A shows the cyclic route TR1 and the total cyclic cost Y1 obtained in step S304. First, since the current traveling cost X is smaller than the minimum total traveling cost Ymin (= ∞) (step S801), the “leading” traveling point N1 (see FIG. 11A) is set as the current traveling destination CP. This is selected (step S802), whereby a point to be visited next to the departure point NSP is selected. Next, the cost (its value is C01) between the previous destination PP (= departure point NSP) and the current destination (tour point N1) is read from the cost matrix of FIG. It is added to the cost X (= 0) (step S803). As a result, the current cyclic cost X is X = C01.
[0080]
Since the traveling point N1 is not the last traveling destination LP as shown in FIG. 11A (step S804), the traveling point N1 as the current traveling destination CP is set as the previous traveling destination PP, and then the management information (Step S805). The deleted tour point N1 is not selected as the current visit destination CP because it is excluded from the management information in the next and subsequent tour recursion processes. As a result, the traveling route TR is constructed without selecting the same traveling point N redundantly. In addition, as shown in FIG. 11B, the management information is arranged in the order of the other tour points N2 to N5 and the <end> mark as much as the tour point N1 is removed.
[0081]
After step S805, the second cyclic recursion process is called and executed to select the next cyclic destination of the cyclic point N1 and calculate the cyclic cost X up to that point (step S806). The current traveling cost X is C01. The traveling route TR1 is configured from the departure point NSP to the traveling point N1. Further, the traveling point N1 is selected as the previous traveling destination PP. Furthermore, the “head” of the management information is the traveling point N2 as shown in FIG. In this state (initial state), the second cyclic recursion process is recursed.
Even in the second cyclic recursion process (step S807), no cyclic route TR is obtained, so the cyclic cost X (= C01) is smaller than the minimum total cyclic cost Ymin (= ∞) (step S801). . Therefore, the tour point N2 (see FIG. 11B) currently at the “top” is selected as the current tour destination CP, that is, the next tour destination of the tour point N1 (step S802). Next, the cost (its value is C12) between the previous destination PP (= tour point N1) and the current destination CP (tour point N2) is added to the tour cost X (= C01). . As a result, the current cyclic cost X is X = C01 + C12.
[0082]
As shown in the management information of FIG. 11B, the tour point N2 is not the last visit destination LP (step S804), so the visit point N2 (current visit destination CP) is set as the previous visit destination PP. Is deleted from the management information (step S805). As a result, the management information transitions to the state shown in FIG. 11C, and the other tour points N3 to N5 and the <end> mark are arranged in order as the tour point N2 is removed.
[0083]
After step S805, the third cyclic recursion process is called (step S806). At this point, the cyclic cost X is C01 + C12. The traveling route TR1 is configured from the departure point NSP → the traveling point N1 → the traveling point N2. Further, the traveling point N2 is set as the previous traveling destination PP. Furthermore, the “head” is the traveling point N3 as shown in FIG. When the third cyclic recursion process is executed in this state (initial state), after steps S801 and S802 are executed, the previous circulation destination PP (= circulation point N2) and the current circulation destination (= The cost to the traveling point N3) (the value is C23) is added to the traveling cost X (= C01 + C12) (step S803). As a result, the current cyclic cost X is X = C01 + C12 + C23.
Further, after the execution of step S804, the traveling point N3 (current traveling point CP) is set as the previous traveling destination PP and then deleted from the management information as shown in FIG. 11 (d) (step S805).
[0084]
Thereafter, the fourth and fifth cyclic recursion processes (step S806) are called. For convenience, the description of the fourth cyclic recursion process is omitted. Thereafter, when the fifth cyclic recursion process is called, the cyclic cost X is C01 + C12 + C23 + C34. Further, the traveling route TR1 is configured from the departure place NSP → the traveling point N1 → the traveling point N2 → the traveling point N3 → the traveling point N4. Further, the traveling point N4 is set as the previous traveling destination PP. Furthermore, the management information is in a state as shown in FIG. 11E, and the traveling point N5 is set at the “top”. When the fifth cyclic recursion process is executed in this state (initial state), in step S803, between the previous circulation destination PP (circulation point N4) and the current circulation destination CP (circulation point N5). (The value is C45) is added to the cyclic cost X (step S803), and the cyclic cost X is updated from C01 + C12 + C23 + C34 to C01 + C12 + C23 + C34 + C45.
[0085]
Next, since the management information is currently in the state of FIG. 11E, the “leading” tour point N5 is determined as the last visit destination LP (step S804). Then, the cost (the value is C56) between the last circulation destination LP (circulation point N5) and the destination ND is added to the current circulation cost X (step S811). Thus, a cyclic route TR1 is formed. The traveling route TR1 is a route from the departure point NSP to the destination ND in the order of the traveling point N1, the traveling point N2, the traveling point N3, the traveling point N4, and the traveling point N5. In addition, since the cyclic order TR1 is completely configured, the current cyclic cost X (= C01 + C12 + C23 + C34 + C45 + C56) is held as the total cyclic cost Y1.
[0086]
Next, since the total cyclic cost Y1 is smaller than the minimum total cyclic cost Ymin (= ∞) (step S812), the cyclic route TR1 and the total cyclic cost Y1 configured this time are the optimum cyclic route and the minimum total cost. It is recorded as the traveling cost Ymin (step S813). As a result, the fifth cyclic recursion process (step S807) ends, and the cyclic recursion process executed for the cyclic path TR1 ends. When the fifth cyclic recursion process is closed, step S807 and subsequent steps of the fourth cyclic recursion process are performed, and preparation for the cyclic recursion process related to the second cyclic path TR2 is performed.
First, the current circulation destination CP (circulation point N4) deleted in step S805 of the fourth cyclic recursion process is returned to the management information (step S807). As a result, the management information transitions from the state shown in FIG. 11E to the state shown in FIG.
Also, the current cyclic cost X is returned to the state before step S803 of the fourth cyclic recursion process is executed. That is, the current traveling cost is returned to the state immediately before C34 is added in step S803. As a result, the current traveling cost X is returned to X = C01 + C12 + C23.
In step S809, the previous cyclic PP is returned to the one when the fourth cyclic recursion process is called. Therefore, the previous tour destination PP is returned to the tour point N3. Furthermore, the current circulation destination CP is updated from the one set in the fourth cyclic recursion process to a new one (step S809). More specifically, in the current management information (see FIG. 13A), the traveling point N5 connected immediately after the traveling point N4 used in the fourth cyclic recursion process is the new current traveling destination. Set as CP.
[0087]
By executing Steps S807 to S809, as shown in FIG. 12B, the second circuit route TR2 is configured from the departure point NSP → the circuit point N1 → the circuit point N2 → the circuit point N3 → the circuit point N5. become. Note that the traveling cost X of the departure point NSP → the traveling point N1 → the traveling point N2 → the traveling point N3 has already been obtained. This cyclic cost X is also used when obtaining the total cyclic cost Y of the second cyclic route TR2. Then, step S803 is executed to calculate the traveling cost X of the traveling route TR2 up to this middle. That is, as shown in FIG. 12B, the cost (its value is C35) between the previous circulation destination PP (= circulation point N3) and the current circulation destination (circulation point N5) is the current value. It is added to the cyclic cost X (= C01 + C12 + C23), and as a result, the current cyclic cost X becomes C01 + C12 + C23 + C35.
[0088]
As shown in the management information in FIG. 13A, the “first” tour point N4 is not the last tour destination LP (step S804), so the tour point N5 (current visit destination CP) is the previous visit destination. After being set as PP, it is deleted from the management information (step S805). As a result, the management information transitions from the state of FIG. 13A to the state of FIG. 13B, and the <end> mark is reduced in order by the amount of the round trip point N5 being removed.
[0089]
After step S805, the sixth cyclic recursion process is called (step S806). At this time, the cyclic cost X is C01 + C12 + C23 + C35. Further, the traveling route TR2 is configured from the departure point NSP → the traveling point N1 → the traveling point N2 → the traveling point N3 → the traveling point N5. Further, the traveling point N5 is set as the previous traveling destination PP. Furthermore, the “head” is the traveling point N4 as shown in FIG. In this state (initial state), the sixth cyclic recursion process is executed.
First, it is determined whether or not the current cyclic cost X (= C01 + C12 + C23 + C35) is equal to or greater than the minimum total cyclic cost Ymin (= Y1) (step S801). Now assume that X ≧ Ymin (= Y1). Under this assumption, the cyclic route TR2 cannot be optimal. Therefore, it is not necessary to perform the subsequent processing (steps S802 to S813), and the sixth cyclic recursion processing is terminated and ends. As a result, useless processing is not performed, and the number of calculations is reduced.
[0090]
Thereafter, the processing shown in the flowchart of FIG. 8 is performed until the first cyclic recursion processing is closed. When the first cyclic recursion process is closed, an optimal cyclic route having the minimum total cyclic cost Ymin is recorded. As a result, step S304 (cyclic order determination processing) in FIG. 3 ends.
In the next step S305, one traveling route from the departure point NSP to the destination ND is configured according to the optimum traveling order currently recorded. In step S306, the obtained tour route is presented to the user, or the current position accompanying the movement of the vehicle is displayed on the tour route.
[0091]
As described above, in step S304 (cyclic order determination processing) of the present embodiment, processing using the depth-first search method as a basic method is performed, and the optimal cyclic order is investigated. In this cyclic order determination process, the investigation process derived from the current circulation destination CP is terminated when the currently obtained cyclic cost X exceeds the minimum total cyclic cost Ymin. That is, it is possible to omit the investigation process of the combination that cannot be the minimum total traveling cost Ymin. As a result, the time required for selecting the optimum tour order can be shortened.
[0092]
Further, in this step S304 (touring order determination process), the tour point Ni as a selection target is deleted from the management information (list structure) in FIG. 7 when selected. Therefore, at the time when each cyclic recursion process is started, only unselected cyclic points Nj (j ≠ i) are managed in the management information of FIG. Further, in each cyclic recursion process, the current cyclic destination CP is simply selected from the “top” of the management information in FIG. Thereby, an unselected patrol point Ni is efficiently selected. As a result, it is possible to shorten the time required for selecting the optimum tour order.
[0093]
Further, in this step S303 (inter-point route calculation process), each point starts searching when obtaining the minimum cost route between two points included in the departure point NSP, the destination ND, and the n traveling points N1 to Nn. The points are sequentially selected (step S402), and then a unidirectional search process (steps S403 and S404) is performed using the Dijkstra method. As a result, the search process between each of two points at m points (m is the total number of the starting point NSP, the destination ND, and the n traveling points N1 to Nn) from m × (m−1) times to m. Can be reduced to times. This also shortens the time required to select the optimal tour order.
[0094]
(2) Second route selection method
In the second route selection method, the minimum cost between points determined in advance in the data storage unit 102 (FIG. 1) in the first embodiment and the data storage unit 202 (FIG. 2) in the second embodiment. Route results (cost between points and route data) are additionally recorded. Further, the predetermined points include points on the main road for accessing the vicinity of the points. This point is called the entrance / exit point.
Here, FIG. 14 is an example of a predetermined point. In FIG. 14, a sightseeing spot SP indicated by a circle is selected as a predetermined spot within the shaded sightseeing area α. In addition, as roads to access this sightseeing area α, points on the national highway (C1 to C3) and the point (S1) on the high-speed IC of Miemaru are added to the predetermined points as the entrance / exit points. Yes. The second patrol route selection method assumes a case where an optimal patrol order for patroling a plurality of sightseeing spots SP in the sightseeing area α is obtained on the in-vehicle navigation system. The outline of the second patrol route selection method is different from that of the first, and the tour points N1 to Nn are selected from predetermined points and then stored in the data storage unit 102 (or 202). The path cost between points is read out. After that, a cost matrix as shown in FIG. 5 is created, and an optimal tour order is determined in the same manner as in the first tour route selection method. Hereinafter, the second cyclic route selection method of the present invention will be described in more detail.
[0095]
FIG. 15 is a main flowchart showing a processing procedure of the second cyclic route selection method. In FIG. 15, first, the starting point NSP and the destination ND are selected by the user in the same manner as in the first patrol route selection method (step S1001). Next, n traveling points (delivery points) N1 to Nn are selected from a predetermined point group (step S1002). That is, this predetermined point is a candidate point for selecting the tour points N1 to Nn. Next, the point-to-point cost and the route data included in each traveling point Ni selected in step S1002 are read from the previously stored minimum cost route result between the points (step S1003). The read point-to-point cost and route data are stored on the RAM as point-to-point route information.
[0096]
Next, end point addition processing is performed in order to obtain a tour route TR including the departure point NSP and the destination ND that are not predetermined points (step S1004). In short, the end point addition process is a process for obtaining inter-point cost and route data from the departure point NSP to each of the traveling points N1 to Nn and from each of the traveling points N1 to Nn to the destination ND.
FIG. 16 is a flowchart illustrating a processing procedure of end point addition processing. In FIG. 16, first, it is determined whether or not the departure point NSP is included in a predetermined point (step S1101).
If it is included in a predetermined point, the minimum cost route result from the departure point NSP to each of the traveling points N1 to Nn is recorded in the data storage unit 102 (or 202). move on.
On the other hand, if the departure point NSP is not included in the predetermined points, steps S1102 and S1103 are performed to obtain the minimum cost route from the departure point NSP to each of the selected traveling points N1 to Nn. In step S1102, first, map data of an area including the departure place NSP and the selected patrol points N1 to Nn is read onto a RAM (not shown). Further, using the Dijkstra method with the departure point NSP as the search start point, departure point side search processing (search processing for expanding the search in the forward direction with respect to the vehicle traveling direction) is performed (step S1102). Next, after the route configuration is performed from each of the selected tour points N1 to Nn, the minimum cost route from the departure point NSP to each of the tour points N1 to Nn is obtained. Then, the point-to-point cost and route data as the minimum cost route are added to the point-to-point route information (step S1103). Thereafter, step S1104 is performed.
[0097]
Next, it is determined whether or not the destination ND is included in a predetermined point (step S1104).
If included in a predetermined point, since the minimum cost route result from each of the traveling points N1 to Nn to the destination ND is recorded in the data storage unit 102 (or 202), the end point addition process ends. To do.
On the other hand, if the destination ND is not included in a predetermined point, steps S1105 and S1106 are performed, and the minimum cost route from each selected traveling point N1 to Nn to the destination ND is obtained. In step S1105, first, map data of an area including the destination ND and the selected tour points N1 to Nn is read onto the RAM. Furthermore, the destination side search process (search process for expanding the search in the direction opposite to the vehicle traveling direction) is performed using the Dijkstra method with the destination ND as the search start point (step S1105). Next, in step S1106, after the route configuration is performed from each selected tour point N1 to Nn, the minimum cost route from each selected tour point N1 to Nn to the destination ND is obtained. The point-to-point cost and route data as the minimum cost route are added to the point-to-point route information.
Thus, the end point addition process (step S1004) is completed, and thereby, the inter-point information between the two points included in the departure point NSP, the destination ND, and each of the traveling points N1 to Nn is obtained.
[0098]
Refer to FIG. 15 again. When step S1004 (endpoint addition processing) ends, it is determined whether or not the minimum cost route between the endpoint (starting point NSP and destination ND) and each of the traveling points N1 to Nn has been obtained (step S1005). ). In step S1005, in the end point addition process in step S1004, within the range of the map data read into the RAM, from the departure point NSP to each of the tour points N1 to Nn, or from each of the tour points N1 to Nn to the destination ND. This is done in case the route is not found.
If a route has been obtained in step S1005, the process proceeds directly to step S1007 and proceeds to a cyclic order determination process.
On the other hand, if a route is not obtained in step S1005, entry / exit point selection processing is performed to set a temporary departure point DNSP instead of the departure point NSP or a temporary destination point DND instead of the destination ND. .
[0099]
FIG. 17 is a flowchart showing a detailed processing procedure of the entrance / exit point selection processing (step S1006). In FIG. 17, first, it is determined whether or not a route from the departure point NSP to each of the traveling points N1 to Nn has been obtained (step S1201). If it is determined in step S1201 that the route is obtained, the process proceeds to step S1205 without processing. On the other hand, if it is determined in step S1201 that a route has not been obtained, step S1202 is performed.
In step S1202, the one-way hierarchical search process is expanded using the departure point NSP as a search start point. In other words, using the Dijkstra method or the like, the search process is expanded from the starting point NSP with the map data of the lowest hierarchy, and if a certain range is searched, a higher level is recorded with a coarser road network. The search process using the upper map data is repeated up to the highest level. During the search process, after the search process for one layer is completed, it is checked whether or not the entry / exit points (for example, C1 to C3 and S1 in FIG. 14) are in the search end state, and then the connection cost is recorded. Is done.
Next, an entrance / exit point that can be reached from the departure point NSP with the minimum route cost is set as a temporary departure point DNSP (step S1203). At this time, if the points (C1 to C3) on the boundary of the tourist area α are in the search end state, priority may be given to the point (S1) on the main road away from the tourist area α. Next, the route cost and route data from the departure point NSP to the temporary departure point DNSP are recorded (step S1204), and the process on the departure point side ends.
[0100]
Next, it is determined whether or not a route from each of the traveling points N1 to Nn to the destination ND has been obtained (step S1205). If it is determined in step S1205 that a route has been obtained, the entry / exit point selection processing (step S1006) is terminated. On the other hand, if it is determined in step S1205 that a route has not been obtained, step S1206 is performed.
In step S1206, the above-described one-way hierarchical search process is expanded using the destination ND as a search start point. During the search process, after the search process for one layer is completed, it is checked whether the entry / exit points (for example, C1 to C3 and S1 in FIG. 14) are in the search end state, and then the connection cost is recorded. The Next, in step S1207, an entry / exit point that can reach the destination ND with the minimum route cost is set as the temporary destination DND. At this time, if the points (C1 to C3) on the boundary of the tourist area α are in the search end state, priority may be given to the point (S1) on the main road away from the tourist area. Next, the route cost and route data from the temporary destination DND to the destination ND are recorded (step S1208), and the processing on the destination side ends. Above, step S1006 (entrance / entrance point selection process) of FIG. 15 is complete | finished.
[0101]
In the above description, the one-way search method according to hierarchy is adopted. However, in the area above the map where the roads to which the points (C1 to C3) on the boundary of the tourism area α belong are recorded, A bidirectional search method may be employed in which a reverse search is performed with the initial cost of points (C1 to C3) on the boundary of α being 0. In this case, the point (S1) on the main road away from the tourist area α may not be included in a predetermined point as an entrance / exit point.
In the above description, the entrance / exit point that can be reached with the minimum route cost from the departure point NSP (or the destination ND) is automatically adopted as the temporary departure point DNSP (or the temporary destination DND). You may choose which access road you want to choose.
[0102]
Refer to FIG. 15 again. After step S1006 is performed, a tour order determination process for determining the order in which the tour points N1 to Nn are visited is performed (step S1007). The processing in step S1007 is the same as the processing in step S304 in FIG. However, in step S1007, if a temporary departure point DNSP or temporary destination DND is selected in the entry / exit point selection process in step S1006, these are treated as equivalent to the departure point NSP or the destination ND. . Next, the route data obtained in step S1003, step S1004, and step S1006 are connected in accordance with the optimum circulation order obtained in step S1007, and the destination ND (including the temporary departure place DNSP) is connected to the destination ND ( A single traveling route TR (including a temporary destination DND) is configured (step S1008). Finally, the obtained tour route TR is presented to the user, or the current position accompanying the movement of the vehicle is displayed on the tour route, and the direction to be followed is guided (step S1009).
[0103]
As described above, according to the second route selection method, as preconditions, points that can be selected as the tour points N1 to Nn are determined in advance. Further, point-to-point information (point-to-point cost and route data) regarding these points is obtained in advance. As a result, there is a case where it is not necessary to calculate a route between the respective tour points, and the time required for selecting a tour order having the minimum total tour cost Ymin can be shortened.
[0104]
Further, according to the second tour route selection method, the search process based on the Dijkstra method is performed using an arbitrary departure point NSP or destination ND as a search start point, so that each of the tour points N1 to Nn from the departure point NSP. Alternatively, the route from each of the traveling points N1 to Nn to the destination ND is obtained by one search process. As a result, it is possible to shorten the selection time of the minimum cost cyclic route from an arbitrary departure point NSP or the minimum cost cyclic route to an arbitrary destination ND.
[0105]
Further, according to the second patrol route selection method, a point on the road that accesses the patrol area (corresponding to the tourist area α in FIG. 14) is determined in advance as an entrance / exit point. Similarly to the predetermined traveling points, the points between the entrance points and the points that can be selected as the traveling points N1 to Nn, the points that can be selected as the traveling points N1 to Nn, and the doorways The route between the points is obtained in advance and recorded as route information. Therefore, by selecting the entrance point to the patrol area as the temporary departure point DNSP, or the exit point from the patrol area as the temporary destination DND, the cyclic route TR from the temporary departure point DNSP or the temporary destination A traveling route TR to the ground DND can be obtained. As a result, in an environment with a small amount of RAM, such as an in-vehicle device, even when a route from the departure point NSP to each of the traveling points N1 to Nn or a route from each of the traveling points N1 to Nn to the destination ND is not required. The traveling route TR of the selected traveling points N1 to Nn can be obtained.
[0106]
Further, according to the second patrol route selection method, a plurality of points on the main road crossing the border of the patrol area are selected as the entrance / exit points. This makes it possible to mechanically easily set the minimum necessary entry / exit points, simplifying the selection of entry / exit points, and using the search process to automatically set the appropriate temporary departure point. A DNSP or a temporary destination DND can be selected.
[0107]
Further, according to the second embodiment, as the entrance / exit point, a point on the main road that is not included in the circulation area but passes through the vicinity thereof is selected. Thus, an entrance / exit point that can reach the departure point NSP or the destination ND at a minimum cost can be selected by a one-way hierarchical search method from the departure point NSP or the destination ND. As a result, the cyclic route selection process can be simplified, and the time required to select the optimal cyclic order can be shortened.
[0108]
Furthermore, according to the second embodiment, the temporary departure point DNSP or the temporary destination DND can be selected by the user from the entrance / exit points to enter the patrol area from a desired direction or Since it is possible to escape, it is possible to select an approach road to the tour area or an exit road from the tour area according to the user's preference.
[0109]
(3) Third route selection method
In the third route selection method, the minimum cost between points determined in advance in the data storage unit 102 (FIG. 1) in the first embodiment and the data storage unit 202 (FIG. 2) in the second embodiment. Record the route result (only the cost between points). The predetermined point includes a point on the main road for accessing the vicinity of the point. This point on the main road is hereinafter referred to as an entrance / exit point. The difference between the third cyclic route selection method and the second one is that route data between predetermined points is not recorded in advance.
In addition, when selecting a tour location, a tour location is selected from predetermined locations. Thereafter, the route cost between the points stored in the data storage unit 102 (or 202) is read, and a cost matrix is created. Thereafter, as in the second route selection method, the order in which the selected tour points N1 to Nn are visited is determined. Hereinafter, the third method for selecting a tour route will be described.
[0110]
FIG. 18 is a main flowchart showing the processing procedure of the third cyclic route selection method. The flowchart of FIG. 18 is almost the same as the main flowchart (FIG. 15) showing the second cyclic route selection method. Hereinafter, different points will be described. The difference between the third route selection method (FIG. 18) and the second one (FIG. 15) is the following two points.
The first difference is that step S1303 is executed instead of step S1003 of FIG. 15 in the flowchart of FIG. In the second tour route selection method, the route cost and route data between the selected tour points N1 to Nn are obtained in advance in step S1003 in FIG. However, in the third route selection method, since only the route cost between predetermined points is recorded in the data storage unit 102 (or 202) as described above, the selection is made in step S1303 in FIG. Only the route cost between two points included in each of the round points N1 to Nn is read into the RAM.
[0111]
The second difference is that the point-to-point route selection process in step S1308 is performed between step S1307 in FIG. 18 (corresponding to step S1007 in FIG. 15) and step S1309 (corresponding to step S1008 in FIG. 15). is there. In the third cyclic route selection method, route data between points is not recorded in the data storage unit 102 (or 202) as described above. For this reason, after the tour order of the tour points N1 to Nn is determined in the tour order determination process in step S1307, the tour route TR from the departure point NSP to the destination ND cannot be connected in step S1309. Therefore, in the point-to-point route selection process in step S1308, the part from the departure point NSP to the first rounding point Ni, the first rounding point Ni to the second rounding point Nj,..., The last rounding point Nk to the destination ND The route is obtained by repeating a known point-to-point search method.
[0112]
Here, FIG. 19 is a flowchart showing a detailed processing procedure of step S1308 (point-to-point route selection processing). In FIG. 19, first, a departure point NSP is selected as a starting point (step S1401). Next, according to the tour order determined in the tour order determination process in step S1307 (FIG. 18), the next tour point Ni that is the starting point is set as the end point (step S1402). Here, in the entrance / exit point selection process in step S1306 (FIG. 18), when the temporary departure point DNSP or the temporary destination DND is mechanically set, it is preferable not to set these as the end points. However, when the temporary departure point DNSP and the temporary destination DND are set by the user, these are set as the end points.
[0113]
Next, the minimum cost route from the start point to the end point is calculated using a conventional point-to-point search method represented by the Dijkstra method, and then the route data is recorded on the RAM (step S1403). Next, it is determined whether or not the end point is the destination ND (step S1404). If it is determined in step S1404 that the destination is ND, the point-to-point route selection process ends, assuming that all partial routes of the cyclic route TR from the departure point NSP to the destination ND have been obtained.
On the other hand, if it is determined in step S1404 that it is not the destination ND, step S1405 is performed, the current end point is selected as the next start point, the process returns to step S1402, and the above process is repeated.
[0114]
By connecting the partial route data obtained by the above-described point-to-point route selection process in step S1309 in FIG. 18, the route TR that passes from the departure point NSP to the destination ND through each of the circulation points N1 to Nn. Is configured.
[0115]
As described above, according to the third tour route selection method, points that can be selected as the tour points N1 to Nn are determined in advance. In addition, only the route cost between each selectable point is recorded in advance. The partial route of the cyclic route TR is obtained by repeating the point-to-point search process. As a result, unlike the second cyclic route selection method, route data need not be recorded in advance, so that the storage area of the data storage unit 102 or 202 can be used effectively.
[0116]
Further, according to the third route selection method, the temporary departure point DNSP or the temporary destination DND determined in the entry / exit point selection process (step S1306) is skipped, and the route from the departure point NSP to the destination ND is performed. A partial route of the route TR is obtained. Thereby, it is possible to select only the traveling route TR having the minimum total traveling cost Ymin regarding the traveling points N1 to Nn that the user really wants to drop in. As a result, it is possible to prevent the selection of the cyclic route TR that unnaturally passes through the temporarily selected temporary departure point DNSP or temporary destination DNS.
[0117]
Further, according to the third route selection method, the partial route of the route TR from the departure point NSP to the destination ND can be obtained without skipping the temporary departure point DNSP or temporary destination DND set by the user himself / herself. Desired. As a result, it is possible to select a traveling route TR that passes through the provisional departure point DNSP or provisional destination DNS set by the user. As a result, it is possible to obtain a traveling route TR that always enters the traveling area through the access road intended by the user or exits the traveling area.
[0118]
(4) Fourth route selection method
The main flowchart of the present route selection method is the same as the main flowchart (FIGS. 3, 15, and 18) of the first to third route selection methods. Here, in the first to third cyclic route selection methods, the same cyclic order determination process (FIG. 6) is used. This cyclic order determination process has a problem that the processing time increases when the number of points to be visited becomes more than ten, and it is difficult to obtain an optimal cyclic order in a practical time. Therefore, in order to solve this problem, the following cyclic order determination process is adopted instead of the cyclic order determination process of FIG.
[0119]
FIG. 20 is a flowchart showing a cyclic order determination process according to the fourth cyclic route selection method. In FIG. 20, all selected tour points N1 to Nn are selected as target point groups (step S1501). Next, a recursive grouping process is performed for dividing a large number of all round points N1 to Nn into groups of a certain number or less (step S1502).
FIG. 21 is a flowchart showing a detailed procedure of the recursive grouping process. In FIG. 21, first, it is determined whether or not the target point group includes a certain number or more of traveling points N1 to Nn (step S1601). The certain number of values depends on the hardware performance of the device to be used, but is assumed here to be ten. However, the fixed number may be any number as long as the traveling route can be obtained in a sufficiently short time. If it is determined in step S1601 that the number is less than the predetermined number, the recursive grouping process is terminated because it is not necessary to perform grouping.
[0120]
On the other hand, if it is determined in step S1601 that the number is equal to or greater than a certain number, step SS1602 is executed. In step S1602, first, two traveling points Ni with the lowest cost between two points in the target point group are selected. The two selected points are collected as the same group (step S1602). Next, it is determined whether or not the number of groups is a certain number or more (step S1603). Here, with respect to the traveling points Nj (j ≠ i) that are not grouped, one point is converted as one group. Here, the fixed number of values is the same as the constant value in step S1601, and 10 values are adopted. In step S1603, if the number of groups is equal to or greater than a certain number, it is determined that further grouping is necessary, and the process returns to step S1602. On the other hand, in step S1603, if the number of groups is less than a certain number, it is determined that further grouping is not necessary, and the process proceeds to step S1604. In subsequent steps S1604 to S1608, the following investigation is performed on the group configured in steps S1602 and S1603.
[0121]
First, one of unexamined groups is selected as an investigation target (step S1604). Next, it is determined whether or not the number of patrol points Ni included in the group to be investigated is a certain number or more (step S1605). Here, the fixed number of values is the same as the constant value in step S1601, and 10 values are adopted. If it is determined in step S1605 that the number of traveling points Ni is less than a certain number, it is determined that there is no need to further group the groups to be investigated (that is, it is not necessary to create a lower layer group). Is done.
[0122]
On the other hand, in step S1605, if the number of points is a certain number or more, it is determined that the processing time is increased, and it is determined that it is necessary to further create a lower layer group and divide the group to be investigated, and steps S1606 and S1607 Done. First, all the traveling points Ni included in the investigation target group are set as processing targets (step S1606). Next, the recursive grouping process of FIG. 21 is recursively executed in order to divide the traveling point Ni to be processed into lower layer groups. By this recursive grouping process, groups are hierarchically configured. The recursive grouping process is repeated until the number of traveling points Nj (j⊂i) included in each lower layer group becomes less than a certain number.
[0123]
Next, proceeding to step S1608, it is determined whether or not all groups have been investigated in step S1604 (step S1608). If all the groups have not been checked in step S1608, the process returns to step S1604, the next group to be checked is selected, and the subsequent processing is repeated. On the other hand, if all the groups have been checked in step S1608, the recursive grouping process ends.
In the present embodiment, when grouping is performed (step S1602), grouping is performed from points with low cost between two points, but grouping may be performed from points that are close in spatial distance. Moreover, as a result of grouping two points belonging to different groups, a large group may be formed. If this group is much larger than the other groups, the grouping may not be performed.
[0124]
Here, an example of the points grouped by the recursive grouping process is shown in FIG. In FIG. 22, there are 17 traveling points N (see white circles), but eight groups G1 to G8 (shaded portions) are formed by the recursive grouping process of FIG. Further, the departure point NSP is indicated by ☆ and the destination ND is indicated by ★.
Reference is again made to FIG. When the recursive grouping process in step S1502 ends, the hierarchical cyclic order determination process in step S1503 is performed according to the grouped result. In step S1503, after the cyclic order of the upper layer group is determined, the process of determining the cyclic order of the lower layer group is repeatedly executed. As a result, finally, a cyclic order having a minimum total cyclic cost Ymin is obtained. The detailed procedure of the hierarchical cyclic order determination process (step S1503) will be described below with reference to FIG. Note that this hierarchical cyclic order determination process is also a recursive process.
[0125]
When the first hierarchical cyclic recursion process is called, first, in step S1801 in FIG. 23, a group including the first cyclic point Nt is selected as the first cyclic group. Further, the group including the final tour point Ne is selected as the final tour group (step S1801). However, in the first hierarchical traveling process, the group including the traveling point that can be reached from the departure point NSP with the minimum route cost is the first traveling group, and the group including the traveling point that can reach the destination ND with the minimum route cost is the final traveling group. Selected as.
After step S1801, groups other than the first and last tour groups are set to an unselected state. Further, initial management information (list structure) is created for the unselected group (step S1802). The description of the management information created this time is omitted because it has been described in detail in the first patrol route selection method. However, although the management information is created for the traveling point Ni in the first traveling route selection method, it should be noted that the management information is created for the uppermost group Gi in the first hierarchical traveling order determination process. Cost. That is, in the first hierarchical cyclic order determination process, the highest group Gi is collected and regarded as a cyclic point.
[0126]
Next, the head tour group is selected as the previous tour destination PP. Further, the final tour group is selected as the destination ND for convenience (step S1803). Further, the current traveling cost X is initialized (step S1804). Next, cyclic recursion processing is performed in order to obtain a cyclic route TR that efficiently circulates through the uppermost group Gi without passing through the same group repeatedly. The description of this cyclic recursion process is the same as the cyclic recursion process (FIG. 8) detailed in the first cyclic route selection method, and will not be repeated. However, in this cyclic recursion process, the highest group Gi is handled in the same manner as the cyclic point Ni described in the first cyclic route selection method. In addition, the inter-group cost between one group Ga and another group Gb (corresponding to the point-to-point cost in the first cyclic route selection method) is one of the cyclic points Na belonging to the certain group Ga and the other group. It is assumed that the point-to-point cost is the minimum cost among the combinations with any one of the traveling points Nb belonging to Gb. At this time, the traveling point Nj (jεa) selected in a certain group Ga becomes the final traveling point Ne in the group Ga. Further, the tour point Nk (kεb) selected in the other group Gb is the head tour point in the group Gb.
[0127]
As described above, the processing order of steps S1801 to S1805 can be used to determine the order in which the uppermost group Gi can be visited most efficiently. For reference, an example of the cyclic route TR of the uppermost layer group Gi obtained by this series of processing is shown in FIG. In FIG. 24, first, an outline (route with a thick arrow) of the cyclic route TR that passes through all the highest-level groups G1 to G8 from the departure point NSP to the destination ND can be determined.
In the processing after step S1806, the optimal tour order of the tour points Ni included in each group is determined. First, one of the uninvestigated groups Gi is selected as an investigation target (step S1806). Next, it is determined whether or not two or more tour points Nx (xεi) are included in the investigation target group Gx (xεi) (step S1807). If there is one patrol point Nx in step S1806, the process proceeds to step S1810 without processing. On the other hand, if two or more patrol points Nx are included in step S1806, step S1808 is performed.
[0128]
In step S1808, all the traveling points Nx included in the investigation target group Gx are selected as processing targets (step S1808). Next, in order to determine the optimal tour order of the tour points Nx included in the group Gx to be investigated, the hierarchical tour order determination process of FIG. 23 is recursively executed (step S1809). By recursively performing this hierarchical tour order determination process, finally, all tour points Ni are connected by a single tour route TR.
Next, it is determined whether or not all groups Gi have been examined (step S1810). If there is an uninvestigated group Gi in step S1810, the process returns to step S1806, the next group is set as the subject of investigation, and the above processing is repeated. On the other hand, if all the groups have been investigated in step S1810, the hierarchical cyclic order determination process ends. Thereby, the cyclic order determination process shown in FIG. 20 is completed. For reference, an example of the cyclic route TR obtained by the above processing is shown in FIG. In this way, by combining the routes between groups in each layer, it is possible to obtain the entire traveling route TR in which all the traveling points Ni are connected.
[0129]
In the fourth cyclic route selection method, the cyclic route TR is obtained by advancing the processing from the upper layer group to the lower layer group. However, the processing is performed from the lower layer group to the upper layer group. The traveling route TR may be obtained by proceeding. That is, a partial cyclic route having a minimum route cost with each point (group) as a start point and an end point is obtained and recorded from the lower layer. The partial patrol route is recorded for all combinations of two points (or two groups) included in the same hierarchy. Then, it suffices if a cyclic route TR having a minimum route cost as a whole is selected.
[0130]
As described above, according to the fourth tour route selection method, since the tour routes Ni are obtained after the neighboring tour points Ni are grouped into one group, the number of combinations to be investigated is reduced. In particular, when the number of traveling points Ni is large, the time required to select the traveling route TR can be shortened.
Further, in the fourth cyclic route selection method, those having a short spatial distance between two points are preferentially collected into one group, so that the quality of the cyclic route TR selected by a relatively simple process can be improved. it can.
In addition, according to the fourth route selection method, items with small evaluation values (corresponding to route costs) regarding the distance (or travel time) between two points are preferentially collected into one group. Therefore, it is possible to avoid the grouping of the two traveling points Ni such that the road between the two points is detoured even though the spatial distance is close. As a result, the quality of the selected cyclic route TR can be improved.
[0131]
In addition, according to the fourth cyclic route selection method, as a result of grouping, there is a case where one group having a very wide range as compared with other groups is formed. Since the grouping is invalidated, the size of each group can be averaged. As a result, it is possible to prevent the selection of an irregular traveling route TR. As a result, the quality of the selected cyclic route TR can be improved.
Further, according to the fourth tour route selection method, when the optimum tour order is determined, the rough tour order is first determined in the upper layer by determining the tour order from the upper layer group. As a result, the total amount of calculation can be reduced, so that the processing time required to select the optimum cyclic order can be shortened.
[0132]
(5) Fifth patrol route selection method
Since the car navigation system shown in FIG. 1 is mounted on the vehicle, the capacity of the RAM in the system is limited or the processor is ineffective. Under such circumstances, there is a demand for a cyclic route selection method with a light processing load and a small amount of RAM to be used. Such a demand is realized by the following fifth route selection method.
[0133]
FIG. 26 is a flowchart illustrating a processing procedure of the fifth cyclic route selection method. The processing according to the main flowchart of FIG. 26 is mainly performed by the navigation device 104 of FIG.
In FIG. 26, the user designates a starting point and a destination, and a plurality of patrol points in the same manner as described above (steps S2601, S2602).
[0134]
Next, a point-to-point route calculation process is performed based on the selected departure point, destination, and a plurality of patrol points (step S2603). To outline the point-to-point route calculation process, two points that can be selected are selected from a designated starting point, destination, and a plurality of patrol points, and a minimum cost and a minimum cost route between the two selected points are obtained. . Here, the minimum cost preferably means the minimum value of travel time when the vehicle travels between two selected points. The minimum cost route means a route that minimizes travel time. The obtained minimum cost is described in the cost matrix as an evaluation value when obtaining an optimum patrol route.
[0135]
Here, as already described with reference to FIG. 5, the cost matrix is a matrix in which the evaluation value (minimum cost) from the starting point to the ending point of the two selected points is the starting point and the ending point is the column. Is recorded. FIG. 5 shows an example in which the number of patrol points is n in addition to the departure point and the destination. For example, the minimum cost when starting from the departure point and ending at the traveling point N1 is 50 minutes. Note that there is no need for a cost starting from the destination, a cost starting from the departure point, and a cost from the departure point to the destination, so a “-” mark is described in advance in the cost matrix. The
In addition, the said cost should just represent the evaluation value until it reaches | attains from an origin to an end point, and the movement distance or linear distance between two points may be sufficient, and the thing which considered the passage charge may be considered.
[0136]
FIG. 27 is a flowchart showing a detailed processing procedure of step S2603 of FIG. In FIG. 27, first, a hierarchy of map information for starting a route search is set (step S2701). Here, map information having a hierarchical structure is recorded in the data storage unit 102 according to the fifth tour route selection method. That is, the map information of the lowest layer is data obtained by converting a detailed map expressed in the largest scale. Further, the map information at the highest level represents a predetermined range (for example, about 1.5 km square) in detail up to a narrow street. A number of such top-level map information gathers to form a wide range of maps. As the map information becomes higher, the map becomes a smaller scale and becomes a map that roughly represents a wider range (that is, a more main road is recorded). Therefore, in step S2701, when calculating the evaluation value between each two points, it is determined from which hierarchy. Normally, it is desirable to calculate from the lowest hierarchy, but the search may be started from the middle hierarchy or higher in order to reduce the RAM capacity required for processing and the processing load. If the departure point, the destination, and each patrol point are points on a main road, the search may be started from the level where the main road is recorded.
[0137]
Following step S2701, search processing is performed using the map information of the set hierarchy (step S2702). FIG. 28 is a flowchart showing a detailed processing procedure of step S2702. First, the departure point, the destination, and each patrol point are set as search start point candidates (step S2801). Next, one unselected point is selected from the search start point candidates, and the selected point is set as the search start point. Next, the set search start point and map information around the search start point are read from the data storage unit 102 into the RAM (step S2803). The range represented by the read map information is determined in advance.
[0138]
Subsequent to step S2803, a departure side search process and a destination side search process are performed from a search start point or an upper transition point described later (step S2804). As these search methods, the Dijkstra method which is a well-known technique can be used. However, from the search start point or the upper transition point, the departure side search process for the departure point and the destination side search process for the destination from the search start point or the like may be performed. During the execution of each search process, information on the arrival cost from the search start point to another point and the passing link (or the previous passing node) between the two points is recorded as the search result.
Next, a point on the road that is the end point of the search range and recorded in the upper hierarchy is recorded as an upper transition point, and the arrival cost from the search start point to the upper transition point is recorded. In addition, route data from the search start point to the upper transition point is recorded as a candidate route (step S2805). In step S2805, route data may not be recorded. In that case, in step S2605 (FIG. 26) to be described later, route data is created by repeating the conventional point-to-point search according to the cyclic order determined in step S2604.
[0139]
After step S2805, it is determined whether there is an unselected point, and if there is, the process returns to step S2802, and after the next search start point is selected, the subsequent processing is repeated. If there is no unselected point, the search process within the hierarchy is terminated.
[0140]
Reference is now made to FIG. 27 again. When step S2702 ends, connection determination processing is performed to determine whether a point-to-point route has been obtained in the currently set hierarchy (step S2703). FIG. 29 is a flowchart showing a detailed procedure of the connection determination process. First, an unconnected two points are set as candidates for determination (step S2901).
[0141]
Next, one of the set determination target candidates that has not been determined is selected and set as a determination target (step S2902).
Next, in the intra-hierarchy search process, it is determined whether or not a determination point-to-point route has been obtained (step S2903). For example, in order to determine whether a route between two points from point A to point B has been obtained, refer to the search result on the departure side from point A and the search result on the destination side from point B. Investigate whether there is a searched point to be searched. If a common searched point exists, from point A to point B based on the arrival cost from the point A side to the common searched point and the arrival cost from the point B side to the searched point The total cost is calculated, and the route passing through the point where the total cost is minimum is regarded as the route between the points.
[0142]
If there is no common searched point in step S2903, it is determined that a route between two points has not been obtained, and the process proceeds to step S2906 without processing. On the other hand, if there is a common searched point in step S2903, the total cost of the point-to-point route determined in this step is recorded in the cost matrix as the point-to-point cost (step S2904).
[0143]
Further, route data of the point-to-point route obtained in step S2903 is configured (step S2905). At this time, if the end point of the searched route in the currently set layer is an upper transition point from the lower layer, the corresponding one is selected from the lower layer candidate route data recorded in step S2805 as the end point. By adding, all routes from the point A to the point B are configured, and the route is recorded as the point-to-point route information.
Note that step S2905 may not be performed. In that case, in step S2605 (FIG. 26) to be described later, route data is created by repeating the conventional point-to-point search according to the cyclic order determined in step S2604. Of course, it is not necessary to record route data in step S2805 (FIG. 28).
[0144]
Next, it is determined whether there is an undetermined point among the determination target candidates (step S2906), and if there is, the process returns to step S2902, and after another determination target is selected, the subsequent processing is repeated. It is. On the other hand, if there is no undetermined point in step S2906, the connection determination process ends.
[0145]
Refer to FIG. 27 again. When step S2703 is completed, it is determined whether a necessary route between all points (for example, a route between the start point and the end point in the column in which the numerical values in FIG. 5 are written) has been obtained (step S2704). If there is a route between points that have not been obtained, the hierarchy to be searched is moved to the upper hierarchy (step S2705), the process returns to step S2702, and the search process within the hierarchy is repeated in the hierarchy. On the other hand, if all the necessary routes between points have been obtained in step S2704, the route calculation process between points is terminated. Thus, the blank of the cost matrix is filled.
[0146]
Refer to FIG. 26 again. Following step S2603, the tour order of the tour points is determined (step S2604). When the order of patrol is determined by the car navigation system, the number of patrol points is about several (for example, five points), and therefore a normal depth-first search method may be used.
Next, route data from the departure place to the destination is created in accordance with the determined order of travel (step S2605). In this embodiment, the route data between each two points is recorded in step S2905 in FIG. 29. Therefore, the route data between each two points may be sequentially read in accordance with the circulation order determined in step S2604. .
[0147]
In addition, when recording of route data of the candidate route in step S2805 of FIG. 28 and / or recording of route information of the route between points in step S2906 of FIG. 29 is not performed, it is determined in step S2604. It is also possible to create route data by repeating the normal point-to-point search process for each section in accordance with the circulation order.
Subsequent to step S2605, the obtained tour route is presented to the user, or the tour route is guided as the current location moves.
[0148]
The above-described route calculation process between points in FIG. 26 (step S2603) reduces the load of the route calculation process between points as follows. For example, when there are three tour points (N1, N2, N3), there are six combinations between the two tour points. When these six routes are searched by repeating the point-to-point search six times, the results are as shown in FIGS. Here, FIG. 30 to FIG. 35 are explanatory diagrams of processing for obtaining the route between points by repeating the search between two points. In this explanatory diagram, the map information to be used is assumed to be two layers. In this case, the number of searches for the departure side and destination side in the lower hierarchy is 12 times, and the number of searches for the departure side and destination side in the upper hierarchy is also 12 times.
[0149]
In contrast, the point-to-point route calculation process according to the present embodiment is a process as shown in FIG. FIG. 36 is an explanatory diagram of processing for obtaining a route between points in the present embodiment. In the present embodiment, as shown in FIG. 36, the search process on the departure side and the destination side is expanded from each patrol point in the lower hierarchy, and then the process moves to the upper hierarchy to continue the subsequent search process. Therefore, the number of searches for the departure side and destination side in the lower hierarchy is six times, and the number of searches for the departure side and destination side in the upper hierarchy is also six. In this way, the route between points can be obtained with a relatively light load.
[0150]
Here, in the above-described embodiment, the map information recorded in the data storage unit 102 has a hierarchical structure, but may have a single hierarchy. In that case, the processing of step S2701, step S2704, and step S2705 of FIG. 27 is not necessary, but the search processing of step S2804 of FIG. 28 performs the search processing in a sufficiently wide range so as to obtain the path between points. There is a need.
Note that the evaluation value used in the search process performed in step S2804 in FIG. 28 and the evaluation value recorded in the cost matrix are not necessarily the same. For example, the evaluation value recorded in the cost matrix may be recalculated based on the obtained point-to-point route.
[0151]
As described above, according to the present embodiment, after obtaining the minimum evaluation value route between the points, the evaluation value on the route is adopted as the evaluation value between the two points, thereby assuming the road that actually travels. Therefore, it is possible to obtain an efficient patrol route without a sense of incongruity.
Further, according to this embodiment, each point (departure point, destination, and each patrol point) is set as a search start point, and at least one of the departure point side search process and the destination side search process is performed for all points. Since the number of search processing times can be minimized by going and obtaining the route between the points, the route between all the points can be selected in a short time.
[0152]
In addition, according to the present embodiment, the hierarchical map information is used, and after performing the search process with each point as the search start point in the lower hierarchy for all points, the connection determination is performed and the connection is not established. If there is a route, it is possible to set the search range suitable for the road network recorded in the hierarchy by moving to the upper hierarchy and continuing the search process of related points. A route between points can be selected in a short time.
[0153]
In addition, according to the present embodiment, by re-searching the last required section route, it is possible to configure a cyclic route even if the route data is not left when calculating the minimum evaluation value route between the points. Therefore, the RAM capacity used on the vehicle-mounted device can be reduced.
In addition, according to the present embodiment, it is possible to configure a cyclic route by referring to this route data by leaving the route data when calculating the minimum evaluation value route between the points. Since it is not necessary to perform a search process for the configuration, it is possible to shorten the traveling route selection time.
[0154]
Further, in the description of the first to fifth cyclic route selection methods, the depth-first search method is taken as an example to obtain an optimal cyclic route. However, not only the depth-first search method, but also each of the above-mentioned cyclic route selection methods is added to a conventionally known method such as a simulated annealing method (so-called SA method), a genetic algorithm (so-called GA method) or a neural network method. It may be applied.
[0155]
In the first and second embodiments, each functional block may be configured by hardware, or may be configured as a result of program processing such as multitasking of a microcomputer.
Further, the present invention is realized by a program, and can be easily implemented by another independent computer system by recording it on a recording medium such as a floppy disk and transferring it. In this case, the recording medium is not limited to the floppy disk, and any recording medium such as an optical disk, an IC card, and a ROM cassette that can record the program can be similarly implemented.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of a car navigation system according to a first embodiment of the present invention.
FIG. 2 is a block diagram showing a configuration of a delivery vehicle management system according to a second embodiment of the present invention.
FIG. 3 is a main flowchart showing a processing procedure of a first cyclic route selection method.
4 is a flowchart showing a detailed procedure of a point-to-point route calculation process (step S303) in FIG. 3;
FIG. 5 shows an example of a cost matrix.
6 is a flowchart showing a detailed procedure of a cyclic order determination process (step S304) in FIG. 3;
FIG. 7 shows an example of management information having a list structure.
FIG. 8 is a flowchart showing a detailed procedure of the cyclic recursion process (step S604) in FIG. 6;
FIG. 9 shows three examples of cyclic routes.
FIG. 10 shows a cost matrix configured when tour points N1 to N5 are selected.
11 shows state transition of management information in the process of cyclic recursion processing of FIG.
12 shows how processing is terminated for the total cyclic cost Y1 and the cyclic route TR2 calculated in the course of the cyclic recursion process of FIG.
13 shows state transition of management information in the process of cyclic recursion processing of FIG.
FIG. 14 shows a sightseeing spot SP indicated by a circle in the shaded sightseeing area α as an example of a predetermined spot.
FIG. 15 is a main flowchart showing a processing procedure of a second cyclic route selection method.
FIG. 16 is a flowchart showing a detailed procedure of end point addition processing (step S1004) in FIG. 15;
FIG. 17 is a flowchart showing a detailed procedure of the entrance / exit point selection process (step S1006) in FIG. 15;
FIG. 18 is a main flowchart showing a processing procedure of a third cyclic route selection method.
FIG. 19 is a flowchart showing a detailed procedure of the point-to-point route selection process (step S1308) in FIG. 18;
FIG. 20 is a flowchart showing a detailed procedure of a cyclic order determination process according to a fourth cyclic route selection method.
FIG. 21 is a flowchart showing a detailed procedure of the recursive grouping process (step S1502) in FIG. 20;
FIG. 22 shows an example of a tour point N grouped by the recursive grouping process of FIG.
FIG. 23 is a flowchart showing a detailed procedure of the hierarchical cyclic order determination process (step S1503) in FIG. 20;
24 shows an example of a cyclic route between groups of the highest hierarchy obtained by the hierarchical cyclic order determination process of FIG. 23. FIG.
FIG. 25 shows an example of a cyclic route obtained by the hierarchical cyclic order determination process of FIG.
FIG. 26 is a main flowchart showing a processing procedure of a fourth cyclic route selection method.
27 is a flowchart showing a detailed procedure of the point-to-point route calculation process (step S2603) of FIG.
28 is a flowchart showing a detailed procedure of search processing within a hierarchy (step S2702) of FIG.
FIG. 29 is a flowchart showing a detailed procedure of connection determination processing (step S2703) in FIG. 27;
FIG. 30 is a diagram for explaining a process for obtaining a route from the traveling point N1 to N2 in the repetition of the search between two points.
FIG. 31 is a diagram for explaining a process for obtaining a route from the traveling point N1 to N3 in the repetition of the search between two points.
FIG. 32 is a diagram for explaining a process for obtaining a route from the traveling point N2 to N3 in the repetition of the search between two points.
FIG. 33 is a diagram for explaining a process for obtaining a route from a traveling point N2 to N1 in the repetition of the search between two points.
FIG. 34 is a diagram for explaining a process for obtaining a route from a traveling point N3 → N1 in the repetition of the search between two points.
FIG. 35 is a diagram for explaining a process for obtaining a route from a traveling point N3 → N2 in the repetition of the search between two points.
FIG. 36 is a diagram for describing processing for obtaining a route between two points of the tour points N1 to N3 in the fourth tour route selection method;
FIG. 37 is a flowchart showing a processing procedure of a conventional cyclic route search method.
FIG. 38 is a flowchart showing a processing procedure of another conventional cyclic route search method.
[Explanation of symbols]
101 Locator
102 Data storage unit
103 Operation unit
104 Navigation device
105 Information presentation section
201 wireless device
202 Data storage unit
203 Operation unit
204 Computer
205 Information presentation section

Claims (3)

A method for efficiently patroling multiple points and selecting a patrol route from the starting point to the destination,
A first step of selecting a starting point, a destination and a plurality of patrol points;
A second step of obtaining an evaluation value between two necessary points out of two points selectable from the starting point, the destination and each of the patrol points;
A third step of determining an order of patrol the plurality of patrol points based on the depth-first search method using each evaluation value obtained in the second step;
According to the tour order determined in the third step, the starting point, the plurality of tour points, and the destination are connected to form a single route from the starting point to the destination. Including the steps of
The third step includes
A total evaluation value (hereinafter referred to as a first total evaluation value) up to the middle of the cyclic order currently being investigated is calculated based on each evaluation value obtained in the second step; And the steps
When the cyclic order that is currently the subject of the survey is configured, a total evaluation value of the configured cyclic order (hereinafter referred to as a second total evaluation value) is obtained in each of the second steps. A sixth step of calculating based on the evaluation value;
A seventh step of recording a minimum value of the second total evaluation value calculated in the sixth step,
If the first total evaluation value calculated in the fifth step exceeds the minimum value of the second total evaluation value recorded in the seventh step, that derived from it for the cyclic route A patrol route selection method in which the subsequent investigation process is terminated. A device for efficiently patroling a plurality of points and selecting a patrol route from the starting point to the destination,
A point selection unit for selecting a starting point, a destination, and a plurality of patrol points;
An evaluation value acquisition unit that acquires an evaluation value between two necessary points among the two points selectable from the starting point, the destination, and each of the traveling points;
Using each evaluation value obtained by the evaluation value acquisition unit, a traveling order determination unit that determines the order of circulating the plurality of traveling points based on a depth-first search method;
In accordance with the tour order determined by the tour order determination unit, the tour route, the plurality of tour points, and the destination are connected to form a tour route from the departure location to the destination. Including components,
The patrol order determining unit
A total evaluation value (hereinafter referred to as a first total evaluation value) up to the middle of the cyclic order currently being investigated is calculated based on each evaluation value obtained in the second step. A total evaluation value calculator of
When the cyclic order that is currently the subject of the survey is configured, the total evaluation value of the configured cyclic order (hereinafter referred to as the second total evaluation value) is obtained by the evaluation value acquisition unit. A second total evaluation value calculation unit for calculating based on the evaluation value;
A minimum value recording unit that records a minimum value of the second total evaluation value calculated by the second total evaluation value calculation unit,
When the first total evaluation value being currently calculated exceeds the minimum value of the second total evaluation value recorded by the recording unit, the first total evaluation value calculation unit determines whether the first total evaluation value is being calculated. A patrol route selection device that terminates the subsequent investigation process derived from. A recording medium on which a program for realizing an environment for efficiently traveling around a plurality of points and selecting a patrol route from a departure place to a destination on a computer device is recorded,
A first step of selecting a starting point, a destination and a plurality of patrol points;
A second step of obtaining an evaluation value between two necessary points out of two points selectable from the starting point, the destination and each of the patrol points;
A third step of determining an order of patrol the plurality of patrol points based on the depth-first search method using each evaluation value obtained in the second step;
According to the tour order determined in the third step, the starting point, the plurality of tour points, and the destination are connected to form a single route from the starting point to the destination. Including the steps of
The third step includes
A total evaluation value (hereinafter referred to as a first total evaluation value) up to the middle of the cyclic order currently being investigated is calculated based on each evaluation value obtained in the second step; And the steps
When the cyclic order that is currently the subject of the survey is configured, a total evaluation value of the configured cyclic order (hereinafter referred to as a second total evaluation value) is obtained in each of the second steps. A sixth step of calculating based on the evaluation value;
A seventh step of recording a minimum value of the second total evaluation value calculated in the sixth step,
If the first total evaluation value calculated in the fifth step exceeds the minimum value of the second total evaluation value recorded in the seventh step, that derived from it for the cyclic route A recording medium on which the program is recorded, from which the subsequent investigation process is terminated.

Images