JP2005071276A - Car allocation determination method and car allocation determining device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To achieve efficient car allocation, such as a taxi and a load carrying vehicle, at a car allocation center. <P>SOLUTION: A car allocation determining device comprises a means 107 for determining time until car allocation is determined after a request for car allocation; a means 111 for storing each car allocation target position to each car allocation request generated within the time; a means 106 for acquiring the current position of each vehicle; a means 104 for calculating running costs from the current position of each vehicle to each car allocation target position that is stored when the time passes; and means 108, 110 for determining a car that is allocated to each car allocation target position so that the total of the running costs of the entire cars can be minimized by preventing cars from being allocated to two or more car allocation target positions overlappingly. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a vehicle allocation determination method and a vehicle allocation determination device that efficiently allocate a taxi, a cargo transportation vehicle, and the like.

In order to increase the operating efficiency of taxis and baggage transport vehicles, it is important to reduce the time of empty vehicles (when no passengers or luggage are on board). Signals from signposts or GPS (Global Positioning System: Global Positioning System) The location information of each mobile station, such as taxi, measured using such information, and information such as empty cars (no passengers or luggage) or actual vehicles (customers or luggage) Via the AVM system (Automatic)
Vehicle Monitoring System (automatic display system for vehicle position etc.) is used for vehicle allocation planning.

  FIG. 6 is a block diagram illustrating a general configuration example of a conventional vehicle allocation system. In FIG. 6, each vehicle corresponds to the mobile station 10 and the dispatch center corresponds to the base station 100. The mobile station 10 includes a vehicle position detection unit 11, an operation state detection unit 12, and a communication unit 13. The base station 100 includes a communication unit 101, a display output unit 102, a dispatch destination input unit 103, an operation route search unit 104, a map database 105, a vehicle operation monitoring unit 106, and the like.

  As the vehicle position detector 11 attached to the vehicle, there is a device that acquires the position of the vehicle as the vicinity of the position of the sign post from the GPS system or the received identification signal of the sign post as described above. A charge meter is an example of the operation state detection unit 12 that detects the state of a vehicle such as an empty vehicle or a real vehicle. The information acquired by the vehicle position detection unit 11 and the operation state detection unit 12 is sent to the communication unit 101 installed on the base station 100 side by radio or the like by the communication unit 13 mounted on the mobile station 10, and the base station Collected in 100 vehicle operation monitoring units 106. The vehicle operation monitoring unit 106 stores the position and operation state of each vehicle in the internal memory as shown in FIG. 7 and replaces the information with the latest information every time new information is sent from each vehicle. The situation is presented to the operator in the base station in real time through the display output unit 102 such as a display connected to the vehicle operation monitoring unit 106 by superimposing it with the map information.

  When the operator in the base station accepts a request for dispatch, the dispatch destination input unit 103 such as a keyboard is used to input the dispatch destination to the operation route search unit 104. When the dispatch destination is input, the operation route search unit 104 makes an inquiry to the vehicle operation monitoring unit 106, receives a list of empty vehicles that can be dispatched as shown in FIG. 8, and maps all the vehicles in the list to the map. While accessing the database 105, the route to the destination and the operating distance of the destination are calculated. As a result, the vehicle with the shortest operating distance is selected, and the vehicle and the destination are notified to the vehicle operation monitoring unit 106. The vehicle operation monitoring unit 106 changes the reception state of the vehicle so that it cannot be dispatched, and at the same time displays the dispatched vehicle on the display output unit 102, and the operator transmits the dispatch destination position to the vehicle by another means such as wireless. To do.

  When the vehicle carrying the person or the baggage arrives at the destination and switches from the actual vehicle to the empty vehicle, the operation state detection unit 12 detects the change from the actual vehicle to the empty vehicle, and is acquired by the vehicle position detection unit 11 Along with the position information, the information is transmitted to the vehicle operation monitoring unit 106 of the base station 100 via the communication unit 13 and the communication unit 101 by wireless or the like. The vehicle operation monitoring unit 106 changes the state of the vehicle held in the internal memory from an actual vehicle to an empty vehicle, changes the reception state from unassignable to unreachable, and updates the position information. The vehicle that has become an empty vehicle waits until a new destination is notified from the operator of the base station 100.

  As described above, the conventional vehicle allocation plan method includes a method of assigning an empty vehicle in which the operation distance to the vehicle allocation destination on the map is the shortest and no vehicle allocation is arranged when a vehicle allocation request occurs. In addition, a method of assigning a vehicle having the shortest travel time to the vehicle destination taking into account traffic jam information between the vehicle destination and the current position of the vehicle (see, for example, Patent Document 1), and calculating the vehicle density within a limited range A method of selecting a vehicle that exists in an area where the vehicle density is high and has the shortest movement time with respect to the vehicle allocation target position (for example, see Patent Document 2) is known.

  However, in the conventional vehicle allocation planning method, the time point when each vehicle allocation request is generated is used as a reference. Therefore, when a plurality of vehicle allocation requests are generated, there is a possibility of inefficient vehicle allocation with respect to the movement distance of the entire vehicle. is there. FIG. 9 shows such an example. In FIG. 9, the star indicates the vehicle, and the circle indicates the destination, and first, a request for dispatch to the point a is entered, and the nearest vehicle A is determined to dispatch to the point a. Immediately after that, a request for dispatching to the point b is entered, so that the vehicle B is determined to be dispatched to the point b. Certainly, when the request for dispatch to the point a is entered, it is most efficient to dispatch the vehicle A to the point. However, when the request for dispatch to the point b is subsequently entered, the vehicle A is If the vehicle is dispatched to the point b and the vehicle B is dispatched to the point a, the travel distance to the points a and b of the vehicles A and B can be shortened. In other words, in the situation where multiple allocation requests occur before and after, the conventional method of directing the closest empty vehicle to the allocation destination at the time when each individual allocation request occurs is the sum of the movement distances of multiple vehicles. In some cases, it is inefficient.

Japanese Patent Laid-Open No. 11-161899 JP-A-10-293900

  In a situation where a plurality of dispatches occur before and after the present invention, even if the travel cost from when a dispatch request is made until the dispatch vehicle reaches the dispatch target position is not minimized, a plurality of dispatch target positions By determining the combination of vehicles to be allocated to the vehicle as a whole so that the total travel cost to each vehicle allocation target position is minimized, the vehicle allocation to the vehicle allocation target position is determined individually for each vehicle allocation request. This is to realize a more efficient vehicle allocation decision than the conventional method.

  In the present invention, when a vehicle allocation request (vehicle allocation request) is generated, the decision as to which vehicle is to be dispatched is delayed, and a plurality of vehicle allocation request destinations newly generated during that time are sequentially stored. After a certain amount of time has elapsed since the vehicle was dispatched, all vehicles overlap so that the total cost of travel to each vehicle destination (vehicle allocation target position) memorized during that time has been minimized. Thus, a configuration is adopted in which a vehicle to be dispatched is determined for each of the dispatch destinations so as not to be assigned to two or more dispatch destinations. Specifically, the current position of each vehicle is set as one node group, each dispatch target position is set as the other node group, and the weight of the branch connecting the vehicle node and the dispatch target position node is set to the travel target position. The matching between the vehicle node and the vehicle allocation target position node, which is the cost and the sum of the weights is minimized, is calculated as a bipartite graph weighted matching, and each vehicle allocation purpose is calculated using the calculation result of the bipartite graph weighted matching. Determine the dispatch vehicle for the location.

  The travel cost here means the cost required for the vehicle to reach the destination. Specifically, the fuel cost including the travel distance, travel time, and body weight to the destination, or those costs Is obtained by multiplying by a weighting factor to obtain a sum, etc., which can be arbitrarily determined in advance by an operator.

As a method of calculating the travel distance of the vehicle between two points, a straight distance on a map may be used, or a distance along a road may be used. As the distance along the road, the distance of the shortest route obtained when the shortest route is obtained is used. The shortest route calculation method includes the Dijkstra method, the Nicolson method based on the Dijkstra method, the A * algorithm, etc. (for example, Kamikawa, Umetsu “Route Search Technology for In-vehicle Navigation Systems”, Measurement and Control, Vol. 36, No. 11) Pp. 790-792). Further, as the sum of the travel distances of the respective vehicles to the respective vehicle allocation target positions, the sum of the travel distances multiplied by a count corresponding to the fuel consumption of each vehicle, which varies depending on the vehicle body weight or the like, may be used. As for the travel time of a vehicle between two points on the map, a value obtained by simply dividing the travel distance by the average speed of the vehicle may be used, or a predicted time that includes traffic information on the travel route between the two points. There may be.

  The entire point can be divided into two point sets, and a point set and a branch set that are not at both ends of a branch even if one arbitrary point is extracted from both point sets is called a bipartite graph. A branch set in which any two branches do not share an end point is called matching. In a bipartite graph in which weights are given to each branch, matching that minimizes the total sum of weights is called minimum weight matching. An algorithm for obtaining this minimum weight matching is, for example, “Algorithm Dictionary” (Goichi Shimauchi et al., Kyoritsu Publication, 1994).

  Each vehicle and each vehicle allocation target position correspond to the two point sets in the bipartite graph, and the vehicle allocation of the vehicle as a weight of a branch with the point corresponding to the vehicle and the point corresponding to the vehicle allocation target position as an end point A bipartite graph of the combination of the vehicle allocation target position and the vehicle allocation vehicle that minimizes the total of the total vehicle travel costs by assigning the travel cost to the target location so that no vehicle is assigned to two or more vehicle allocation target locations As a solution to the minimum weight matching problem.

  As a result, even if the travel cost from when a vehicle allocation request is made until the vehicle arrives at the vehicle allocation target position is reduced, a combination of vehicle allocations for a plurality of vehicle allocation destinations is assigned to each vehicle allocation target position as a whole. By determining so that the sum of the travel costs up to the minimum can be achieved, it is possible to realize more efficient vehicle allocation than individual determination of the vehicle to be allocated to the vehicle allocation target position for each vehicle allocation request.

  Next, it is possible to always set a certain time as the time from when there is a (first) dispatch request until the dispatched vehicle is determined (hereinafter referred to as the dispatch decision delay time). If it is too long, the customer who requested the vehicle dispatch will be kept waiting for a long time. Conversely, if it is too short, a plurality of vehicle dispatch requests will not be generated, and the advantages of the present invention cannot be utilized. Therefore, for example, a method of determining the dispatch determination delay time according to the dispatch request occurrence frequency per unit time by referring to a database that records the occurrence frequency of the dispatch request for each time zone in one day is preferable. This method for determining the vehicle allocation decision delay time is a method that takes into consideration that the frequency of occurrence of vehicle allocation requests varies depending on the time zone, weather conditions, and the like. In times when the frequency of dispatch requests is high, multiple dispatch requests occur in a relatively short time. On the other hand, in times when the frequency of dispatch requests is low, multiple dispatch requests must occur without waiting for a relatively long time. Conceivable. There is no problem in setting the vehicle allocation decision delay time short, but if it is set long, it will cause the customer who requested the vehicle allocation to wait first, so it is advisable to set a certain upper limit.

  In the present invention, instead of immediately assigning the nearest empty vehicle to each vehicle allocation target position alone, the time is set, and after a certain (first) vehicle allocation request (vehicle allocation request) is generated, By accepting further dispatch requests until the time has elapsed, and assigning dispatch vehicles to individual dispatch target positions so that the total travel cost of the entire vehicle is minimized, Compared with the case where a vehicle is allocated independently, more efficient vehicle allocation becomes possible.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

  FIG. 1 is a block diagram showing an embodiment of a vehicle allocation system according to the present invention. In FIG. 1, each vehicle corresponds to a mobile station 10, and a vehicle allocation center as a vehicle allocation determination device corresponds to a base station 100. Each vehicle of the mobile station 10 includes a vehicle position detection unit 11, an operation state detection unit 12, and a communication unit 13. A dispatch center of the base station 100 is a communication unit 101, a display output unit 102, and a dispatch destination. The point provided with the input part 103, the operation route search part 104, the map database 105, and the vehicle operation monitoring part 106 is the same as the conventional dispatch system shown in FIG. The difference from the conventional vehicle allocation system is that the allocation center (vehicle allocation determination device) of the base station 100 further notifies the time 107, the bipartite graph weighted matching calculation unit 108, and the first vehicle allocation request. A vehicle allocation decision delay time determining unit 109 that determines a time from when there is a vehicle allocation (vehicle allocation determination delay time), and a vehicle allocation planning unit that determines a vehicle to be allocated using information from the bipartite graph weighted matching calculation unit 108 110, and a dispatch destination storage unit 111 for storing the dispatch target position of each dispatch request.

  In the mobile station 10 of each vehicle, the current position of the vehicle detected by the vehicle position detection unit 11 such as GPS and a sign post and the operation state detection unit 12 and the operation state such as an actual vehicle or an empty vehicle are transmitted via the communication unit 13 by radio or the like. To the base station 100 of the dispatch center in real time. In the base station 100, the vehicle operation monitoring unit 106 receives the position information and the operation state of each vehicle via the communication unit 101 and stores them in the internal memory, and superimposes the map information obtained from the map database 105 on a display or the like. To the display output unit 102. These operations are the same as those of the conventional example shown in FIG.

  Next, the operation of the vehicle allocation planning unit 110 and its surroundings in the base station 100 when there is a vehicle allocation request will be described based on the flowcharts shown in FIGS. 4 and 5 are flowcharts of different tasks that operate in parallel. FIG. 4 is an example of a flowchart of a task for receiving a vehicle allocation target position. FIG. 5 is a flowchart of each task received after the vehicle allocation decision delay time has passed. It is an example of the flowchart of the task which determines a dispatch vehicle with respect to a dispatch target position.

First, a task for receiving a dispatch destination will be described with reference to FIG.
Step 1001 is a waiting state for dispatch destination input. When an operator in the base station 100 receives a request for dispatch and inputs a dispatch destination to the dispatch destination input unit 103, the process proceeds to step 1002.

  In step 1002, the vehicle allocation plan unit 110 receives the vehicle allocation destination from the vehicle allocation destination input unit 103, receives the time from the clock 107, and stores the vehicle allocation destination position in the vehicle allocation destination storage unit 111.

  In step 1003, the vehicle allocation planning unit 110 checks whether or not the vehicle allocation determination delay is in progress, that is, whether or not the vehicle allocation determination delay state has already been reached by another vehicle allocation request. Returning to, waits for another dispatch request, and if it is not in the dispatch decision delay state, that is, if it is the first dispatch request, the process proceeds to step 1004.

  In step 1004, the vehicle allocation planning unit 110 requests the vehicle allocation determination delay time determination unit 107 to determine the vehicle allocation determination delay time. The simplest method for determining the vehicle allocation decision delay time is that the vehicle allocation determination delay time determination unit 109 always returns a fixed time to the vehicle allocation planning unit 110 as the time until the vehicle allocation determination process. Another method is to finely determine the vehicle allocation decision delay time according to the time at which the vehicle allocation request occurs, based on the statistics for each time zone regarding the time at which the vehicle allocation request occurs. This method will be described in detail with reference to FIGS.

  FIG. 2 shows a configuration example of the vehicle allocation determination delay time determination unit 109, which includes a delay time determination unit 210 and a vehicle allocation request frequency database 220. FIG. 3 is a conceptual diagram of the dispatch request frequency database 220, in which the average number of dispatch requests at each time is stored in advance based on statistics. The delay time determination unit 210 obtains the current time from the clock 107, inquires the dispatch request frequency database 220, examines how much time has elapsed from the current time to reach the dispatch request accumulation target number, and determines the maximum dispatch With the decision delay time as the upper limit, the time interval is calculated as the vehicle assignment decision delay time. In other words, assuming that the current time is H hours and M minutes, if A + B is less than the dispatch request accumulation target number and A + B + C exceeds the dispatch request accumulation target number, the delay time determination unit 210 sets the dispatch decision delay time as 3 minutes. Decide. However, when the maximum vehicle allocation decision delay time is exceeded, the delay time determination unit 210 takes the maximum vehicle allocation decision delay time as the vehicle allocation decision delay time. The delay time determination unit 210 returns the determined vehicle allocation determination delay time to the vehicle allocation planning unit 110. It should be noted that the operator can arbitrarily determine in advance the target number of dispatch request accumulation and the maximum dispatch decision delay time.

  In step 1005, the vehicle allocation planning unit 110 sets a time obtained by adding the vehicle allocation determination delay time determined by the vehicle allocation determination delay time determination unit 109 in step 1004 to the current time as a vehicle allocation determination processing start time in a register or the like. In step 1006, the vehicle allocation planning unit 110 sets the vehicle allocation determination delay in the flag register or the like, returns to step 1001, and waits for the next vehicle allocation request.

Next, a task of determining a dispatch vehicle for each accepted dispatch destination will be described with reference to FIG.
In step 2001, the vehicle allocation planning unit 110 acquires the current time from the clock 107. In step 2002, the vehicle allocation planning unit 110 compares the current time acquired in step 2002 with the vehicle allocation determination processing start time set in the previous step 1005. If the vehicle allocation determination processing start time has not yet been reached, the vehicle allocation planning unit 110 returns to step 2001. If the vehicle allocation determination process start time has been reached, the process proceeds to step 2003.

  In step 2003, the vehicle allocation planning unit 110 cancels the vehicle allocation decision delay set in the previous step 1006. Then, the process proceeds to Step 2004. Note that step 3003 and step 2004 may be reversed.

  In step 2004, the vehicle allocation planning unit 110 acquires the vehicle allocation target position stored so far in the vehicle allocation destination storage unit 111 and sends it to the operation route search unit 104. At this time, the information on the destination position in the destination storage unit 111 is deleted, and the task of accepting the destination in FIG. 4 waits for input of a new destination in step 1001. Then, the first dispatch request after this time is handled again as the first dispatch request, and the dispatch decision delay state is entered from the time when the dispatch request is generated.

  In step 2005, the operation route search unit 104 acquires a list of empty vehicles that can be dispatched as shown in FIG. As described above, FIG. 7 and FIG. 8 show the current position, the vehicle state such as the actual vehicle / empty vehicle and the reception state regarding the vehicle request for each vehicle, but the vehicle target position taking into account the fuel for each vehicle. In addition, a fuel efficiency coefficient indicating the amount of fuel consumed per unit distance for each vehicle is also included.

  In step 2006, the operation route search unit 104 determines the vehicle allocation purpose for each empty vehicle that can be allocated in step 2005 for each vehicle allocation target position sent from the vehicle allocation planning unit 110 in step 2004. The travel cost for the shortest route to the position is calculated using the information in the map database 105 and the result is sent to the vehicle allocation planning unit 110.

As described above, the shortest route search method performed by the operation route search unit 104 includes the Dijkstra method (“Algorithm Dictionary” by Goichi Shimauchi et al., Kyoritsu Shuppan, 1994) and the Nicolson method based on this Dijkstra method. ("Automobile Route Guidance System Research Expert Committee: IEEJ Technical Report Automotive Route Guidance System 26", The Institute of Electrical Engineers of Japan (1994)) and A * algorithm
Techniques ", T Ikeda et al. Vehicle Navigation & Information Systems Conference 1994).

In step 2007, the vehicle allocation planning unit 110 uses the bipartite graph weighted matching calculation unit 108 to set the current position of each vehicle as one node group, each vehicle allocation target position as the other node group, and the vehicle node and vehicle allocation purpose. Using the weight of the branch connecting the position nodes as the travel cost of the vehicle with respect to the vehicle allocation target position, the matching between the vehicle node and the vehicle allocation target node that minimizes the sum of the weights is calculated, and the result is stored in the internal memory. By using the bipartite graph matching calculation, no vehicle is assigned to more than one vehicle destination position.
As described above, a calculation algorithm for weighted matching of a bipartite graph is described in, for example, “Algorithm Dictionary” (Shinichi Shimauchi et al., Kyoritsu Shuppan, 1994).

  In step 2008, the vehicle allocation planning unit 110 determines which vehicle is allocated to which vehicle allocation destination by matching the vehicle acquired in step 2007 with the vehicle allocation target position, and outputs it to the display output unit 102. By looking at the display on the display output unit 102, the operator of the dispatch center reads the information, and notifies the dispatch destination directly to the dispatched vehicle using a separate radio device or the like using voice.

  Regarding the transmission of the vehicle allocation target position to the corresponding vehicle allocation vehicle, the vehicle allocation monitoring unit 106 notifies the vehicle operation monitoring unit 106 of the matching result of the vehicle allocation target position acquired by the vehicle allocation planning unit 110 in step 2007 to the vehicle operation monitoring. The configuration may be such that the unit 106 transmits the vehicle allocation target position to the corresponding vehicle through the communication unit 101.

  Although one embodiment of the present invention has been described above, the base station 100 shown in FIG. 1 is a computer, and some or all of the processing functions of each unit in the base station 100 are configured by a computer program. Needless to say, the present invention can be realized by executing it using a computer, or the processing procedures shown in FIGS. 4 and 5 can be configured by a computer program and the program can be executed by the computer. In addition, a computer-readable recording medium such as an FD, MO, ROM, memory card, CD, or the like is stored in the computer. In addition, the program can be recorded and stored on a DVD, a removable disk, etc., and the program can be distributed through a network such as the Internet.

It is a block diagram which shows the structural example of the vehicle allocation system by this invention. It is a block diagram which shows the structural example of a vehicle allocation determination delay time determination part. It is a conceptual diagram of a vehicle allocation request frequency database. It is an example of the flowchart of the vehicle allocation purpose position reception task in this invention. It is an example of the flowchart of the dispatch vehicle determination task in this invention. It is a block diagram which shows the structural example of the conventional vehicle allocation system. It is a conceptual diagram which shows the state of each vehicle. It is a conceptual diagram which shows the state of each empty vehicle. It is the conceptual diagram which showed operation | movement of the conventional dispatch system.

Explanation of symbols

10 Mobile station (vehicle)
DESCRIPTION OF SYMBOLS 11 Vehicle position detection part 12 Operation state detection part 13 Communication part 100 of a mobile station Base station (vehicle allocation center)
DESCRIPTION OF SYMBOLS 101 Base station communication part 102 Display output part 103 Allocation destination input part 104 Operation route search part 105 Map database 106 Vehicle operation monitoring part 107 Clock 108 Two parts Graph weighted matching calculation part 109 Allocation determination delay time determination part 110 Allocation plan part 111 Allocation destination storage

Claims (5)

A method of automatically determining a dispatch for a dispatch request at a dispatch center,
The time from when there is a certain vehicle allocation request until the vehicle allocation decision is determined, each vehicle allocation target position for each vehicle allocation request generated within the time is stored, and when the time has elapsed, each stored vehicle allocation purpose is stored. Calculate the travel cost from the current position of each vehicle with respect to the position, so that no vehicle is duplicated and assigned to two or more target allocation positions, and the total travel cost of the entire vehicle is minimized. A vehicle allocation determination method characterized by determining a vehicle to be allocated to a vehicle allocation target position.   A certain number of times depending on the time at which the vehicle allocation request is generated, based on the statistics for each time zone regarding the time at which the vehicle allocation request occurs, as the time from when the vehicle allocation request is generated until the vehicle allocation is determined. 2. The vehicle allocation determination method according to claim 1, wherein a time interval at which a vehicle allocation request is generated is set.   Means for determining a time from when there is a vehicle allocation request until vehicle allocation is determined, means for storing each vehicle allocation target position for each vehicle allocation request generated within the time, means for acquiring the current location of each vehicle, When the determined time has elapsed, the means for calculating the running cost from the current position of each vehicle for each stored vehicle allocation target position and any vehicle are assigned to two or more vehicle allocation target positions in an overlapping manner. And a means for determining a vehicle to be allocated to each vehicle allocation target position so as to minimize the total traveling cost of the entire vehicle.   In the means for determining the time from when a certain vehicle allocation request is generated to when the vehicle allocation is determined, a constant is determined according to the time at which the vehicle allocation request occurs, based on the statistics for each time zone regarding the time at which the vehicle allocation request occurs. 4. The vehicle allocation determination device according to claim 3, wherein a time interval at which the number of vehicle allocation requests is generated is set.   The means for determining a vehicle to be dispatched with respect to the vehicle allocation target position includes the current position of each vehicle as one node group, each vehicle allocation position as the other node group, and a branch line connecting the vehicle node and the vehicle allocation target node. Means for calculating a matching of a vehicle node and a vehicle allocation target position node as a bipartite graph weighted matching, wherein the weight is a running cost of the vehicle with respect to the vehicle allocation target position, and the sum of the weights is minimized, and the bipartite graph A vehicle allocation determining device comprising means for determining a vehicle to be allocated to each vehicle allocation target position using a calculation result of weighted matching.

Images