JP2005112609A - Delivery plan preparation method, delivery plan preparation device, delivery plan preparation program, and physical distribution system - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To prepare a delivery plan having excellent transportation efficiency when delivering articles to be delivered to a plurality of delivery destinations from a plurality of delivery origins and even when loading other delivery articles after unloading delivery articles at each delivery destination and delivering them to the next delivery destination. <P>SOLUTION: This delivery plan preparation method comprises initial plan preparation processing for preparing an initial delivery plan based on one or a plurality of delivery origins, one or a plurality of delivery destinations, the articles to be delivered, a transportation system, and restriction conditions related to them, plan candidate preparation processing for preparing a delivery plan candidate by correcting the initial delivery plan, evaluation value calculating processing for obtaining an evaluation value for evaluating the delivery plan candidate, and plan adoption determining processing for comparing evaluation values of each delivery plan candidate and adopting the delivery plan candidate having the best evaluation value as a final delivery plan. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a delivery plan creation method, a delivery plan creation device, and a delivery plan creation program for creating a delivery plan for delivering various delivery items from a delivery source to a delivery destination using a transportation system such as an automobile, a freight car, and a ship. The present invention also relates to a physical distribution system provided with such a delivery plan creation device.

  In a conventional delivery plan creation method, a genetic algorithm is used to first assign a delivery item to a transport agency (distribution), and then create a delivery route of a delivery destination assigned to each transport agency ( For example, see Patent Document 1.) Here, the genetic algorithm refers to a technology that imitates the genetic mechanism of an organism and applies it to engineering. Hereinafter, this technique is referred to as a first conventional example. Also, some conventional delivery plan creation methods attempt to simultaneously optimize the delivery route of the delivery destination assigned to the vehicle and each transport using a simulated annealing method (for example, Patent Documents). 2). For details of the simulated annealing method, see S. Kirkpatrick, CDGelatt Jr., MPVecchi: "Optimization by Simulated Annealing", Science, Vol. 220, No. 4598, pp. 671-680 (1983). I want to be. Hereinafter, this technique is referred to as a second conventional example.

  Further, the conventional delivery plan creation method includes a first step of creating an initial delivery plan based on the input delivery request data and vehicle data, and the initial delivery plan as initial data of a delivery plan that is subsequently updated, Using the tabu search method, the second step to modify the delivery plan and create a revised plan, the third step to evaluate the revised plan modified in the second step, and the revised plan to date If it is determined whether or not it is the best delivery plan that has been created, if it is the best delivery plan, it is stored as the best delivery plan and the delivery plan used in the second step is updated with this amendment. Determines whether or not this proposed amendment improves the current delivery plan, and if so, updates the delivery plan used in the second step with this amendment, and if not, Determine whether the amendment is the best among the amendments to the current delivery plan. If it is the best, update the delivery plan used in the second step with this amendment. If not, update the current delivery plan. And a fourth step for performing a series of processes that are not performed, and the second to fourth steps are repeatedly executed to select the best delivery plan that is finally stored (for example, see Patent Document 3). .) Here, the tabu search method is a metaheuristic solution proposed by Glover (1985). In the tabu search method, first, a function Q for converting one solution into another solution is determined starting from an initial solution, and then a set called a neighborhood having a solution as an element from the solution using the function Q. Create In each stage, a neighborhood is created from a given solution, and one solution is selected from them. Let the solution obtained here be the original solution in the next stage. This series of operations is repeated. In order to prevent cycles in this repetition, the past history is stored in a storage area called a taboo list. Hereinafter, this technique is referred to as a third conventional example.

JP-A-10-55349 (Claim 7, [0030] to [0072], FIGS. 2 to 16) Japanese Patent No. 2816802 (Claim 1, Claim 2, [0063], [0064], FIGS. 1 to 4) JP 2002-302257 A (Claim 1, Claim 23, [0075] to [0122], FIGS. 5 to 15)

  By the way, recently, in addition to delivering a delivery item from a single distribution center (delivery source) to a plurality of delivery destinations as in the past, it is desired to deliver a delivery item from a plurality of delivery sources to a plurality of delivery destinations. There has been a demand or a desire to load a delivery product at each delivery destination and then load another delivery product and deliver it to the next delivery destination. However, since the above-described first to third conventional examples are intended only to deliver a delivery item from one delivery source to a plurality of delivery destinations, they cannot respond to the above-described demand. There was a problem that a delivery plan with good transportation efficiency could not be created.

  The present invention has been made to solve the above-described problems. The purpose of the present invention is to deliver a delivery product from a plurality of delivery sources to a plurality of delivery destinations after the delivery products are wholesaled at each delivery destination. A delivery plan creation method, a delivery plan creation device, a delivery plan creation program, and a distribution system that can create a delivery plan with good transportation efficiency even when other delivery items are loaded and delivered to the next delivery destination. is there.

  In order to solve the above-described problem, a delivery plan creation method according to the first aspect of the present invention is based on one or more delivery sources, one or more delivery destinations, delivery items, transportation facilities, and constraints on these. An initial plan creation process for creating a delivery plan; a plan candidate creation process for modifying the initial delivery plan to create a delivery plan candidate; an evaluation value calculation process for obtaining an evaluation value for evaluating the delivery plan candidate; It is characterized in that it has a plan adoption judgment process in which the evaluation values of the delivery plan candidates are compared and the delivery plan candidate having the best evaluation value is used as the final delivery plan.

  The invention according to claim 2 relates to the delivery plan creation method according to claim 1, and based on the delivery plan candidates, the time for setting the arrival time and the departure time at each delivery source and each delivery destination. It has a management process.

  The invention according to claim 3 relates to the delivery plan creation method according to claim 1 or 2, wherein the initial delivery plan and the delivery plan candidate are a transport agency mark and date indicating a boundary between the transport agency and the transport agency. A date mark representing the boundary of the order is used, and it is created for each of the above-mentioned transportation facilities in charge of delivery of each order.

  The invention according to claim 4 relates to the delivery plan creation method according to claim 3, wherein the initial delivery plan and the delivery plan candidate have a first value related to the order number assigned to each order. In the delivery product stack, the second value associated with the order number represents a wholesale of the delivery product corresponding to the delivery.

  The invention according to claim 5 relates to the delivery plan creation method according to any one of claims 2 to 4, wherein a mesh point is set for each predetermined distance in a delivery target area of the delivery product, and any two A mesh point-to-mesh information database that pre-calculates the travel distance and travel time of the transport between the mesh points is included, and in the time management process, the mesh point-to-mesh information database is referred to and any of the delivery target areas The travel distance and the travel time of the transport between the delivery source and the delivery destination are calculated.

  The invention according to claim 6 relates to the delivery plan creation method according to claim 5, wherein the predetermined distance is different between a part of the delivery target area and another area. Yes.

  The invention according to claim 7 relates to the delivery plan creation method according to any one of claims 1 to 6, and in the plan candidate creation process, the evaluation value calculation process, and the plan adoption determination process, a simulated annealing process is performed. It is characterized by the use of a ring method.

  In addition, the delivery plan creation device according to the invention described in claim 8 is an initial delivery plan based on the input one or more delivery sources, one or more delivery destinations, delivery items, transport facilities, and constraint conditions related thereto. Initial plan creation means for creating a plan candidate creation means for correcting the initial delivery plan to create a delivery plan candidate, evaluation value calculation means for obtaining an evaluation value for evaluating the delivery plan candidate, The present invention is characterized by comprising a plan adoption determining means for comparing the evaluation values of the delivery plan candidates and setting the delivery plan candidate having the best evaluation value as the final delivery plan.

  The invention according to claim 9 relates to the delivery plan creation device according to claim 8, and the time for setting the arrival time and departure time at each delivery source and each delivery destination based on the delivery plan candidate. It is characterized by having management means.

  The invention according to claim 10 relates to the delivery plan creation device according to claim 8 or 9, wherein the initial delivery plan and the delivery plan candidate are a transportation agency mark and a date representing a boundary between the transportation agency and the transportation agency. A date mark representing the boundary of the order is used, and it is created for each of the above-mentioned transportation facilities in charge of delivery of each order.

  The invention according to claim 11 relates to the delivery plan creation device according to claim 10, wherein the initial delivery plan and the delivery plan candidate have a first value related to the order number assigned to each order. In the delivery product stack, the second value associated with the order number represents a wholesale of the delivery product corresponding to the delivery.

  According to a twelfth aspect of the present invention, there is provided the delivery plan creation device according to any one of the ninth to eleventh aspects, wherein a mesh point is set for each predetermined distance in a delivery target area of the delivery product, and any two A mesh point-to-mesh information database that pre-calculates the travel distance and travel time of the transport between the mesh points is provided, and the time management means refers to the mesh point-to-mesh information database, and any of the delivery target areas The travel distance and the travel time of the transport between the delivery source and the delivery destination are calculated.

  The invention according to claim 13 relates to the delivery plan creation device according to claim 12, characterized in that the predetermined distance is different between a part of the delivery target area and another area. Yes.

  The invention according to claim 14 relates to the delivery plan creation device according to any one of claims 8 to 13, wherein the plan candidate creation means, the evaluation value calculation means, and the plan adoption judgment means are simulated annealing. It is characterized by the use of a ring method.

  According to a fifteenth aspect of the present invention, there is provided a delivery plan creation program that causes a computer to realize the function according to any one of the first to fourteenth aspects.

  According to a sixteenth aspect of the present invention, a physical distribution system includes the delivery plan creation device according to any one of the eighth to fourteenth aspects.

  As described above, the present invention provides an initial plan creation process for creating an initial delivery plan based on one or a plurality of delivery sources, one or a plurality of delivery destinations, a delivery item, a transport agency, and constraints related thereto, A plan candidate creation process that modifies a delivery plan to create a delivery plan candidate, an evaluation value calculation process that calculates an evaluation value for evaluating the delivery plan candidate, and an evaluation value of each delivery plan candidate are compared and evaluated most. And a plan adoption determination process in which a good delivery plan candidate is used as a final delivery plan. Therefore, when delivering a delivery item from a plurality of delivery sources to a plurality of delivery destinations, even when a delivery item is wholesaled at each delivery destination and another delivery item is loaded and delivered to the next delivery destination, the transportation efficiency is high. A delivery plan can be created.

FIG. 1 is a block diagram showing a configuration of a delivery plan creation device 1 according to an embodiment of the present invention. The delivery plan creation apparatus 1 in this example includes a control unit 2, an operation unit 3, a display unit 4, a communication control unit 5, and a storage unit 6.
The control unit 2 includes a CPU (Central Processing Unit) and the like, and executes a delivery plan creation process, a delivery plan correction process, and the like based on various programs stored in the storage unit 6 to control the entire apparatus. That is, for example, taking the delivery plan creation process which is a feature of the embodiment of the present invention as an example, the delivery plan creation program is read from the storage unit 6 and read into the control unit 2 to control the operation of the control unit 2. Control. When the delivery plan creation program is activated, the control unit 2 executes delivery plan creation processing under the control of the delivery plan creation program.

  The delivery plan creation process includes a data conversion process, a geographic data creation process, an initial plan creation process, a plan candidate creation process, a time management process, an evaluation value calculation process, and a plan adoption determination process. The data conversion process is a process for converting the data format of order data and other data (described later) supplied from the host computer 11 (see FIG. 2), which will be described later, into a data format suitable for use in the initial plan creation process or the like. is there. The geographic data creation process is a process of creating geographic data used when executing the time management process and the evaluation value calculation process. As will be described later, the geographic data is composed of point code 1, point code 2, general time, general distance, high speed time, and high speed distance. The initial plan creation process is a process for creating an initial delivery plan based on order data and other data supplied from the host computer 11. Here, an order refers to a delivery request relating to a delivery item to be delivered, and order data refers to data relating to the delivery item. The plan candidate creation process is a process for creating a delivery plan candidate by correcting the initial delivery plan. The time management process is a process for setting the arrival time, work start time, and departure time at each delivery source and each delivery destination based on the delivery plan candidates. The evaluation value calculation process is a process for obtaining an evaluation value for evaluating the delivery plan candidate. The plan adoption determination process is a process for comparing the evaluation values of the delivery plan candidates and setting the delivery plan candidate having the best evaluation value as the final delivery plan. The delivery plan correction process is a process for an operator who has viewed the delivery plan created by executing the delivery plan creation process and displayed on the display unit 4 to correct the delivery plan.

  The operation unit 3 includes a keyboard including a numeric keypad, an enter key, or a function key, a pointing device such as a mouse, a touch pad, or a pen device. The display unit 4 includes a CRT display, a liquid crystal display (LCD), an electroluminescence (EL) display, a plasma display panel (PDP), or the like. As shown in FIG. 2, the communication control unit 5 performs data communication with the host computer 11 via a cable.

  The storage unit 6 includes a semiconductor memory such as a ROM, a RAM, or a flash memory, an FD drive to which an FD (floppy (registered trademark) disk) is mounted, an HD drive to which an HD (hard disk) is mounted, and an MO (optical MO disk drive to which a magnetic disk is mounted, or CD (compact disk) -ROM, CD-R (Recordable), CD-RW (ReWritable), DVD-ROM, DVD-R, DVD-RW, etc. CD / DVD drive. The storage unit 6 stores various programs to be executed by the control unit 2 in advance, and is used for work when the control unit 2 executes the various programs, and stores various data.

  FIG. 2 is a block diagram showing a configuration of a physical distribution system to which the delivery plan creation device 1 is applied. The physical distribution system of this example is particularly suitable for use in filling and delivering petroleum, edible oil, natural gas, etc. in industrial containers such as drums, pails, gas cylinders and the like. The physical distribution system of this example includes a host computer 11 to which the delivery plan creation device 1 is connected, a plurality of shipper clients 12, a plurality of shipping company clients 13, and a network 14. Further, in this example, it is assumed that the transportation system used by the shipping company is a truck, and not only the outbound flight from the shipping company to the delivery destination via the shipper but also the delivery destination (in some cases, a new shipper. The same applies hereinafter. )) And return flights to other shipping destinations to return to the shipping company.

  The host computer 11 includes a control unit, a storage unit, a display unit, an operation unit, and a communication control unit. The control unit includes a CPU and the like, and executes basic information management processing, financial resource grasping processing, delivery instruction processing, results management processing, accounting management processing, and the like based on various programs stored in the storage unit. Control the whole. The basic information management process is a process for managing basic information of the shipper and the delivery company. Financial resources grasping processing includes industrial containers (new cans) that are newly produced and delivered to shippers, industrial containers (original cans) that are currently used in logistics, and trucks that can be used for delivery at each shipping company This is a process for grasping financial resources such as a truck that can be dispatched. The delivery instruction process is a process of transmitting data related to a delivery instruction sheet instructing delivery of each delivery item to be delivered by each truck belonging to the transportation company in response to a request from the transportation company client 13.

  The results management process confirms delivery completion registration data indicating that delivery of a delivery sent from the shipping company client 13 has been completed, manages the fare related to delivery of each delivery, This is a process of transmitting completion report data indicating that the delivery has been completed to the shipper client 12 and managing the delivery results of the delivery items at each shipping company. In the accounting management process, the shipper client 12 is charged for the fare generated by the delivery of each delivery item, and the shipper client 12 and the shipping company client 13 are charged for the management fee necessary for managing the logistics system. Or paying the freight paid by the shipper for the delivery of each delivery item to the shipping company client 13.

  The storage unit is ROM, RAM, or flash memory, FDD with FD attached, HDD with HD attached, MO disk drive with MO disc attached, CD-ROM, CD-R, CD-RW, DVD A CD / DVD drive or the like to which a ROM, DVD-R, DVD-RW, or the like is mounted. The storage unit stores in advance various programs to be executed by the control unit, and is used for work when the control unit executes the various programs, and stores various data. The display unit is composed of a CRT display, LCD, EL display, PDP, or the like. The operation unit includes a pointing device such as a keyboard and a mouse. The communication control unit performs data communication with the shipper client 12 and the shipping company client 13 via the network 14.

  The shipper client 12 includes a control unit, a storage unit, a display unit, an operation unit, and a communication control unit. The shipper client 12 is installed in the shipper's head office, branch office, or sales office. The control unit is composed of a CPU or the like, and executes delivery item registration / change processing, progress status inquiry processing, delivery report reception processing, invoice reception processing, accounting processing, and the like based on various programs stored in the storage unit. . The delivery item registration / change process is a process for accessing the host computer 11 and registering a delivery item or changing its registration. The progress status inquiry process is a process of accessing the host computer 11 and inquiring the progress status of the requested delivery. The delivery report reception process is a process of receiving a delivery report that is accessed from the host computer 11 and reports that the requested delivery has been delivered to a desired delivery destination. The bill receiving process is a process of receiving a bill that is accessed from the host computer 11 and charges a fare generated by the delivery of the requested delivery item or a management fee necessary for managing the physical distribution system. The accounting process is a process of accessing the host computer 11 and paying the fare and the management fee.

  The storage unit is ROM, RAM, or flash memory, FDD with FD attached, HDD with HD attached, MO disk drive with MO disc attached, CD-ROM, CD-R, CD-RW, DVD A CD / DVD drive or the like to which a ROM, DVD-R, DVD-RW, or the like is mounted. The storage unit stores in advance various programs to be executed by the control unit, and is used for work when the control unit executes the various programs, and stores various data. The display unit is composed of a CRT display, LCD, EL display, PDP, or the like. The operation unit includes a pointing device such as a keyboard and a mouse. The communication control unit performs data communication with the host computer 11 and the like via the network 14.

  The shipping company client 13 includes a control unit, a storage unit, a display unit, an operation unit, a printing unit, and a communication control unit, and is installed at the headquarters, branch office, or sales office of the shipping company. ing. The control unit consists of a CPU and the like, and based on various programs stored in the storage unit, track registration / change processing, delivery instruction / invoice reception processing, delivery completion registration processing, invoice reception processing, accounting processing, etc. Execute. The track registration / change process is a process for accessing the host computer 11 and registering a vehicle dispatchable track or changing its registration. In the delivery instruction / delivery receipt reception process, the host computer 11 is accessed, and a delivery instruction for instructing delivery of a delivery assigned to each truck belonging to the shipping company and an invoice for each delivery are received. In addition, the printing unit prints each of them. The delivery completion registration process is a process of accessing the host computer 11 and registering that the delivery of each delivery item has been completed. The bill receiving process is a process of receiving a bill that is accessed from the host computer 11 and charges a management fee necessary for managing the physical distribution system. The accounting process is a process that is accessed from the host computer 11 and receives the fare paid from the shipper for the delivery of each delivery item, or accesses the host computer 11 and pays the management fee.

  The storage unit is ROM, RAM, or flash memory, FDD with FD attached, HDD with HD attached, MO disk drive with MO disc attached, CD-ROM, CD-R, CD-RW, DVD A CD / DVD drive or the like to which a ROM, DVD-R, DVD-RW, or the like is mounted. The storage unit stores in advance various programs to be executed by the control unit, and is used for work when the control unit executes the various programs, and stores various data. The display unit is composed of a CRT display, LCD, EL display, PDP, or the like. The operation unit includes a pointing device such as a keyboard and a mouse. The communication control unit performs data communication with the host computer 11 and the like via the network 14.

The network 14 includes, for example, a local area network (LAN), a wide area network (WAN), the Internet, and the like. The network 14 is wired or wirelessly connected to the host computer 11, the shipper client 12, and the shipping company client 13 by, for example, a twisted pair cable, a coaxial cable, an optical fiber cable, or the like.
There are many security problems in data communication via the network 14, but countermeasures (for example, firewalls and authentication processing) are not directly related to the present invention and are not particularly mentioned.

  Next, the configuration and structure of various data used in this embodiment will be described. FIG. 3 shows an example of the configuration of order data supplied from the host computer 11 to the delivery plan creation apparatus 1. Order data includes order number, delivery source code, delivery destination location code, loading date (from), loading time (from), loading date (to), loading time (to), delivery date (from), delivery time (from ), Delivery date (to), delivery time (to), can type code, product type code, contents code, number, car number designation / transport company designation. The order number is a number assigned to each order data, and is, for example, 1 to 100. The delivery source code is a code assigned to a point where the truck of the shipping company should load the delivery product, such as the consignor's head office, branch office, or sales office. The delivery destination location code is a code given to a point where the delivery product should be delivered.

  The loading date (from) and the loading date (to) are the date when the delivery product should be loaded into the truck. If there is a large amount of the delivery product and it cannot be loaded in one day, the loading date should be started. As a loading date (from), a loading date (to) is set as the date of loading. On the other hand, if the delivery of the delivery item can be completed in one day, only the loading date (from) is set. The loading time (from) and the loading time (to) are set as the loading start time (from) when either or both of the loading start time and the loading end time are set according to the request of the shipper. However, the loading time (to) is set as the loading end time.

  The delivery date (from) and the delivery date (to) are the dates on which the delivery products should be delivered from the truck at the delivery destination. The delivery date (from) is set as the date on which the delivery should be terminated. On the other hand, when delivery of a delivery item can be completed in one day, only the delivery date (from) is set. The delivery time (from) and the delivery time (to) are set as the delivery start time (from from) when either or both of the delivery start time and the delivery end time are set according to the request of the delivery destination. ) Is set as the delivery end time.

  The can type code is a code assigned to each type of industrial container, that is, a drum can, an 18 liter (18 L) can, a pail can, and a gas cylinder. The product type code is a code assigned to each product type, for example, petroleum, lubricating oil, chemical product, adhesive, paint, wax, cosmetics, pharmaceuticals, edible oil, juice, gas and the like. The content code is a code assigned to each specific content filled in an industrial container, for example, heavy oil, light oil, engine oil, caustic soda, ketone, sesame oil, natural gas, LP gas, etc. Applicable to industrial containers (filled cans) filled with contents. The number is the number of delivery items to be delivered. The vehicle number designation / transport company designation is set when a transport company to which a delivery item is to be delivered and a truck belonging to the transport company are designated, and a maximum of 10 cases can be set.

  FIG. 4 shows an example of the structure of the track data supplied from the host computer 11 to the delivery plan creation apparatus 1. Track data includes window store company code, transportation company code, working company code, car number, Joban / Limited division, base factory code, long distance sign, tonnage, vehicle type, size, maximum number of loads, preload type Code, preload contents code, desired start date, desired end date, desired start location, desired end location, usable time 1 (from), usable time 1 (to), usable It consists of a time 2 (from), a usable time 2 (to), a preloading sign, and a usability.

  The contact store company code is a code assigned to a store serving as a contact among the head office, branch office, or sales office of the transportation company to which the truck belongs. The transportation company code is a code assigned to the transportation company to which the truck belongs. The company code is a code assigned to the head office, branch office or sales office of the transport company to which the truck belongs. The vehicle number is a multi-digit number (in this case, four digits) assigned to the truck. The Job / Limitation category indicates whether the truck driver is a permanent employee or has a limited employment period. The base factory code is a code given to the shipper's factory where the truck is mainly responsible for delivery. The long distance impossible sign is set when the truck cannot be transported for a long distance due to predetermined circumstances.

  The tonnage is the weight that can be loaded by the truck, the vehicle type is the type of the truck, the size is the total height / length / width of the truck, and the maximum number is the maximum number of industrial containers that the truck can load. . The pre-load product type code is a code given to the product type that was filled in the industrial container delivered last time by the truck. The pre-load content code is a code assigned to the content filled in the industrial container delivered last time by the truck. The desired start date of use is the date on which the use of the truck is desired. The desired end date of use is the date on which the end of use of the truck is desired. The desired use start place is a place where the truck driver desires to start using the truck when the employment period is limited. The end-of-use desired place is a place where the truck driver wishes to end use of the truck when the employment period is limited.

  The available time 1 (from) is the first start of the time in which the track can be used. The available time 1 (to) is the first end of the time in which the track can be used. The available time 2 (from) is the second start of the time that the track can be used. The available time 2 (to) is the second end of the time in which the track can be used. The preloading sign indicates that a delivery product cannot be loaded (loaded) on the truck on the day before the delivery date. The availability indicates whether or not the track can be used.

  FIG. 5 shows an example of the configuration of location data supplied from the host computer 11 to the delivery plan creation apparatus 1. The location data is composed of a point code, an operation start time, an operation end time, a prefecture code, a city code, a track limit for entry, a loading work time, and a wholesale work time. The point code is a code given to the delivery source or delivery destination of the delivery item. The operation start time and the operation end time are the time when the factory or sales office of the delivery source or the delivery destination starts the operation and the time when the operation ends, both of which can be set for each type of industrial container. The prefecture code and the city code are codes assigned to the prefecture and city where the delivery source or delivery destination exists, respectively. The restrictions on the trucks that can be entered are the tonnage, the overall height, the overall length, and the overall width of the truck that can enter the delivery source or delivery destination. The loading operation time is the time taken to load the delivery item on the truck at the delivery source. The wholesale work time is the time taken to wholesale the delivery product from the truck at the delivery destination. Both the loading operation time and the wholesale operation time can be set for each kind of industrial container.

  FIG. 6 shows an example of the configuration of parameter data used when the delivery plan creation device 1 executes an initial plan creation process described later. The parameter data includes a simulation execution point and a simulation target date. The simulation execution point indicates a factory for producing a new can, a shipper's head office, a branch office, a sales office, or the like (hereinafter referred to as an execution point when collectively referred to) as a target for creating a delivery plan. This is because a delivery plan is created for each execution point. The conditions necessary for creating a delivery plan include conditions common to all execution points and conditions specific to each execution point. Therefore, in order to reflect conditions specific to each execution point in the creation of the delivery plan, it is set for which execution point the delivery plan is created using the parameter data.

  In FIG. 6, the simulation target date indicates a reference date for creating a delivery plan. In order to determine when the control unit 2 should load the delivery item on the truck at the execution point. Use. That is, for example, when there is a delivery product on the delivery date three days after the date of creating the delivery plan, and the simulation target date is set to one day after the date of creating the delivery plan (the day after the execution date), the control unit No. 2 selects an appropriate date as a date for loading a delivery item in consideration of the holiday of the delivery source, etc., from the day before the simulation target date to 3 days after the delivery date. Therefore, when the operator changes the simulation target date, for example, a delivery plan can be created in the case where the delivery date is loaded 6 days after the delivery plan is created and the delivery date is 7 days later. The parameter data is input by the operator operating the operation unit 3 by displaying a screen prompting the input of the simulation execution factory and the simulation target date on the display unit 4 when the control unit 2 starts the delivery plan creation program. As a result, the data may be stored in a predetermined area of the storage unit 6, or set in advance on the host computer 11 side and transferred together with the order data or the like to be stored in the storage unit 6 after data conversion. It may be stored in this area.

  FIG. 7 is a diagram illustrating an example of the configuration of geographic data used when executing a time management process and an evaluation value calculation process, which will be described later, in the delivery plan creation apparatus 1. The geographic data is composed of point code 1, point code 2, general time, general distance, high speed time, and high speed distance. The point code 1 is a code given to the point that is the delivery source of the delivery item. The point code 2 is a code given to the point that is the delivery destination of the delivery item. The general time is a travel time required when a general road is used for delivery. The general distance is a distance traveled by the truck when a general road is used for delivery of the delivery goods. The high-speed time is a travel time required when a highway is used for delivery. The high-speed distance is the distance traveled by the truck when the expressway is used for delivery.

  FIG. 8 is a diagram showing an example of a configuration of a delivery plan created by executing a time management process described later in the delivery plan creation device 1. As shown in FIG. 8, the delivery plan is created for each car number of a truck belonging to each shipping company. The order number, point, loading / unloading classification, arrival time, work start time, departure time It is composed of The order number indicates the order number constituting the order data. The point indicates a point where a delivery product is loaded or wholesaled. The loading / unloading category indicates whether the delivery product is loaded or unloaded. The arrival time indicates the time when the truck arrives at the point. The work start time indicates the time at which the truck starts loading work or wholesale work at the above point. The departure time indicates the time at which the truck departs from the point after the work is completed.

  FIG. 9 is a diagram illustrating an example of a constraint condition when executing a plan candidate creation process and an evaluation value calculation process, which will be described later, in the delivery plan creation apparatus 1. Restrictions include loading order, car number designation / shipping company designation, anti-shipper exclusive vehicle, long-distance availability, truck availability on the date of dispatch, loadable cans and varieties for each vehicle type, delivery area, exclusive delivery destination Car, location restrictions, loading date / loading time designation, delivery date / delivery time designation, truck load / number of loading, no loading the previous day, truck usable time, priority vehicle, dummy truck, working time by location It consists of the direction at the time of stacking, the constraint at the time of stacking, the preloading constraint, the order of patrol, the fare, the share, the average unit price, the loading rate, the number of unallocated orders, the number of used trucks, the mixed loading of different types, and the ferry.

  Among these, loading order, car number designation / transport company designation, exclusive car, long distance availability, truck availability on the day of delivery, loadable cans and varieties by car type, delivery area, exclusive car, by location The entry restriction constitutes a constraint condition (first constraint condition) when executing a plan candidate creation process described later in the delivery plan creation device 1. On the other hand, loading date / loading time designation, delivery date / delivery time designation, truck loading capacity / number of loading, previous day loading impossible, truck usable time, priority vehicle, dummy truck, operation time by location, direction at the time of loading , Loading restrictions, preload restrictions, order of travel, freight, share, average unit price, loading rate, number of unallocated orders, number of used trucks, mixed loading of different types, ferry, plan evaluation processing described later in the delivery plan creation device 1 The constraint condition (2nd constraint condition) at the time of performing is comprised.

  When loading industrial containers with different can types, varieties, contents, etc. on a single truck, delivery products can only be wholesaled in the reverse order of loading, so the loading order is based on this order. It is a constraint condition. Car number designation / transport company designation is a car number designation / transport company designation is a restriction condition based on this when there is a designation of the shipping company to which the delivery item should be delivered and the truck belonging to that shipping company. It is. A shipper's exclusive vehicle is a constraint based on this when there is a truck of a transportation company that is exclusively contracted with the shipper. Long distance availability is a constraint based on whether or not a truck can carry long distances. The availability of the truck on the day of dispatch is a constraint based on whether the truck can be used on the day of dispatch. The loadable can type / variety for each vehicle type is a restriction condition based on this in the case where the can type and product type of industrial containers that can be loaded are set for each vehicle type of the truck.

  The delivery possible area is a restriction condition based on this when an area where the truck can deliver the delivery product is set. The exclusive vehicle for the delivery destination is a restriction condition based on this when there is a truck of a transport company that is exclusively contracted with the delivery destination. The entry restriction for each place is a constraint condition based on this when the entry of the truck is restricted for each delivery source or delivery destination. The loading date / loading time specification is a constraint condition based on this when the shipper has specified the loading date / time of the delivery item. The delivery date / delivery time designation is a restriction condition based on this when the delivery destination designates the delivery date / time of the delivery product. The truck loading capacity / loading capacity is a constraint condition based on this when the maximum loading capacity of each truck and the maximum loading capacity of industrial containers are set. Unloading on the previous day is a restriction condition based on this in the case where a delivery product cannot be loaded on the truck on the day before the delivery date.

  The track usable time is a restriction condition based on this when a usable time of the track is set. The priority vehicle is a constraint condition based on this when a truck to be preferentially delivered for a specific delivery item is set. The dummy track is a constraint condition based on this when a virtual track that is not an actual track is set. The operation time for each place is a constraint condition based on this when the operation time of the delivery source or the delivery destination is set. The direction at the time of stacking is a restriction condition based on this when a delivery item to be delivered to a plurality of delivery destinations is mounted on one truck and a delivery destination direction is set. The restrictions at the time of stacking are restrictions based on this when there is a restriction other than the direction at the time of stacking when a delivery product to be delivered to a plurality of delivery destinations is mounted on one truck. It is a condition.

  The preload constraint is a constraint condition based on this when a constraint is set based on a delivery product delivered by each truck last time. For example, if the delivery product delivered last time is a chemical product such as powerful medicine, but the delivery product to be delivered this time is edible oil or the like, there is a restriction due to that. The circulation order is a constraint condition based on this when a plurality of delivery destinations are set in advance so as to visit a predetermined route. The fare is a constraint condition based on a fare constraint when delivering a delivery item, for example, when a fare that does not assume the use of an expressway is set. Share is a constraint based on this when it is set so that a specific shipping company does not handle the delivery of one shipper exclusively. The average unit price is a constraint condition based on this when an average value of a fare paid by one shipper for delivery of one delivery item is set.

  The loading rate is a constraint condition based on this when a loading rate of a delivery product mounted on each truck is set. The number of unassigned orders is a constraint condition based on this when the number of order data that is not assigned among the order data is set. The number of used trucks is a constraint based on this when the number of trucks used in each shipping company is set. Different types of mixed loading is a constraint condition based on this when it is set whether or not it is possible to load different types of conveyed products on one truck. Part of this mixed-mixed loading is also used in the plan candidate creation process. The ferry is a constraint based on this when each truck is set to be able to use the ferry and is used in connection with posting the delivery distance and delivery time.

  In this embodiment, the delivery plan is represented by a one-dimensional data array (hereinafter referred to as plan array data). FIG. 10 is an example of planned arrangement data when the number of orders is 4 and the number of trucks is 2. In FIG. 10, “0” is a track mark representing a boundary between tracks, and an array of elements sandwiched between “0” and “0” is data of one track. “−1” is a date mark indicating a date boundary, and each element on the left side of the drawing from “−1” relates to a day (x day) immediately before a target date for creating a delivery plan. Each element on the right side of the figure from “−1” relates to a target day (x + 1 day) on which a delivery plan is created. Numbers greater than or equal to “1” and less than or equal to the number of orders indicate the value corresponding to the order number that constitutes the order data (in the case of “loading”) and the value obtained by adding all the orders to the order number (in the case of “wholesale”) .

  In the example of FIG. 10, the numbers “1, 4, −1, 8, 5” are consecutive between the first “0” from the left and the second “0” from the left. For the first truck, the delivery of order number <1> was loaded on day x, the delivery of order number <4> was loaded, and the delivery of order number <4> was wholesaled on (x + 1) day. Thereafter, it indicates that the delivery of the order number <1> is wholesaled. Next, because the numbers “−1, 2, 6” are continuous between the second “0” from the left and the third “0” from the left, , (X + 1) days after delivery of the order number <2> is loaded, the order number <2> delivery is wholesaled. To the right of the third “0” from the left, “3, 7” continues to indicate that the “stack” and “wholesale” of the delivery of order number <3> is unallocated. Yes. It should be noted that the specific dates and times of the “loading” and “wholesale” of the delivery products are set by a time management process described later.

  Next, a delivery plan creation process executed by the delivery plan creation device 1 configured as described above will be described. When the operator operates the operation unit 3 constituting the delivery plan creation device 1 to instruct creation of a delivery plan, the control unit 2 first proceeds to the process of step SP1 shown in FIG. 11 and executes a data conversion process. That is, the control unit 2 accesses the host computer 10 and requests transfer of order data, track data, location data, and constraint conditions. When order data, track data, location data, and constraint conditions are supplied from the host computer 11, the control unit 2 converts each data format of the order data, track data, location data, and constraint conditions in an initial plan creation process to be described later. After conversion to a data format suitable for use, the data is stored in a predetermined area of the storage unit 6. And the control part 2 progresses to step SP2.

In step SP2, the control unit 2 executes the geographic data creation process, and then proceeds to step SP3. This geographic data creation process will be described later. In step SP3, the control unit 2 executes an initial plan creation process. That is, the control unit 2 generates an initial delivery plan (initial plan array data) based on the order data, the track data, and the location data whose data format has been converted, and the parameter data created by the operator operating the operation unit 3. ), The process proceeds to step SP4.
Next, more detailed processing of the initial plan creation processing will be described with reference to the flowchart shown in FIG. First, the control unit 2 proceeds to the process of step SA1 shown in FIG. 12, reads track data from a predetermined area of the storage unit 6, and sets 1 to a value corresponding to the number of tracks based on the read track data. After setting “0” (track mark) corresponding to the added value to create new planned array data, the process proceeds to step SA2. FIG. 13 (a) shows an example of planned array data in which the above processing has been executed. In this example, since the number of tracks is two, in FIG. 13A, three “0” s are arranged. In FIG. 13A, the leftmost “0” is a track mark (first track mark) indicating that it is the top, and the first track is from “0” at the left end to the second “0” from the left. Data from the second “0” from the left to the third “0” from the left to the data immediately before the third “0” from the left. The data from the third “0” to the right end from the left is the track. Indicates an unassigned order.

  In step SA2, the control unit 2 reads the simulation target date from the parameter data stored in the predetermined area of the storage unit 6, and reads the order data stored in the predetermined area of the storage unit 6, Get the latest delivery date that composes the order data. The reason for obtaining the latest delivery date is as follows. In other words, in this example, a delivery plan is created for a delivery product whose delivery date is the day after the date on which the delivery plan is created, but on Saturdays and Sundays, the execution point is a closed day, and the shipping company does not dispatch vehicles. Therefore, when the date on which the delivery plan is created is Friday, it is necessary to create a delivery plan for delivery items assuming that Saturday, Sunday, and Monday are delivery dates. In this case, the date for which the delivery plan for the delivery product is created is three days. Therefore, when there are a plurality of dates for which the delivery plan for the delivery product is created, the latest delivery date is acquired. Next, the control unit 2 sets date marks for the number of days from the read simulation target date to the latest delivery date in the planned arrangement data for each track. And the control part 2 progresses to step SA3. FIG. 13B shows an example of the planned arrangement data in which the above processing is executed. In this example, it is assumed that the date on which the delivery plan is created is not Friday, and the date mark “−1” for two days on the simulation target date and the next day is set for each track.

  In step SA3, the control unit 2 reads out the order data stored in a predetermined area of the storage unit 6, values corresponding to all the order numbers constituting the order data, and a value obtained by adding all the order numbers to the order numbers Are set as “stack” and “wholesale” in the column where the track of the planned sequence data is not allocated, and after creating the initial planned sequence data, the process returns to the main routine shown in FIG. 11 and proceeds to step SP4. FIG. 13 (c) shows an example of planned arrangement data in which the above processing is executed. In this example, “1, 5, 2, 6, 3, 7, 4, 8” are consecutive to the right from the fourth “0” from the left, so that the order numbers <1> to <4>. This indicates that the “shipment” and “wholesale” of the delivery items of the item are not allocated.

  In step SP4 shown in FIG. 11, the control unit 2 executes a plan candidate creation process. That is, the control unit 2 corrects the initial plan arrangement data created in step SP3 or the delivery plan candidate (plan candidate arrangement data) previously created in the plan candidate creation process in step SP4, that is, is mounted on the same track. After creating a new delivery plan candidate (plan candidate array data) by changing the loading / unloading order of the delivery items that are being loaded or by changing the delivery items that were loaded on one track to be loaded on other tracks The process proceeds to step SP5.

  Next, a more detailed process of the plan candidate creation process will be described with reference to the flowchart shown in FIG. First, the control unit 2 proceeds to the process of step SB1 shown in FIG. 14, and uses a random number to correspond to an order number to be moved (hereinafter referred to as a movement target order), that is, “stack”. After selecting, the process proceeds to step SB2. FIG. 15A is an example of plan candidate array data created by the plan candidate creation process of step SP4 last time. On the other hand, FIG. 6B shows that the element “2” is hatched, and “2” is selected as the movement target order in the process of step SB1.

  In step SB2, the control unit 2 selects, using a random number, the insertion destination of the “stack” of the movement target order selected in step SB1, moves the “stack” of the movement target order to the insertion destination, Proceed to SB3. In this case, the delivery plan is clarified by referring to the order data, the track data, and the constraint conditions stored in a predetermined area of the storage unit 6 and moving the “stack” of the order to be moved to the insertion destination. If it cannot be executed, the selection of the insertion destination is performed again. The reason for re-referencing the order data and re-selecting the insertion destination in this way is to avoid unnecessary evaluation of a delivery plan that is clearly unexecutable in the plan evaluation process described later. FIG. 15C shows that the elements “2” and “4” are hatched, and the element “4” is selected as the insertion destination of the movement target order “2”. FIG. 15D shows that the adjacent elements “2” and “4” are hatched, and the movement target order “2” is inserted on the right side of the element “4”.

  In step SB3, the control unit 2 moves the “loading” of the movement target order by the process of step SB2, and as a result, the “wholesale” corresponding to the “loading” of the movement target order violates the loading order. After searching for the insertion destination of “Wholesale” of the target order and moving the “Wholesale” of the movement target order to the insertion destination, the process returns to the main routine shown in FIG. 11, and proceeds to Step SP5. In this case, if there are a plurality of insertion destination candidates searched, the insertion destination is selected using a random number. In this case, as a result of moving the “stack” of the order to be moved by the process of step SB2, the element “6” and the order to be moved “2” (“stack”) are hatched in FIG. In this way, the movement target order “2” (“loading”) is inserted on the right side of the corresponding “wholesale” element “6”, which violates the loading order. Therefore, when the insertion destination of the element “6” that is “wholesale” of the movement target order “2” is searched in the process of step SB3, as shown in FIG. 15F, the second date mark “−1” from the left is displayed. Only the right side of "is searched for as an insertion destination. Therefore, as shown in FIG. 15G, the element “6”, which is “wholesale” of the movement target order “2”, is moved to the right side of the second date mark “−1” from the left.

  In step SP5 shown in FIG. 11, the control unit 2 executes a time management process. That is, the control unit 2 creates geographic data, and also creates a delivery plan candidate (plan candidate array data) created in step SP4, order data, track data, and location data converted in data format, and step SP2. After setting the arrival time, work start time, and departure time at each delivery source and each delivery destination based on the geographical data, the process proceeds to step SP6.

Next, a more detailed process of the time management process will be described with reference to a flowchart shown in FIG. First, the control unit 2 proceeds to the process of step SC1 shown in FIG. 16, refers to the geographic data, and then proceeds to step SC2. Hereinafter, the creation process of the geographic data referred to in step SC1 will be described in detail. This geographic data creation process is executed between the data conversion process and the initial plan creation process as shown in FIG. 11 (step SP2). In order to create an optimal delivery plan that delivers deliveries reliably at a specified time in a short time and at a low cost, the travel time and travel distance of the truck between the point of delivery and the point of delivery It is essential to calculate the travel time and travel distance between any two points, but generally several hours are required. For example, when there are 100 orders, there are a delivery source and a delivery destination for each order. Therefore, as shown in FIG. 17, two arbitrary points out of 200 points (points P ₁ and P ₂ at the maximum). , Between points P ₁ and P ₃ ,... Between points P ₁ and P ₁₉₉ , between points P ₁ and P _200, and between points P ₂ and P ₃ ...,... (Between point P ₁₉₉ and point P ₂₀₀ ), that is, it is necessary to calculate travel time, travel distance, and the like for 19900 ways represented by equation (1).
₂₀₀ C ₂ = (200 × 199) / 2 = 19900 (1)
A car navigation program currently on the market takes about one second to calculate the travel time between any two relatively short distances. Therefore, if such a car navigation program is made to calculate the travel distance between any two of the 200 points, it takes 19900 seconds, that is, about 5.5 hours.

  In this regard, when the delivery source and the delivery destination are fixed, they can be calculated in advance, so that the calculation time does not matter. On the other hand, in this embodiment, both the delivery source and the delivery destination are arbitrary, and can not be calculated in advance because it is not determined until the time when the delivery plan is specifically created. Recently, since mobility and quickness are required for creating a delivery plan, it needs to be performed in a very short time, for example, several seconds.

Therefore, in this embodiment, by adopting the following methods (A) to (C), it is possible to deliver between any two points at high speed and with high accuracy despite being targeted for delivery throughout Japan. Geographic data consisting of travel time and travel distance can be calculated.
(A) A representative point (mesh point) is defined and divided (mesh division) for every predetermined distance on land in Japan, and the travel time and travel distance are calculated in advance only between each mesh point and stored in the storage unit 6 deep. The travel time and travel distance between the created mesh points are referred to as a mesh point information database.

(B) A distance between mesh points in a short-distance transportation area where a delivery item is delivered between points with a relatively short distance (a mesh point in this case is called a short mesh point) and a point with a relatively long distance The distance between mesh points in a long-distance transportation area where deliveries are delivered between them (the mesh points in this case are called long mesh points) is made different. This is for the following reason. That is, if the distance between mesh points is shortened, the travel time between the mesh points and the calculation accuracy of the travel distance are improved, but the creation time of the inter-mesh point information database is increased accordingly, which is not practical. On the other hand, when a delivery item is delivered between a user's location and a relatively short distance, and when a delivery item is delivered between a user's location and a relatively long point, Usually, the required accuracy is different. That is, when delivering a delivery item to a point that is only 10 km away from a certain point, there is a practical problem if there is an error of 10 minutes in the calculation result of the travel time, but it is a point that is 300 km away from a certain point. When delivering a delivery item, there is no practical problem even if there is an error of 10 minutes in the calculation result of the travel time. Therefore, Japan is divided into short-distance transportation areas and long-distance transportation areas, and the travel time and travel distance are calculated in advance considering only the combination of two mesh points in each of the short-distance transportation area and long-distance transportation area. By configuring the mesh point information database, the number of mesh point combinations whose travel time and travel distance should be calculated can be greatly reduced. As a result, the creation time of the mesh point information database can be greatly shortened. For example, the distance between short mesh points (short mesh size) is 5 km with the metropolitan area and the Kansai area as short distance transport areas, and the distance between long mesh points (long mesh size) is 20 km with the other areas as long distance transport areas. And FIG. 18 shows an example of dividing a part of the Tokyo metropolitan area, which is a short-distance transportation area, with short mesh points, and FIG. 19 shows an example of dividing a part of Shizuoka Prefecture, which is a long-distance transportation area, with long mesh points. FIG. In FIG. 18 and FIG. 19, the short mesh points or the long mesh points that are mechanically arranged on the sea are rearranged on the nearest land.

(C) For any two points, the travel time and travel distance between mesh points in the vicinity are read from the mesh point information database, and statistical processing is performed to obtain geographical data as a highly accurate approximate value. . That is, for example, as shown in FIG. 20, when the distance between mesh points is L, and mesh points M _{1 to} M ₄ exist around four points (target points) P that are targets for creating a delivery plan. The mesh point closest to the target point P by the straight line distance is the mesh point M ₁ . On the other hand, assuming that the target point P exists at the point P ′, the mesh point closest to the point P ′ in terms of the straight line distance is not the mesh point M ₁ but the mesh point M ₂ . Similarly, if it is assumed that the target point P exists at the point P ″, the mesh point closest to the point P ″ by the straight line distance is not the mesh point M ₁ or the mesh point M ₂ but the mesh point M ₃ . In other words, when an arbitrary target point is taken on the map where the mesh points are arranged, the distance from the arbitrary target point to the mesh point closest to the straight line distance is the accuracy when the travel distance is approximately calculated. It will be. In this case, assuming that an expected value from which an accurate travel distance can be calculated is H, the expected value H is expressed by Equation (2) using a statistical method.
H = 0.3826 × L (2)
Therefore, since the short mesh size is 5 km in the case of the short-distance transportation area, the expected value H is about 1.9 km, and in the case of the long-distance transportation area, the long mesh size is 20 km. 7.7 km, both of which are more practical than conventional methods.

As described above, if the short mesh size and the long mesh size are appropriately set, practical accuracy can be obtained as compared with the conventional method. However, in this example, several mesh points that exist near the target point are searched, and the multiple regression method, which is one of the statistical prediction methods for the distance between each mesh point and the corresponding target point, is used. To obtain a predicted value with higher accuracy. That is, first, as shown in FIG. 21, two target points P ₁ and P ₂ and four mesh points existing around each target point P ₁ and P ₂ (hereinafter referred to as surrounding mesh points). ) consider the Q ₁ to Q ₄ and R ₁ to R _4. Knowing respective coordinates of the target point P ₁ and P ₂ can be calculated travel distance and the travel time and the like between the object point P ₁ and the target point P ₂ required by time management process. Here, when the coordinates of the target point P ₁ are (x ₁ , y ₁ ) and the coordinates of the target point P ₂ are (x ₂ , y ₂ ), the distance between the target point P ₁ and the target point P ₂ is such travel distance and travel time, as a function f of the coordinate target point _{P 1 (x 1, y 1} ) and object point P ₂ of the coordinates (x _2, y _2), the formula (3).
f (x ₁ , y ₁ , x ₂ , y ₂ ) (3)

In FIG. 21, the surrounding mesh point closest to the target point P _{1 in the} straight line distance is the surrounding mesh point Q ₂ , and the surrounding mesh point closest to the target point P _{2 in the} straight line distance is the surrounding mesh point R ₃ . Here, the coordinates of the surrounding mesh point Q ₂ are (cx ₁ , cy ₁ ), the coordinates of the surrounding mesh point R ₃ are (cx ₂ , cy ₂ ), and the distance Δx ₁ from the surrounding mesh point Q _{2 in the} x direction, A point (cx ₁ + Δx ₁ , cy ₁ + Δy ₁ ) separated by a distance Δy _{1 in} the y direction and a point (cx ₂ + Δx) separated from the surrounding mesh point R ₃ by a distance Δx _{2 in} the x direction and a distance Δy _{2 in the} y direction. ₂ , cy ₂ + Δy ₂ ), the travel distance, travel time, and the like are expressed by equation (4) as in equation (3).
f (cx ₁ + Δx ₁ , cy ₁ + Δy ₁ , cx ₂ + Δx ₂ , cy ₂ + Δy ₂ ) (4)

When the first-order expansion is performed for the variation of the x coordinate and the y coordinate for the equation (4), the equation (5) is obtained.
f (cx ₁ + Δx ₁ , cy ₁ + Δy ₁ , cx ₂ + Δx ₂ , cy ₂ + Δy ₂ )
_{= F (cx 1, cy 1} , cx 2, cy 2) + a 1 · Δx 1 + b 1 Δy 1 + a 2 · Δx 2 + b 2 · Δy 2 ··· (5)
The travel distance and travel time between the two points are considered to have a strong correlation with the position coordinates due to their properties, and have a property of continuously changing, so the approximation using equation (5) is high. It can be considered to have accuracy.

In equation (5), the function f (cx ₁ , cy ₁ , cx ₂ , cy ₂ ) is a function between the surrounding mesh point Q ₂ and the surrounding mesh point R _3, and is previously obtained by the method (A) described above. Since it is calculated and stored in the storage unit 6 as a mesh point information database, it is known. Therefore, in equation (5), the four coefficients a ₁ and b ₁ , a ₂ , b ₂ are unknown. Therefore, in order to estimate and calculate these four coefficients, attention is paid to _four surrounding mesh points Q _{1 to} Q ₄ and R _{1 to} R ₄ existing around the target points P ₁ and P ₂ shown in FIG. To do. That is, the function f for any one of the surrounding mesh points Q _{1 to} Q ₄ and any one of the surrounding mesh points R _{1 to} R ₄ is calculated in advance by the above-described method (A), and is used as an information database between mesh points. It is stored in the storage unit 6. In the example of FIG. 21, the function f between two surrounding mesh points (hereinafter referred to as the surrounding mesh point function f) is all except for the function f between the surrounding mesh point Q ₂ and the surrounding mesh point R _3. There are 15 pieces. Therefore, as shown below, the unknown coefficients a ₁ and b ₁ , a ₂ , and b ₂ are estimated and calculated by the multiple regression method using the fifteen surrounding mesh point functions f.

First, in Equation (5), the difference between the left term and the first term on the right side is expressed by Equation (6) as Δf.
Δf = f (cx ₁ + Δx ₁ , cy ₁ + Δy ₁ , cx ₂ + Δx ₂ , cy ₂ + Δy ₂ )
−f (cx ₁ , cy ₁ , cx ₂ , cy ₂ )
= A ₁ · Δx ₁ + b ₁ Δy ₁ + a ₂ · Δx ₂ + b ₂ · Δy ₂ (6)
Consider the result vector ΔF shown in equation (7), where n is the number of surrounding mesh point functions f (n is a natural number). Here, the result vector ΔF is a value for any one of the travel distance and travel time of the truck. This result vector ΔF can be created using the values stored in the storage unit 6 as the mesh point information database described above.

Further, with respect to the n neighboring mesh point functions f, Δx ₁ and Δy ₁ , Δx ₂ , Δy ₂ constituting the equation (5) are arranged, and as shown in the equation (8), an n × 4 matrix X Create

The value of each element of the matrix X can be easily calculated because the coordinates of surrounding mesh points are known. Therefore, a vector (unknown coefficient vector) θ constituted by unknown coefficients a ₁ and b ₁ , a ₂ , b ₂ to be obtained is expressed by Expression (9).

The determinant shown in the equation (10) is obtained from the equations (7) to (9).
ΔF = Xθ (10)
Therefore, the most probable value Θ obtained using the least square method for the unknown coefficient vector θ is expressed by Expression (11).
Θ = (X ^T X) ⁻¹ X ^T ΔF (11)
The result of Expression (11) is an unknown coefficient (predicted value) to be obtained.
On the other hand, Δx ₁ and Δy ₁ are determined from the positional relationship between the target point P ₁ and the surrounding mesh point Q _2, and Δx ₂ and Δy ₂ are determined from the positional relationship between the target point P ₂ and the surrounding mesh point R ₃ . Accordingly, the function f (cx ₁ , cy ₁ , cx ₂ , cy ₂ ) stored as the mesh point information database in the storage unit 6, the result of the expression (11), and the expression By substituting Δx ₁ and Δy ₁ , Δx ₂ , Δy ₂ , the function f between the target point P ₁ and the target point P ₂ can be obtained.

  Next, the creation processing of the mesh point information database will be described with reference to the flowchart shown in FIG. First, the control unit 2 proceeds to the process of step ST1 shown in FIG. 22, and the above-described methods (A) and (B) are performed on the map data of Japan stored in a predetermined area of the storage unit 6 in advance. Based on the short mesh points and the long mesh points, the process proceeds to step ST2. In step ST2, the control unit 2 divides Japan into a plurality of regional blocks in stages, for example, nationwide, large-scale areas (Tohoku, Kanto, Chubu, etc.), prefectures, municipalities, etc. The travel distance and travel time between any two mesh points created in the process of step ST1 for each regional block are calculated, and the mesh point-to-mesh information database is created for each regional block. After storing in the predetermined area of 6, the series of processing ends.

  Next, more detailed processing of the geographic data creation processing will be described with reference to the flowchart shown in FIG. First, the control unit 2 proceeds to the process of step ST11 shown in FIG. 23, selects two of a plurality of target points that are targets of delivery plan creation, and then proceeds to step ST12. In step ST12, the control unit 2 selects between the mesh points of the narrowest regional block including one target point selected in step ST11 from the mesh point-to-mesh information database stored in the predetermined area of the storage unit 6. After selecting the information database, the process proceeds to step ST13. In step ST13, the control unit 2 determines whether or not both target points selected in step ST11 exist in the mesh point information database selected in step ST12. When the determination result is “NO”, the control unit 2 returns to step ST12, and is the mesh point information database stored in a predetermined area of the storage unit 6, and is the one selected in step ST11. After selecting the mesh point information database of the regional block one step wider than the previously selected regional block from those including the target point, the process proceeds to step ST13. In step ST13, the control unit 2 determines whether or not both target points selected in step ST11 exist in the mesh point information database selected in step ST12. Then, by repeating the processes in steps ST12 and ST13 described above, if there are two target points selected in step ST11 in the mesh point information database selected in step ST12, the determination result in step ST13 is “ In the case of “YES”, the control unit 2 proceeds to step ST14.

  In step ST14, the control unit 2 selects four surrounding mesh points of each of the two target points selected in step ST11 from the mesh point information database selected in step ST12, and then proceeds to step ST15. In step ST15, the control unit 2 reads data related to the eight surrounding mesh points selected in step ST14 from the mesh point information database, and based on the above prediction method, between the two target points selected in step ST11. After obtaining the function f, the process proceeds to step ST16. In step ST16, the control unit 2 stores the function f between the two target points obtained in step ST15 as geographic data in a predetermined area of the storage unit 6, and then proceeds to step ST17. In step ST17, the control unit 2 determines whether or not the above-described processing of steps ST12 to ST16 has been executed for all combinations of two points for a plurality of target points for which a delivery plan is to be created. If the determination result is “NO”, the control unit 2 returns to step ST11 and repeats the processes of steps ST11 to ST16 described above. And when the process of step ST12-ST16 mentioned above was performed about the combination of all 2 points | pieces about the some target point which is the object of delivery plan preparation, the judgment result of step ST17 will be "YES", and the control part 2 After finishing the geographic data creation process, the process returns to the main routine shown in FIG. 11 and proceeds to step SP3.

  Here, a specific example of geographic data is shown in FIG. In this embodiment, the delivery source is the MNO Chemical Shizuoka Office located in Shimizu, Shizuoka City, Shizuoka Prefecture, the JKL Warehouse Ogishima Office located in Ogishima, Tsurumi-ku, Kanagawa Prefecture, and Kachidoki, Chuo-ku, Tokyo. Suppose that it is a certain GHI warehouse Kachidoki sales office. On the other hand, it is assumed that the delivery destination is the YZ Industrial Kakegawa Factory located in Shimooka Prefecture Kakegawa City Shimo-rafter, and the PQR Chemical Fuchu Factory located in Koyanagicho, Fuchu-shi, Tokyo. As can be seen from FIG. 24, the JKL Warehouse Ogishima Sales Office and the GHI Warehouse Katachidoki Sales Office are also delivery destinations.

In step SC2 shown in FIG. 16, the control unit 2 is based on the plan candidate sequence data created in step SP4, the order data, the track data, and the location data converted in data format, and the geographic data referenced in step SC1. After setting the arrival time, work start time and departure time for the initial stack, the process proceeds to step SC3. Here, the initial stack means a “stack” that is first positioned after the track mark (first track mark) indicating the head in the plan candidate array data. First, as a premise, in this time management process, the work time Tw required for the delivery work and the wholesale work of the delivery product is expressed by Expression (12).
Tw = Tp + Ta + Ts (12)
In Expression (12), Tp is the time (actual work time) required for the operator to actually load or wholesale the delivery product, and is set separately for each delivery source and delivery destination. Ta is the time (post work time) required for the work to be performed after the delivery work or the wholesale work of the delivery product, and is set separately for each delivery source and delivery destination. Ts is a time (additional time) required for various reasons other than the actual work time Tp and the post-work time Ta, and is set separately for each delivery source and delivery destination.

The date and time for the initial stacking is set as shown below.
(I) When the right next to the first track mark is “stacking” The date of “stacking” is the day before the simulation target date constituting the parameter data. In this case, “loading” means that a delivery product is loaded on the truck on the day before the delivery date (loading), so the arrival time is uniformly set to 5 pm (17:00). In the case of “loading”, there is no need to consider the work start time, so it is not set. The departure time is set by calculating backward from the work start time of the delivery product that is the first to be shipped among the plurality of delivery products that are stacked together with the delivery product to be loaded first.
(II) When the date mark is to the right of the first track mark The “stack” date is the sum of the number of date marks on the first day of each track. Here, the first day means a date between the first track mark and the date mark positioned first. When there is a date mark between “stack” and “wholesale” corresponding thereto, the arrival time, work start time, and departure time are the same as in the case of (I) described above. On the other hand, when “loading” and the corresponding “wholesale” are on the same day, the arrival time is set to the operation start time of the delivery source, the work start time is set to the same time as the arrival time, and the departure time Sets the work start time plus work time.

In step SC3, the control unit 2 performs initial loading based on the plan candidate sequence data created in step SP4, the order data, the track data, and the location data converted in data format, and the geographic data referenced in step SC1. After the arrival time, work start time, and departure time are set later, the process returns to the main routine shown in FIG. 11 and proceeds to step SP6.
The date and time setting after the initial loading is performed as follows.
(I) When the date mark is on the left of the “loading” or “wholesale”, the arrival time is set to the time if a time is specified, otherwise the delivery source or destination (hereinafter generically) When this is done, it is called the work place.) The work start time is set to the same time as the arrival time. The departure time is set by adding the work time to the work start time.
(II) When the left side of the “loading” or “wholesale” is “loading” or “wholesale” The arrival time is the travel time from the previous work place to the work place at the departure time of the previous work place. Set to the sum. If the time is set and the arrival time is earlier than the specified time, the specified time is set as the work start time. If the arrival time is later than the specified time, the arrival time is set as the work start time. . On the other hand, when the time is not set and the arrival time is earlier than the operation start time of the work place, the operation start time of the work place is set as the work start time, and the arrival time from the operation start time of the work place If it is late, the arrival time is set as the work start time. The departure time is set by adding the work time to the work start time.
(III) When there are multiple date marks If there are “loading” and “wholesale” for the same delivery items on both sides of the date mark, the delivery items are treated as “loading”, otherwise The date and time are set using the method (I) or (II) described above.

  Hereinafter, the time management process will be described using specific data. FIG. 25 shows an example of plan candidate array data created by the above-described plan candidate creation process (step SP4) and used in this time management process (step SP5). FIG. 26 shows a specific example of order data, FIG. 27 shows a specific example of track data, and FIG. 28 shows a specific example of location data. The geographic data shown in FIG. 24 is used. First, the simulation target date, that is, the delivery date is 23 days as shown in FIG. Further, in the plan candidate array data shown in FIG. 25, the track marks appear from the left to the right in the order of the track data array shown in FIG. That is, the first truck is a truck with a car number 1432 belonging to ABC transportation, and the second truck is a truck with a car number 7253 belonging to DEF transportation.

  Next, the time setting of the track of the car number 1432 belonging to ABC transportation which is the first truck will be described. As can be seen from the plan candidate sequence data shown in FIG. 25, the truck of the car number 1432 is delivered by the delivery product of the order number <4>, that is, as shown in FIG. Perform “loading” at the office and “wholesale” at the YZ Kakegawa Factory. In the plan candidate sequence data shown in FIG. 25, since the date mark “−1” is inserted between the first track mark “0” and the “stack” of the order number <4>, the “#” of the order number <4> The loading date of “loading” is 23 days. In addition, since “loading” of order number <4> is “loading” on the day of delivery, the operation start time of the MNO Chemical Shizuoka office shown in FIG. 28 is 8:30 am (8:30). Set to arrival time and work start time.

  Next, as shown in FIG. 26, since the delivery number of order number <4> is 120, the work start time is set to 40 minutes, which is the loading time of 150 pieces of MNO Chemical Shizuoka office shown in FIG. The added 9:10 am (9:10) is set as the departure time. Next, the arrival time of “Wholesale” of order number <4> is the departure time of “loading” of order number <4> from the MNO Chemical Shizuoka office of the geographic data shown in FIG. 24 to the YZ Kakegawa factory. The travel time is set to 10:28 am, adding 1 hour and 18 minutes. As shown in FIG. 26, the delivery time of the order number <4> is designated from 7:00 am to 12:00 am (7:00 to 12:00). Since the arrival time is entered, “wholesale” can be performed at the same time as arrival at the delivery destination YZ Kakegawa Factory. Accordingly, the work start time of “wholesale” at the YZ Kakegawa factory, which is the delivery destination, is set to 10:28 am (10:28). And the work time of “wholesale” of 150 industrial containers in the YZ Kakegawa factory takes 70 minutes as shown in FIG. 28, so the work start time is 10:28 (10:28). The departure time is set to 11:38 am (11:38) with 70 minutes added.

  Next, the time setting of the truck of the car number 7253 belonging to the DEF transportation which is the second truck will be described. As can be seen from the plan candidate arrangement data shown in FIG. 25, the truck of the car number 7253 performs the sequential delivery of the delivery products of the order numbers <2>, <1>, and <3>. In other words, as shown in FIG. 26, the truck with the car number 7253 is “stacked” at the MNO Chemical Shizuoka Office for each of the 100 industrial containers corresponding to the order numbers <2> and <1>. After “wholesale” 100 industrial containers corresponding to order number <1> at warehouse Ogishima sales office, “100 industrial containers corresponding to order number <2>” at GHI warehouse Katachidoki sales office “Wholesale” and “stacking” of 200 industrial containers corresponding to order number <3>, and “wholesale” of 200 industrial containers corresponding to order number <3> at PQR Chemical Fuchu Factory.

  In the plan candidate arrangement data shown in FIG. 25, the “stack” of order number <2> and the order number <1 between the second track mark “0” from the left and the second date mark “−1” from the left. > “Stack” is inserted, so these two “stacks” become “loading” of “22 days” which is the day before the delivery date “23 days”. Thus, since “loading” of order number <2> becomes “loading”, the arrival time at the MNO Chemical Shizuoka office is set to 5:00 pm (17:00), and the work start time is not set. . The departure time from the MNO Chemical Shizuoka Office is the arrival time at the delivery destination of the order number <1> for the first “wholesale” on the 23rd, which is the delivery date. 24:49, which is the time obtained by subtracting 4 hours and 11 minutes, which is the travel time from the MNO Chemical Shizuoka Office to the JKL Warehouse Ogishima Office in the geographic data shown in FIG. 24, from 9 am (9:00). (4:49).

  The “loading” of order number <1> becomes “loading” in the same way as the “shipping” of order number <2> at the MNO Chemical Shizuoka office. The arrival time is set to 5 pm (17:00), the work start time is not set, and the departure time is set to 4:49 am (4:49). In addition, since “loading” of order number <1> is “loading” and the delivery time is specified at 9:00 am (9:00) as shown in FIG. Both the arrival time and the work start time are set at 9:00 am (9:00). In addition, as shown in FIG. 28, at the JKL warehouse Ogishima sales office where the “wholesale” of order number <1> is performed, the work time of “wholesale” of 100 industrial containers is 20 minutes. The departure time of “Wholesale” in <1> is set to 9:20 am (9:20), which is the work start time 9:00 am (9:00) plus the above 20 minutes.

  Next, the arrival time of “wholesale” of order number <2> is 9:20 am (9:20), which is the departure time of “wholesale” of order number <1>, and the geographic data shown in FIG. It is set to 9:50 am (9:50), which is 30 minutes, which is the travel time from the JKL warehouse Ogishima sales office to the GHI warehouse Katachidoki sales office. In addition, as shown in FIG. 26, the delivery time of the order number <2> is specified at 10:00 am (10:00) and is later than the arrival time, so the delivery time at the GHI warehouse Kachidoki sales office that is the delivery destination The work start time is set to the designated 10:00 am (10:00). The work time of “wholesale” of 100 industrial containers at the GHI warehouse Kachidoki sales office takes 25 minutes as shown in FIG. 28, so the work start time is 10:00 am (10:00) 10:25 am (10:25) obtained by adding 25 minutes to the time is set as the departure time.

  Next, since the “loading” of the order number <3> is performed at the GHI warehouse Kachidoki sales office which is the same point as the “wholesale” of the order number <2>, the arrival time of the order number <3> is the order number <3> Set to 10:25 am (10:25), which is the same as the departure time of 2>. Further, as can be seen from FIG. 26, the order number <3> has no designated loading time, so the work start time is set to 10:25 am (10:25) which is the same as the arrival time. Next, as shown in FIG. 28, at the GHI warehouse Kachidoki sales office where “stacking” of order number <3> is performed, the working time of “stacking” of 200 industrial containers is 35 minutes. The departure time of the “stack” with the number <3> is set to 11:00 am (11:00), which is the work start time 10:25 am (10:25) plus 35 minutes.

  Next, the arrival time of “wholesale” of order number <3> is the GHI warehouse of the geographic data shown in FIG. 24 at 11:00 am (11:00), which is the departure time of “loading” of order number <3>. It is set to 11:56 am (11:56), which is 56 minutes, which is the travel time from the Kachidoki sales office to the PQR Chemical Fuchu factory. In addition, as shown in FIG. 26, the delivery time of the order number <3> is designated from 1 pm to 5 pm (13:00:00 to 17:00) and is later than the arrival time. The work start time at a certain PQR chemical Fuchu factory is set to 1 pm (13:00) as specified. And, as shown in FIG. 28, the work time for “wholesale” of 200 industrial containers at the PQR Chemical Fuchu Factory takes 50 minutes, so it takes 50 minutes at 1 pm (13:00), which is the work start time. Is set to 1:50 pm (13:50) as the departure time. Therefore, as a result of executing the time management process, a delivery plan shown in FIG. 29 is created.

In step SP6 shown in FIG. 11, the control unit 2 executes an evaluation value calculation process. That is, the control unit 2 obtains an evaluation value used for evaluating the delivery plan candidate (plan candidate array data) created in step SP4 in the plan adoption determination process performed in step SP7, and then proceeds to step SP7. . The lower the evaluation value, the better the plan candidate sequence data.
Next, more detailed processing of the evaluation value calculation processing will be described with reference to the flowchart shown in FIG. First, the control unit 2 proceeds to the process of step SD1 shown in FIG. 30 and creates the order data and fare table stored in a predetermined area of the storage unit 6 and the time management process of step SP5, Based on the geographic data stored in the predetermined area, the cost (fare, mileage, etc.) is calculated for the delivery plan created in step SP5 and set as the initial value of the evaluation value, and then the process proceeds to step SD2. Here, the fare table is determined according to the type of goods to be conveyed, the type of can, and the like according to the contract concluded between each shipper and each shipping company, and a predetermined area of the storage unit constituting the host computer 11 Are stored in advance.

  In step SD2, the control unit 2 refers to the order data and the track data stored in the predetermined area of the storage unit 6 for the delivery plan created in step SP5, and among the constraints shown in FIG. . 10-No. It is examined whether or not it is violated by 29, and if it is violated, it is weighted according to the importance of the constraint condition as a penalty, added to the evaluation value obtained in step SD1, and then proceeds to step SD3. . In step SD3, the control unit 2 refers to the order data and the track data stored in the predetermined area of the storage unit 6 for the delivery plan created in step SP5, and evaluates other than steps SD1 and SD2 (for example, The number of unallocated orders and the like are added, weighted according to the importance for each item, added to the evaluation value obtained in step SD2, and then the process returns to the main routine shown in FIG. 11 and proceeds to step SP7.

In step SP7 shown in FIG. 11, the control unit 2 executes a plan adoption determination process. That is, the control unit 2 calculates the evaluation value of the delivery plan (not the best delivery plan that is held) adopted until the previous iteration and the step SP6 of the delivery plan candidate newly created in step SP4. After comparing with the evaluated value and holding the delivery plan candidate with the best evaluation value as the delivery plan, the process proceeds to step SP8.
Next, a more detailed process of the plan adoption determination process will be described with reference to the flowchart shown in FIG. First, the control unit 2 proceeds to the process of step SE1 shown in FIG. 31, calculates the employment probability, and then proceeds to step SE2. The above-described simulated annealing method is used in the plan candidate creation process including the adoption probability calculation process, the evaluation value calculation process, and the plan adoption determination process.

Here, an outline of the simulated annealing method will be described. In the simulated annealing method, the initial state of the system, the initial temperature of the system, and the cooling schedule are set, and the calculation is repeated while the temperature is lowered until the system reaches the thermal equilibrium state, and the state with the lowest energy value is optimized. Judge as a solution. In each iteration, the following (i) to (iii) are performed.
(I) Create a neighboring state from the current state.
(Ii) Calculate the energy (cost) value of the created state.
(Iii) Adopting the neighboring state as a new state.
In the determination of (iii), the probability p to be adopted is expressed by the equation (14) based on the difference ΔE (see equation (13)) between the current state energy E ′ and the energy E of the nearby state and the current temperature T. ) And (15), and is performed according to the probability.
ΔE = E′−E (13)
If ΔE <0,
p = 1 (14)
In other cases,
p = exp (ΔE / T) (15)
As described above, the simulated annealing method can prevent the local optimum solution having high energy from being determined as the optimum solution by probabilistically adopting the wrong solution.

In this embodiment, the simulated annealing method is applied by regarding the above “state” as a “delivery plan”. That is, the difference ΔE between the evaluation value E of the delivery plan adopted up to the previous iteration and the evaluation value E ′ of the newly created delivery plan candidate is calculated by the above equation (13). Next, a certain number of iterations are collectively referred to as “stages”, and the temperature T _i of each “stage” is calculated by equation (16).
T _{i + 1} = βT _i (16)
In Expression (16), β is a cooling coefficient.
Next, based on the temperature T _i , the probability to be adopted is calculated by the equations (14) and (15) in the same manner as the probability p described above. However, T in equation (15) is T _i . Then, whether or not the delivery plan can be adopted is determined based on the probability p and the random number. Note that the initial temperature, the cooling coefficient β, and the number of repetitions per stage are set to appropriate values by trial and error.
In step SE3, the control unit 2 holds an executable delivery plan having the lowest evaluation value in the iterative calculation, that is, stores it in a predetermined area of the storage unit 6, and then returns to the main routine of FIG. The process proceeds to step SP8.

  In step SP8 shown in FIG. 11, the control unit 2 determines whether or not the processes in steps SP4 to SP7 described above have been executed a predetermined number of times. The predetermined number of times is determined in advance by trial and error. If the determination result in step SP8 is “NO”, the control unit 2 returns to step SP4 and repeats the processes in steps SP4 to SP7 described above. In this case, one “stage” described above is updated. Then, when the processes in steps SP4 to SP7 described above are executed a predetermined number of times, the determination result in step SP8 is “YES”, and the control unit 2 proceeds to step SP9. In step SP9, the control unit 2 outputs that the delivery plan last stored in the predetermined area of the storage unit 6 in step SP7 is finally adopted, that is, after displaying on the display unit 4, a series of The process ends.

  When the delivery plan created by executing the delivery plan creation process described above is displayed on the display unit 4, the operator looks at the delivery plan and examines whether correction is necessary. When it is determined that correction is necessary, the operator operates the operation unit 3 to correct the delivery plan. Specifically, for the purpose of adjusting the delivery date, the operator reassigns the delivery items to the truck, considers the order data conditions, the delivery source or delivery destination conditions, the truck conditions, etc. Unassigned delivery items, such as delivery items that were not ordered at the time of planning, are allocated.

  The operator can confirm on the map data what route each truck is supposed to take by operating the operation unit 3 for the delivery plan displayed on the display unit 4. For example, on the map data, it is possible to confirm what number of national roads a certain truck runs on, and from which interchange when a highway is used, and from which interchange. Further, the operator can confirm various conditions such as the delivery source, delivery date (date and time), and the number of transports of the delivery product by operating the operation unit 3 with respect to the delivery plan displayed on the display unit 4. The confirmation of this condition is mainly performed when the operator corrects the delivery plan described above.

  In addition, the operator operates the operation unit 3 with respect to the delivery plan displayed on the display unit 4 to display management data such as the total number of transports per day of a certain shipping company on the display unit 4. The delivery plan can be evaluated. For example, it is evaluated whether each delivery item can be delivered according to the delivery date, or whether more delivery items can be stacked in one truck for delivery date adjustment or the like. When the operator determines that actual delivery is possible as a result of correcting the delivery plan displayed on the display unit 4, the operator gives an instruction for approval by operating the operation unit 3. As a result, the delivery plan is transmitted to the host computer 10 via the cable. If there is a new delivery request after the approval, the delivery plan can be corrected on the host computer 11 side.

  The host computer 10 to which the delivery plan is transmitted creates data relating to necessary forms such as delivery instructions and delivery notes based on the delivery plan, and each shipper client 12 and each shipping company client via the network 14. 13 to send. As a result, each shipping company client 13 receives a delivery instruction for instructing delivery of a delivery assigned to each truck belonging to the shipping company and a delivery for each delivery, and the printing section respectively To print. Each truck to which these delivery instructions and delivery forms correspond is handed to the truck driver. Thereby, the delivery of the delivery goods according to a delivery plan is started.

In this embodiment, as shown in FIG. 10, a track mark representing a boundary between tracks and a date mark representing a date boundary are used, and “loading” is performed with a value corresponding to the order number assigned to each order. "Wholesale" is expressed by adding the total number of orders to the value corresponding to the order number, so that the one-dimensional planned array data according to the number of orders of delivery goods and the number of trucks in charge of delivery Have created. In addition, the entire country of Japan, which is the delivery target of deliveries, is divided into long-distance transportation areas and short-distance transportation areas, and each is divided into different mesh sizes, and travel between any two short mesh points and between any long mesh points. The distance and travel time are calculated in advance and a mesh point-to-mesh information database is created. With reference to this mesh point-to-point information database, the travel distance and travel time between any delivery source and destination in Japan are calculated. Calculated using the multiple regression method. Then, a delivery plan is created based on the above-described plan arrangement data and the travel distance and travel time between the delivery source and the delivery destination, and an evaluation value is calculated for the obtained delivery plan, and simulated annealing is performed. The optimal delivery plan is determined by evaluating using the method.
Therefore, when delivering a delivery item from a plurality of delivery sources to a plurality of delivery destinations, even when a delivery item is wholesaled at each delivery destination and another delivery item is loaded and delivered to the next delivery destination, the transportation efficiency is high. A delivery plan can be created.

The embodiment has been described in detail with reference to the drawings. However, the specific configuration is not limited to the embodiment, and there are design changes and the like without departing from the scope of the invention. Are also included in the present invention. For example, in the above-described embodiment, an example of creating a delivery plan in the case where the number of orders is 4 and the number of trucks is 2 has been described. However, the present invention is not limited to this, and the number of orders may be any number. Any number of can be used. In this case, a plurality of trucks may belong to each transportation company.
Further, in the above-described embodiment, an example in which a truck is used as a transportation vehicle has been described.
Further, in the above-described embodiment, the example in which the delivery object of the delivery product is the whole country of Japan has been shown. However, the present invention is not limited to this, and the delivery object of the delivery product may be a part of Japan or the whole world, Asia, Europe. , East Asia, Scandinavia, etc.

Further, in the above-described embodiment, an example in which the planned array data is one-dimensional has been shown, but the present invention is not limited to this and may be two-dimensional.
Further, in the above-described embodiment, the planned array data represents “stack” with a value corresponding to the order number, and represents “wholesale” with a value obtained by adding the total number of orders to the value corresponding to the order number. However, the present invention is not limited to this, and “loading” is represented by the first value related to the order number, and “wholesale” corresponding to the above “loading” is represented by the second value related to the order number. In other words, “load” may be expressed in association with the order number, and “wholesale” is a value that can be expressed as a one-to-one correspondence with the corresponding “load”. Anything can be used. For example, for “wholesale”, a predetermined value may be added to, subtracted from, or multiplied by, or a flag may be set for the predetermined value.

Moreover, although the example which applies this invention to the exclusive delivery plan preparation apparatus 1 was shown in the above-mentioned embodiment, it is not limited to this. For example, the present invention can be applied to a general-purpose personal computer. In this case, the delivery plan creation program and the like are installed in the personal computer as one of application programs.
Moreover, in the above-mentioned embodiment, although the example which delivers industrial containers, such as a drum can, a pail can, and a gas container which filled oil, edible oil, natural gas, etc. inside was shown, it is not limited to this. For example, in the present invention, from the standpoint of environmental protection, the above-mentioned industrial containers are washed, shaped, painted, and reused (reused), or damaged industrial containers are scrapped to produce new iron products. The present invention can also be applied to physical distribution when used as a raw material to be produced (recycled).

It is a block diagram which shows the structure of the delivery plan preparation apparatus which is embodiment of this invention. It is a block diagram which shows the structure of the physical distribution system to which a delivery plan preparation apparatus is applied. It is a figure which shows an example of a structure of order data. It is a figure which shows an example of a structure of track data. It is a figure which shows an example of a structure of place data. It is a figure which shows an example of a structure of parameter data. It is a figure which shows an example of a structure of geographic data. It is a figure which shows an example of a structure of a delivery plan. It is a figure which shows an example of the constraint conditions at the time of performing a plan candidate creation process and an evaluation value calculation process in a delivery plan creation apparatus. It is a figure which shows the specific example of plan arrangement | sequence data. It is a flowchart which shows the delivery plan creation process which a delivery plan creation apparatus performs. It is a flowchart which shows an initial plan preparation process. It is a figure which shows an example of the plan arrangement | sequence data produced by the initial plan production process. It is a flowchart which shows a plan candidate creation process. It is a figure showing an example of plan arrangement data changed by plan candidate creation processing. It is a flowchart which shows a time management process. It is a figure for demonstrating the necessity to produce geographic data. It is a figure which shows an example which divides | segments a short-distance transport area by a short mesh point. It is a figure which shows an example which divides | segments a long-distance transport area by a long mesh point. It is a figure for demonstrating the precision of a mesh division | segmentation. It is a figure for demonstrating the calculation method of the required time and distance between two points. It is a flowchart which shows the process which calculates the required time and distance between mesh points. It is a flowchart which shows geographic data creation processing. It is a figure which shows the specific example of geographic data. It is a figure which shows the specific example of plan arrangement | sequence data. It is a figure which shows the specific example of order data. It is a figure which shows the specific example of track data. It is a figure which shows the specific example of location data. It is a figure which shows the specific example of a delivery plan. It is a flowchart which shows an evaluation value calculation process. It is a flowchart which shows plan adoption judgment processing.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Delivery plan creation apparatus, 2 Control part (Initial plan creation means, Plan candidate creation means, Time management means, Evaluation value calculation means, Plan adoption judgment means) 3 Operation part 4 Display part 5 Communication control part 6 Storage Department, 11 Host computer, 12 Shipper client, 13 Shipping company client, 14 Network.

Claims (16)

One or more delivery sources, one or more delivery destinations, delivery items, transportation facilities, and a periodical plan creation process for creating an initial delivery plan based on constraints related thereto,
A plan candidate creation process for creating a delivery plan candidate by correcting the initial delivery plan;
An evaluation value calculation process for obtaining an evaluation value for evaluating the delivery plan candidate;
And a plan adoption determination process for comparing the evaluation values of the respective delivery plan candidates and setting the delivery plan candidate having the best evaluation value as a final delivery plan.   2. The delivery plan creation method according to claim 1, further comprising a time management process for setting an arrival time and a departure time at each delivery source and each delivery destination based on the delivery plan candidate.   The initial delivery plan and the delivery plan candidate are created for each transportation agency responsible for delivery of each order, using a transportation agency mark representing a boundary between the transportation agency and a date mark representing a date boundary. 3. The delivery plan creation method according to claim 1, wherein the delivery plan is created.   The initial delivery plan and the delivery plan candidate correspond to the stack with the first value related to the order number assigned to each order and the stack with the second value related to the order number. 4. The delivery plan creation method according to claim 3, wherein the wholesale of the delivered products is represented.   A mesh point is set for each predetermined distance in the delivery target area of the delivery item, and the information on the distance between the mesh points is calculated in advance. In the management process, the travel distance and the travel time of the transportation facility between any delivery source and the delivery destination in the delivery target area are calculated with reference to the mesh point information database. The delivery plan creation method according to any one of claims 2 to 4.   6. The delivery plan creation method according to claim 5, wherein the predetermined distance is different between a part of the delivery target areas and another area.   The delivery plan creation method according to any one of claims 1 to 6, wherein a simulated annealing method is used in the plan candidate creation process, the evaluation value calculation process, and the plan adoption determination process. An initial plan creation means for creating an initial delivery plan based on the input one or more delivery sources, one or more delivery destinations, delivery items, transportation facilities, and constraints related thereto,
Plan candidate creation means for modifying the initial delivery plan to create a delivery plan candidate;
Evaluation value calculation means for obtaining an evaluation value for evaluating the delivery plan candidate;
A delivery plan creation device comprising: plan adoption judgment means for comparing the assessment values of the delivery plan candidates and setting the delivery plan candidate having the best assessment value as a final delivery plan.   9. The delivery plan creation device according to claim 8, further comprising time management means for setting an arrival time and a departure time at each delivery source and each delivery destination based on the delivery plan candidate.   The initial delivery plan and the delivery plan candidate are created for each transportation agency responsible for delivery of each order, using a transportation agency mark representing a boundary between the transportation agency and a date mark representing a date boundary. 10. The delivery plan creation device according to claim 8 or 9, wherein   The initial delivery plan and the delivery plan candidate correspond to the stack with the first value related to the order number assigned to each order and the stack with the second value related to the order number. 11. The delivery plan creation apparatus according to claim 10, wherein each of the delivered goods wholesale is represented. A mesh point is set for each predetermined distance in the delivery target area of the delivery item, and a mesh point-to-mesh information database that pre-calculates the travel distance and travel time of the transport between any two mesh points is provided,
The time management means calculates the travel distance and the travel time of the transportation facility between any delivery source and delivery destination in the delivery target area with reference to the mesh point information database. 12. The delivery plan creation device according to claim 9,   13. The delivery plan creation device according to claim 12, wherein the predetermined distance is different between a part of the delivery target areas and another area.   The delivery plan creation apparatus according to any one of claims 8 to 13, wherein the plan candidate creation unit, the evaluation value calculation unit, and the plan adoption determination unit use a simulated annealing method.   15. A delivery plan creation program for causing a computer to realize the function according to any one of claims 1 to 14. 15. A physical distribution system comprising the delivery plan creation device according to claim 8.

Images