Site optimization device, site optimization method, and site optimization program

The site optimization device optimizes base locations by using service data and travel time to maximize workload coverage, addressing the limitations of existing systems in base allocation.

JP7583358B2Active Publication Date: 2024-11-14TOKYO ELECTRIC POWER CO HOLDINGS INC
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
JP2020174442
Authority / Receiving Office
JP · JP
Patent Type
Patents
Current Assignee / Owner
Filing Date
2020-10-16
Publication Date
2024-11-14
Estimated Expiration
2040-10-16

AI Technical Summary

Technical Problem

Existing layout optimization systems only optimize the allocation of maintenance personnel and inventory management of replacement parts, failing to provide a reliable system for optimizing the location of bases.

Method used

A site optimization device and method that utilizes a service number storage unit, prediction model generation, workload calculation, and mathematical engineering models to determine the optimal location of bases based on service operations, travel time, and workload estimation.

Benefits of technology

Enables the optimization of base locations to efficiently provide services by maximizing workload coverage within a service area, ensuring high service levels and reducing the number of bases required.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 0007583358000002
    Figure 0007583358000002
  • Figure 0007583358000003
    Figure 0007583358000003
  • Figure 0007583358000004
    Figure 0007583358000004
Patent Text Reader

Abstract

To enable optimization of a base arrangement.SOLUTION: A method for base optimization includes the steps of: storing a number of base-specific service cases that is a number of service work cases per base generated in a service object area of the base being a foothold in providing predetermined service work; generating a prediction model which models a relationship between attribute information and a number of base-specific service cases by using attribute information of a service object area of a base, and a number of base-specific service cases; calculating a number of region-specific service cases of a relevant service region on the basis of the attribute information representing attributes of a service region being different from a service object area, using the prediction model; estimating an amount of work in a service region on the basis of the number of region-specific service cases; modeling a mathematical engineering model which models movement time and the amount of work in a service region, using the amount of work in the service region and the movement time to the service region from the base; and calculating an optimal solution for an optimal base on the basis of the mathematical engineering model.SELECTED DRAWING: Figure 2
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to a base optimization device, a base optimization method, and a base optimization program, and, for example, to a base optimization device, a base optimization method, and a base optimization program that optimize the placement of bases that are installed or positioned to serve as a base for providing a specified service. [Background technology]

[0002] Conventionally, a placement optimization system has been proposed that optimizes the placement of maintenance personnel and / or replacement parts for maintenance management of multiple devices (see, for example, Patent Document 1). This placement optimization system includes a storage unit that stores number distribution information related to the distribution of the number of devices, regional information related to the regions where each device is installed, seasonal information, device model information, maintenance record information, etc., and a placement determination unit that determines the placement of maintenance personnel and / or replacement parts based on the number distribution information and regional information, etc.

[0003] In this placement optimization system, the placement determination unit determines the placement of maintenance personnel and / or replacement parts for maintaining and managing the equipment by using not only equipment number distribution information but also area information where each piece of equipment is located, equipment model information, maintenance record information, etc. In this way, the placement optimization system optimizes the number of maintenance personnel and the inventory management of replacement parts according to each area, eliminating the occurrence of areas with shortages of maintenance personnel and replacement parts. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] JP 2006-155441 A Summary of the Invention [Problem to be solved by the invention]

[0005] However, the placement optimization system described in Patent Document 1 merely optimizes the number of maintenance personnel and inventory management of replacement parts for each region, but is unable to optimize the placement of bases that serve as bases for providing various services, for example.

[0006] The present invention has been made in consideration of the above problems, and has an object to realize a site optimization device, a site optimization method, and a site optimization program that are capable of optimizing the placement of sites. [Means for solving the problem]

[0007] The base optimization device of the present invention includes a base-specific service number memory unit that stores the number of service cases by base, which is the number of service operations that occurred for each base in the service target area of ​​the base that is the basis for providing a specified service operation; a prediction model generation unit that uses attribute information of the service target area of ​​the base and the number of services by base to generate a prediction model that models the relationship between the attribute information and the number of services by base; a service number prediction unit by area that uses the prediction model to calculate the number of services by area for a service area based on attribute information that represents the attributes of a service area that is different from the service target area; a workload calculation unit that estimates the workload of the service area based on the number of services by area; a mathematical engineering model modeling unit that models a mathematical engineering model that models the travel time and the workload of the service area using the travel time required from the base to the service area and the workload of the service area; and an optimal solution calculation unit that calculates an optimal solution for the optimal base based on the mathematical engineering model.

[0008] It is preferable that the workload calculation unit estimates the workload of the service area based on the number of services provided by each area and the travel time from the base to the service area.

[0009] It is preferable that the optimal solution calculation unit calculates the base where the total value of the workload for the service area is maximum as the optimal solution for the optimal base.

[0010] The base optimization method of the present invention includes a base-specific service number storage step for storing the number of service cases by base, which is the number of service cases for each base that occurred in the service target area of ​​the base that is the basis for providing a predetermined service operation, in a base-specific service number storage unit; a prediction model generation step for generating a prediction model that uses attribute information of the service target area of ​​the base and the number of service cases by base, and models the relationship between the attribute information and the number of service cases by base, in a prediction model generation unit; and a prediction model generation step for generating the service by using the prediction model, based on attribute information representing the attributes of a service area different from the service target area. the number of services by area prediction step of calculating the number of services by area in a service area by an area-by-area service number prediction unit; a workload estimation step of estimating the workload of the service area based on the number of services by area by a workload estimation unit; a mathematical engineering model modeling step of modeling a mathematical engineering model that models the travel time and the workload of the service area by using the travel time required from the base to the service area and the workload of the service area by a mathematical engineering model modeling unit; and an optimal solution calculation step of calculating an optimal solution for the optimal base based on the mathematical engineering model by an optimal solution calculation unit.

[0011] The location optimization program of the present invention causes a computer to execute the following steps: a location-specific service count storage step for storing the number of service cases per location, which is the number of service operations that occurred for each location in the service target area of ​​a location that is the basis for providing a specified service operation; a prediction model generation step for using attribute information of the service target area of ​​the location and the number of services per location to generate a prediction model that models the relationship between the attribute information and the number of services per location; a service number per area prediction step for using the prediction model to calculate the number of services per area for a service area based on attribute information that represents the attributes of a service area that is different from the service target area; a workload estimation step for estimating the workload of the service area based on the number of services per area; a mathematical engineering model modeling step for modeling a mathematical engineering model that models the travel time and the workload of the service area using the travel time required from the location to the service area and the workload of the service area; and an optimal solution calculation step for calculating an optimal solution for the optimal location based on the mathematical engineering model. Effect of the Invention

[0012] According to the present invention, it is possible to realize a site optimization device, a site optimization method, and a site optimization program capable of optimizing the location of sites. [Brief description of the drawings]

[0013] [Figure 1] 1 is a diagram showing an overall configuration of a base optimization system according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a diagram showing a functional block configuration of a site optimization device according to the present embodiment. [Diagram 3] FIG. 2 is a diagram illustrating a hardware configuration of a site optimization device according to the present embodiment. [Figure 4] FIG. 2 is a diagram showing cities, wards, towns and villages that are service areas of bases A to C according to the present embodiment. [Diagram 5] 1 is a diagram showing the area, population, business quantity, and business volume of a base according to an embodiment of the present invention. FIG. [Figure 6] 1 is a diagram showing the business quantity and business volume relative to the area and population of each base in the present embodiment. FIG. [Figure 7] FIG. 2 is a diagram showing travel times (intercity times) between bases and municipalities according to the present embodiment. [Figure 8] 1 is a diagram corresponding to N×M geographic matrix data indicating whether it is possible to reach a city, ward, town, or village within a target travel time in accordance with the present embodiment. [Figure 9] FIG. 2 is a diagram for explaining the principle of calculating the optimal two bases from three bases in this embodiment. [Figure 10] FIG. 13 is a diagram showing an optimal combination of bases according to the present embodiment. [Figure 11] 10 is a flowchart illustrating a processing procedure in an optimal solution calculation unit according to the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] [1] Overview of the embodiment of the present invention First, a typical embodiment of the invention disclosed in this application will be outlined. In the following description, as an example, reference numerals in the drawings corresponding to components of the invention are given in parentheses.

[0015] [1] A location optimization device (2) according to a representative embodiment includes a location-specific service number storage unit (65) for storing the number of service cases by location, which is the number of service cases for each location that occurred in the service target area of ​​the location (A to C) that is the basis for providing a predetermined service operation; a prediction model generation unit (41) for generating a prediction model that models the relationship between the attribute information and the number of service cases by location, using attribute information of the service target area of ​​the location and the number of service cases by location; and a prediction model generation unit (42) for generating a prediction model that models the relationship between the attribute information and the number of service cases by location, using the prediction model, for a service area other than the service target area. The system includes an area-by-area service number prediction unit (42) that calculates the number of services by area in the service area based on attribute information representing attributes, a workload calculation unit (43) that estimates the workload (G) of the service area based on the number of services by area, a mathematical engineering model modeling unit (51) that models a mathematical engineering model that models the travel time and the workload (G) of the service area using the travel time required from the base to the service area and the workload (G) of the service area, and an optimal solution calculation unit (52) that calculates an optimal solution for the optimal base based on the mathematical engineering model.

[0016] [2] In the above [1], the workload calculation unit (43) estimates the workload (G) of the service area based on the number of services by area and the travel time from the base to the service area.

[0017] [3] In the above [1] or [2], the optimal solution calculation unit (52) calculates the base where the sum (Gsum) of the workload (G) for the service area is maximum as the optimal solution of the optimal base.

[0018] [4] In a representative embodiment of a location optimization method, a location-specific service number storage step is used to store a location-specific service number, which is the number of service operations for each location that occurred in the service target area of ​​the location (A to C) that is the basis for providing a predetermined service operation, in a location-specific service number storage unit (65); a prediction model generation step is used to generate a prediction model that models a relationship between the attribute information and the location-specific service number by using attribute information of the service target area of ​​the location and the location-specific service number by a prediction model generation unit (41); and a service optimization step is used to optimize the service by using the prediction model based on attribute information representing attributes of a service area different from the service target area. the number of services by area predicting step of calculating the number of services by area for a service area by an area-by-area service number predicting unit (42); a workload estimating step of estimating the workload (G) for the service area based on the number of services by area by a workload estimating unit (43); a mathematical engineering model modeling step of modeling a mathematical engineering model that models the travel time and the workload (G) for the service area by using the travel time required from the base to the service area and the workload (G) for the service area by a mathematical engineering model modeling unit (51); and an optimal solution calculation step of calculating an optimal solution for an optimal base by an optimal solution calculation unit (52) based on the mathematical engineering model.

[0019] [5] A representative embodiment of a location optimization program causes a computer to execute a location-specific service number storage step for storing the number of service cases by location, which is the number of service operations that occurred for each location (A to C) in the service target area of ​​the location (A to C) that is the basis for providing a specified service operation; a prediction model generation step for generating a prediction model that uses attribute information of the service target area of ​​the location and the number of service cases by location to model the relationship between the attribute information and the number of service cases by location; a service number prediction step for calculating the number of service cases by area for a service area based on attribute information that represents the attributes of a service area different from the service target area using the prediction model; a workload estimation step for estimating the workload (G) of the service area based on the number of service cases by area; a mathematical engineering model modeling step for modeling a mathematical engineering model that models the travel time and the workload (G) of the service area using the travel time required from the location to the service area and the workload (G) of the service area; and an optimal solution calculation step for calculating an optimal solution for an optimal location based on the mathematical engineering model.

[0020] [2] Specific examples of embodiments A specific example of this embodiment will be described below with reference to the drawings. In the following description, the same reference numerals are used to designate components common to this embodiment, and repeated description will be omitted. It should be noted that the drawings are schematic, and the arrangement of each element, data format, communication method, etc. may differ from reality.

[0021] Fig. 1 is a diagram showing the overall configuration of a base optimization system 1 according to this embodiment. The base optimization system 1 shown in the figure is a system that optimizes bases by determining which bases to reduce and which to retain and maintain among the sales bases (hereinafter, these may be simply called "bases") that are already established in the prefecture at present and serve as the bases for personnel performing service operations, such as sales offices, and that are the basis for performing service operations, for example, in the service target area where there is demand for service operations, for example, multiple cities, wards, towns, and villages in the prefecture.

[0022] In this case, service work includes, in utility-related public works such as electricity, gas, water, and communications, targeting access equipment such as power lines, gas pipes, water, and communication lines connected to customer equipment, and customer equipment such as power demand equipment, heat source equipment, water equipment, and communications equipment, and work responding to damage to access equipment caused by natural disasters, investigating malfunctions of customer equipment, and responding to requests for equipment updates, etc. However, service work is not limited to these utility-related public works, but can target any customer equipment, and can include various other work responding to breakdowns, repairs, inspections, etc. of that customer equipment.

[0023] The location optimization system 1 can also optimize the number and locations of locations that are planned to be established in the future. This location optimization system 1 can optimize the service area by leaving only the minimum number of locations necessary to provide the most efficient service operations.

[0024] In this specification, the following description will be given using a city, ward, town, or village in a prefecture as an example, but the present invention is not limited to this and may be applicable to any city, ward, town, or village in any prefecture, or may be applicable to a service area of ​​a base on a regional or national level. The base optimization system 1 is applicable not only to utility-related public service businesses, but also to all business types including dispatching to customers, such as goods delivery businesses, mobile product sales businesses, and food delivery businesses, and is applicable to decisions on the establishment, consolidation, and closure of bases from which customers are dispatched for delivery, sales, takeout, and other related work. In addition, the system can be applicable to decisions on the establishment, consolidation, and closure of facilities in public works such as the operation of fire stations and schools.

[0025] In this case, for example, in the base optimization system 1, a call center (not shown) of an operator that accepts requests for service work receives a direct call from a location such as an ordinary home or business in any city, ward, town, or village that has requested the service work (hereinafter, these may be referred to as locations where there is demand for the service work, i.e., "demand points"), and sets the travel time, which is the number of minutes it takes to travel from any base to the demand point, as the service level evaluation time.

[0026] For example, in this embodiment, the service level evaluation time is set to 40 minutes as an example, but the present invention is not limited to this, and the travel time set as the service level can be arbitrarily determined depending on the location, such as the suburbs of the city center or the suburbs. Also, not only when a service operation is requested, but also when the operator detects information on a service outage such as a power outage, it is possible to set the travel time from the time of detection to the demand point as the service level evaluation time.

[0027] 1, the base optimization system 1 includes a base optimization device 2 and a plurality of client terminal devices 5a to 5d. The base optimization device 2 is configured to be connectable to a network 4, and is connected to external information processing devices such as the plurality of client terminal devices 5a to 5d via the network 4.

[0028] The network 4 is, for example, a wide area network (WAN) such as the Internet. The site optimization device 2 and the client terminal devices 5a to 5d of the site optimization system 1 are configured to be connectable to the network 4 by wired or wireless communication, and are capable of transmitting and receiving data to and from each other.

[0029] The client terminal devices 5a to 5d are information processing devices on the client side that request the site optimization device 2 as a server for optimal site placement. The client terminal devices 5a to 5d can receive and display an output screen that is information on the optimally placed site (the estimated result of the optimal site) estimated by the site optimization device 2. However, the client terminal devices 5a to 5d do not necessarily have to request the site optimization device 2 for optimal site placement, and the site optimization device 2 may itself determine the optimal site placement and simply receive the result from the site optimization device 2. Examples of the client terminal devices 5a to 5d include various information processing devices (computers) such as personal computers, tablet terminals, and smartphones. The client terminal devices 5a to 5d are configured to be connectable to the network 4, and therefore can also be connected to external information processing devices such as the site optimization system 1 via the network 4 for communication.

[0030] Next, a specific description will be given of the configuration of the base optimization device 2 in the base optimization system 1. Fig. 2 is a diagram showing a functional block configuration of the base optimization device 2 according to the present embodiment. Fig. 3 is a diagram showing a hardware configuration of the base optimization device 2 according to the present embodiment.

[0031] The base optimization device 2 is, for example, an information processing device (computer) such as a PC (Personal Computer) or a server, and is a device that estimates the workload G (described later) for each city, ward, town, and village according to a base optimization program installed in advance, and then calculates the optimal base to be maintained based on the workload G. Note that, instead of being performed solely by the base optimization device 2, these base optimization programs may be provided from the base optimization device 2 to the client terminal devices 5a to 5d online or offline (for example, a predetermined package storing the base optimization program), and these client terminal devices 5a to 5d may function as base optimization devices.

[0032] 2, the base optimization device 2 has, as functional blocks, a communication unit 20, a data input / output unit 30, a workload estimation unit 40, an optimal base estimation unit 50, a storage unit 60, and a display unit 70. Each of these functional blocks is realized by the cooperation of hardware resources constituting an information processing device as the base optimization device 2 with software (base optimization program) installed in the information processing device.

[0033] As shown in FIG. 3, the site optimization device 2 includes, as hardware resources, a calculation device 101, a storage device 102, an input device 103, an I / F (Interface) device 104, an output device 105, and a bus 106.

[0034] The program 111 stored in the memory device 102 includes a base optimization program and a basic program that cause a computer to function as each functional unit of the base optimization device 2 in this embodiment, and is pre-installed in the memory device 102, for example.

[0035] In addition, the data 112 stored in the memory device 102 is data that has been stored or will be stored in the city / ward / town / village basic information memory unit 61, the base basic information memory unit 62, the route memory unit 63, the city / ward / town / village service count memory unit 64, and the base service count memory unit 65 of the memory unit 60, which will be described later.

[0036] The program 111 and the data 112 may be distributed via a network, or may be distributed in a state in which they are written on a computer-readable storage medium (non-transitory computer readable medium) such as a CD-ROM.

[0037] The location optimization device 2 will be described below. For example, if multiple cities, wards, towns, and villages in a certain prefecture are currently set as service target areas where there is demand for service operations, there are already three locations ("Location A", "Location B", and "Location C") such as sales offices established in the certain prefecture at present, and these multiple cities, wards, towns, and villages are assigned to each of the locations A to C as their service target areas without overlapping. However, a city, ward, town, and village may partially overlap with multiple locations A to C.

[0038] In such a situation, when there is a request to reduce the number of bases in the service area from the current three to two, the base optimization device 2 estimates which of "base A", "base B" and "base C" is optimal to maintain and leave, that is, it is a device for optimizing or relocating bases. Naturally, the base optimization device 2 can set various numbers of bases to be maintained, such as reducing three bases to two, reducing to one base, or reducing ten bases to seven, etc. Of course, these three bases are not limited to only existing bases, and may be, for example, a case in which three bases, all of which are planned to be installed in the future, are reduced to two bases.

[0039] The following describes each functional block of the base optimization device 2. The communication unit 20 is a functional unit for communicating with an external information processing device. Specifically, the communication unit 20 transmits and receives data to and from the multiple client terminal devices 5a to 5d in the base optimization system 1 via the above-mentioned predetermined network (e.g., LAN) 4.

[0040] The data input / output unit 30 of the site optimization device 2 is an input / output interface such as a keyboard, a pointing device, etc. However, this is not limited to this, and data can be input / output wirelessly via a tablet terminal, a smartphone, etc. The data input / output unit 30 inputs various data to the memory unit 60 and stores it, and reads various data from the memory unit 30 and outputs it to the workload estimation unit 40 and the optimal site estimation unit 50.

[0041] Next, a description will be given of the storage unit 60. The storage unit 60 has a city / ward / town / village basic information storage unit 61, a base basic information storage unit 62, a route storage unit 63, a city / ward / town / village service count storage unit 64, and a base service count storage unit 65, and various data are stored or read out via the data input / output unit 30.

[0042] The city, ward, town, village basic information storage unit 61 is a part of the storage unit 60 in which basic information (hereinafter, this may be referred to as "city, ward, town, village basic information") including the area, population, and the address of the government office used as the center of the city, ward, town, village (hereinafter, this may be referred to as "city, ward, town, village office address") of each city, ward, town, village input by the operator via the data input unit 30 is stored. Here, the city, ward, town, village basic information such as the area, population, and address of the city, ward, town, village office is also attribute information that represents the attributes of the city, ward, town, village. The area and population of the city, ward, town, village basic information are official information that are published on the homepage of the city, ward, town, village, etc. In addition, the city, ward, town, village basic information includes the address of the city, ward, town, village office, but is not limited to this, and various other positions such as the center of gravity of the city, ward, town, village on a map or the center of gravity of the population may be used as the location information of the center representing the city, ward, town, village.

[0043] Furthermore, the basic information of municipalities may include not only the area, population, and location of municipal offices, but also the number of contracts for access equipment and customer equipment, the number of contracts, the household density, the population density, the business density, the age-weighted average, the proportion of aged buildings, etc. This basic information of municipalities may be stored, for example, in any of the client terminal devices 5a to 5d or from an external information processing device via the network 4 and the communication unit 20.

[0044] For example, assume that there are three bases A to C in a certain prefecture, and that hh city, ab city, cd city, gh city, aa town, jj town, cc town, and gg town are set as the service area of ​​base A, kk city, mn city, ef city, and op city are set as the service area of ​​base B, and uv city, ss town, wx town, qr town, st town, and kl town are set as the service area of ​​base C, as shown in Figures 4 and 5. In this case, for example, base A is located in cd city, base B is located in ef city, and base C is located in st town.

[0045] As shown in Figure 5, the service area of ​​base A is one of the cities hh, which has an area of ​​186.96 km 2 The population of 195,633 is stored as basic city, town, and village information (attribute information) in the city, town, and village basic information storage unit 61. 2 The population of 110,046 is stored as basic city, town, and village information in the city, town, and village basic information storage unit 61. The distance from cd city, which is one of the service target areas of the base A, is 194.9 km. 2 The population of 88,078 is stored as city, ward, town, and village basic information in the city, ward, town, and village basic information storage unit 61. Similarly, the areas and populations of the remaining five cities, ward, town, and villages, from gh city to gg town, which are the service target area of ​​base A, are stored as city, ward, town, and village basic information in the city, ward, town, and village basic information storage unit 61.

[0046] For the cities KK, MN, EF, and OP, which are in the service area of ​​base B, city, ward, town, and village basic information (attribute information such as area, population, etc.) for each city, ward, town, and village is stored in the city, ward, town, and village basic information storage unit 61. Similarly, for the cities UV, SS, WX, QR, ST, and KL, which are in the service area of ​​base C, city, ward, town, and village basic information (attribute information such as area, population, etc.) for each city, ward, town, and village is stored in the city, ward, town, and village basic information storage unit 61.

[0047] The base basic information storage unit 62 is a part of the storage unit 60 in which the total area of ​​all cities, wards, towns, and villages assigned to a certain base in the service area and the total population of all the cities, wards, towns, and villages are stored. For example, as shown in FIG. 6, the cities, wards, towns, and villages assigned to base A are hh city, ab city, cd city, gh city, aa town, jj town, cc town, and gg town, so the total area of ​​these eight cities, wards, towns, and villages is 818.3 km2. 2 The total population of these eight cities, wards, towns, and villages, 578,101 people, is stored in the basic information storage unit 62 by the data input unit 30 as basic information of the base A.

[0048] In addition, the cities, wards, towns, and villages assigned to base B are kk, mn, ef, and op, so the total area of ​​these four cities, wards, towns, and villages is 644.5 km. 2 The total population of these four cities, wards, towns, and villages, 185,358 people, is stored in the base basic information storage unit 62 as the base basic information of base B.

[0049] In addition, the cities, wards, towns, and villages assigned to base C are UV City, SS Town, WX Town, QR Town, ST Town, and KL Town, so the total area of ​​these six cities, wards, towns, and villages is 583.6 km2. 2 The total population of these six cities, wards, towns, and villages, 66,438 people, is stored in the base basic information storage unit 62 as the base basic information of base C.

[0050] The route storage unit 63 is a portion of the storage unit 60 that stores one-way travel time (hereinafter, this may be referred to as "intercity time") required to travel a distance corresponding to a route from a certain base to the location of a city, ward, town, or village office in the service area by vehicle such as an automobile (including motorcycles, bicycles, etc.). Such travel time (intercity time) is stored by the operator via the data input / output unit 30, but the travel time may be automatically calculated and stored by a route calculation device (not shown) based on geographic information such as the latitude and longitude of the location of base A and the location of the government office. Note that the route storage unit 63 may store one-way distances between cities instead of travel times.

[0051] For example, as shown in FIG. 7, the intercity time required for one-way travel from base A to No. 3-1, XX-cho, which is the municipality of hh city assigned to base A, is 16 minutes, the intercity time required for one-way travel from base A to No. 2-2, XX-cho, which is the municipality of ab city, is 6 minutes, and so on, and finally, the intercity time required for one-way travel from base A to ▲□57-2, gg town, which is the municipality of gg town, is 57 minutes. These intercity times (one-way travel times) are stored in the route memory unit 63. Similarly, the intercity times required for one-way travel to multiple municipalities assigned to base B and the intercity times required for one-way travel to multiple municipalities assigned to base C are also stored in the route memory unit 63.

[0052] The city, ward, town, and village service number storage unit 64 is a portion of the storage unit 60 in which the city, ward, town, and village service number representing the number of services for the past year for each city, ward, town, and village (the total number of services that would have been requested for all service operations such as damage due to natural disasters, malfunction investigations, and equipment updates) calculated by the area-specific service number prediction unit 42 described later is stored. The city, ward, town, and village service number will be described later.

[0053] The storage unit 65 for storing the number of service cases by location is a portion of the storage unit 60 for storing the number of service cases by location in the past year for service work such as damage caused by natural disasters, malfunction investigations, and equipment updates for each location, which are input by the operator via the data input unit 30. In other words, the number of service cases by location is the value of the number of service cases that occurred at a certain location in the past year, and is known data compiled by the call center of the location.

[0054] As shown in Figure 6, in all cities, wards, towns, and villages in base A, the total number of "damage caused by natural disasters" service operations was 2,088 per year, the total number of "malfunction investigation" service operations was 191 per year, and the total number of "equipment renewal" service operations was 879 per year, so a total of 3,158 service operations existed at base A in the past year, and these 3,158 cases are the number of services by base A.

[0055] At base B, the total number of service operations for "damage due to natural disasters" was 1,903 cases per year, the total number of service operations for "malfunction investigations" was 527 cases per year, and the total number of service operations for "equipment renewals" was 728 cases per year, so a total of 3,158 service operations existed at base B over the past year, and these 3,158 cases are the number of services by base B.

[0056] At base C, the total number of service operations for "damage due to natural disasters" was 417 cases per year, the total number of service operations for "malfunction investigations" was 215 cases per year, and the total number of service operations for "equipment renewals" was 148 cases per year, so a total of 780 service operations existed at base C over the past year, and these 780 cases are the number of services by base C.

[0057] Next, the workload estimation unit 40 will be described. The workload estimation unit 40 is a functional unit that estimates the workload G (described later) for each city, ward, town, and village. The workload estimation unit 40 has a prediction model generation unit 41, a service number prediction unit by area 42, and a workload calculation unit 43.

[0058] The prediction model generation unit 41 is a functional unit that inputs attribute information (explanatory variable data) such as area and population for each of the bases A to C, which is the base basic information (FIG. 6) read from the base basic information storage unit 62 via the data input / output unit 30, and the number of services for each base for the bases A to C (objective variable data) read from the base service number storage unit 65 via the data input / output unit 30 as teacher data, performs machine learning, and generates a prediction model that learns and models the relationship between the attribute information (explanatory variable data) and the number of services for each base (objective variable data), and stores the model in an internal storage area. Incidentally, the attribute information may be either the area or the population, or only some other attribute information. Here, for example, a neural network called a multilayer perceptron may be adopted as a machine learning model for generating the prediction model. The prediction model may be a multiple regression model.

[0059] The area-specific service number prediction unit 42 is a functional unit that predicts the number of services by city, ward, town, and village using the prediction model generated by the prediction model generation unit 41. Here, although the number of services at each base over the past year is stored in the base-specific service number storage unit 65, the actual number of services that occurred in each city, ward, town, and village is unknown, so the area-specific service number prediction unit 42 calculates the number of services by city, ward, town, and village.

[0060] When the area and population of each city, ward, town, or village is input as attribute information via the data input / output unit 30 to the prediction model generated by the prediction model generation unit 41, the area-specific service number prediction unit 42 can calculate the number of services for each city, ward, town, or village corresponding to the city, ward, town, or village. For example, as shown in FIG 5, the area-specific service number prediction unit 42 can calculate the number of services for each city, ward, town, or village corresponding to the city, ward, town, or village via the data input / output unit 30. 2When the area and population of 195,633 are input, the prediction model can calculate the number of services for hh city as 561 based on the area and population. In this way, the area-specific service number prediction unit 42 calculates the number of services by city, ward, town, and village for each city, ward, town, and village, and stores them in the city, ward, town, and village-specific service number storage unit 64.

[0061] In this case, as shown in Fig. 5, the city / ward / town / village-specific service count storage unit 64 stores 561 city / ward / town / village-specific service counts in hh city of base A, 389 city / ward / town / village-specific service counts in ab city, 415 city / ward / town / village-specific service counts in cd city, ..., and 377 city / ward / town / village-specific service counts in gg town. The total number of city / ward / town / village-specific service counts assigned to base A is 3,158, the same as the number of base-specific services for base A (see Fig. 6).

[0062] Furthermore, the city / ward / town / village service count storage unit 64 stores 800 city / ward / town / village service counts for kk city in base B, 781 city / ward / town / village service counts for mn city, 801 city / ward / town / village service counts for ef city, and 775 city / ward / town / village service counts for op city. The total number of city / ward / town / village services assigned to base B is 3158, the same as the number of services by base for base B (see FIG. 6). Incidentally, the number of services by base for base A and the number of services by base for base B are the same value, but this is by coincidence and has no particular meaning.

[0063] Furthermore, the city / ward / town / village service count storage unit 64 stores 133 city / ward / town / village service counts in uv city, 137 city / ward / town / village service counts in ss town, 131 city / ward / town / village service counts in wx town, 131 city / ward / town / village service counts in qr town, 112 city / ward / town / village service counts in st town, and 136 city / ward / town / village service counts in kl town. The total number of city / ward / town / village service counts assigned to these locations C is 780, the same as the location-specific service count for location C.

[0064] The workload calculation unit 43 is a functional unit that converts the number of services by city, ward, town, and village into a workload (task time) for each city, ward, town, and village that is close to the actual situation, based on the number of services by city, ward, town, and village stored in the number of services by city, ward, town, and village storage unit 64, the round trip travel time (intercity time) from the base to the location of the government office stored in the route storage unit 63, the round trip travel time from the location of the government office to the actual demand point, and the work time required at the demand point (local), and stores this in an internal storage area. However, in this embodiment, the workload calculation unit 43 does not include the work time required at the demand point (local) in the calculation in order to shorten and simplify the process, and the following explanation will be given.

[0065] Specifically, the workload calculation unit 43 can estimate the workload G for each city, ward, town, or village using the following formula (1) based on the one-way travel time t1 from the location of a government office with city, ward, town, or village basic information to the nearest base (hereinafter, this may be referred to as the "nearest base"), the total average travel time ta (=(t1+t2) / 2) including the travel time t2 from the next adjacent base to the location of the government office, the travel time tc within the city from the location of the government office to the actual demand point, and the number of services Gc in the city, ward, town, or village. Here, in formula (1), the "2" used in 2·(ta+tc) means the round-trip travel time. G=Gc×2·(ta+tc)……………………………………………………(1) Here, when the travel time is used to calculate the workload G, it is necessary to determine the base from which the travel to the customer's demand point is to be made, but when calculating the optimal solution described below, it is not possible to determine the base from which the travel to be made before the calculation. In other words, it is not known which base should be used. Therefore, in terms of calculation accuracy, it is desirable to calculate the workload G based on the travel time from the city, ward, town, or village that is the service area to multiple bases.

[0066] However, this is not a limitation, and in order to shorten processing time and simplify processing, the workload calculation unit 43 may estimate the workload G according to the following formula (2) by multiplying the one-way travel time t1 (minutes) from the location of the government office in the basic city / ward / town / village information to the nearest base by the number of services by city / ward / town / village Gc in that city / ward / town / village. This simplifies the processing in the workload calculation unit 43, and therefore the processing time can be shortened. Here, in formula (2), the "2" used in 2t means the round-trip travel time. G=Gc×2·t1………………………………………………………………(2)

[0067] Next, a description will be given of the optimal base estimation unit 50. The optimal base estimation unit 50 is a functional unit that estimates which of "base A", "base B", and "base C" is optimal to keep and maintain when a request to reduce the current number of bases from three to two is made by the base optimization device 2 or the operator of the client terminal devices 5a to 5d.

[0068] The optimal location estimation unit 50 has a mathematical and engineering model modeling unit 51 and an optimal solution calculation unit 52. The mathematical and engineering model modeling unit 51 receives, via the data input / output unit 30, the workload G for each city, town, or village estimated by the workload calculation unit 43 of the workload estimation unit 40, and the intercity time from the location to the government office stored in the route storage unit 63.

[0069] The mathematical engineering model modeling unit 51 is a functional unit that models a mathematical engineering model that models the relationship between the intercity time t, which is the travel time required to travel along a route from a base to the location of a government office, and the workload G of the city, ward, town, or village, using the intercity time t and the workload G of the city, ward, town, or village. This mathematical engineering model uses, for example, a maximum cover method. However, it is not limited to the maximum cover method, and a mathematical engineering model using a so-called p-median method that minimizes the sum of travel times between a candidate base and a candidate city, ward, town, or village may be modeled. Note that instead of a mathematical engineering model, a reinforcement learning model may be modeled.

[0070] Here, the maximum coverage method is a technique for finding the optimal placement of N bases (N: an integer greater than or equal to 1) that can provide maximum coverage (provide service operations) to the multiple cities, wards, towns, and villages that make up the service area.In this case, the optimal locations of the bases can be determined so as to maximize the workload G of the multiple cities, wards, towns, and villages that can be reached by the bases in the intercity time t.

[0071] The optimal solution calculation unit 52 is a functional unit that can calculate optimal base candidates to be maintained by using the calculation object parameters input by the operator via the data input / output unit 30 as initial calculation conditions and the workload G for each city, ward, town, and village in a mathematical engineering model. Here, the calculation object parameters are the number of bases N representing the number of bases to be maintained that will ultimately remain (in this case, "N=2" because the number of bases is reduced from three to two), the number of cities, wards, towns, and villages M representing the number of cities, wards, towns, and villages (in this case, the number of all cities, wards, towns, and villages for the service target area of ​​three bases (M=18)), and the service target time X from the base (one-way intercity time t from the base to the location of the city, ward, town, and village office). For convenience of explanation and simplification, the service target time X is not the one-way travel time from the base to the demand point, but the location of the city, ward, town, and village office.

[0072] The optimal solution calculation unit 52 calculates optimal base candidates to be maintained so as to match the calculation objective parameters according to a mathematical engineering model. For example, if it is desired to reduce the current three bases (bases A to C) to two bases (bases A and B or bases A and C or bases B and C), the number of bases N=2 and the number of cities, wards, towns, and villages M=18 (the total of the number of cities, wards, towns, and villages of base A "8", the number of cities, wards, towns, and villages of base B "4", and the number of cities, wards, towns, and villages of base C "6"). The service target time X is the one-way intercity time t from the base to the location of the city, ward, town, and village office, and is set to 30 minutes, 20 minutes, or 10 minutes. This service target time X can be set by the operator to any intercity time t.

[0073] In practice, when the number of locations N, the number of cities, wards, towns, and villages M, and the target service time X are input as calculation object parameters, the optimal solution calculation unit 52 generates N×M geographic matrix data MX as shown in FIG. 8, which lists whether each location can reach all cities, wards, towns, and villages within the target service time X.

[0074] In this geographic matrix data MX, for example, when the service target time X=30 minutes, a cell with a flag of "1" means that the city, ward, town, or village can be reached from the base within 30 minutes, and a cell with a flag of "0" means that the city, ward, town, or village cannot be reached from the base within 30 minutes. In other words, this means that hh city can be reached from base A within 30 minutes, and gg town cannot be reached from base A within 30 minutes. At the same time, hh city and jj town, which are the service target area of ​​base A, can be reached from base B within 30 minutes. This geographic matrix data MX is an example when the service target time X=30 minutes, and naturally the cells with a flag of "1" will change when the service target time X=20 minutes or 10 minutes.

[0075] The optimal solution calculation unit 52 uses the maximum coverage method, which is a heuristic method, while referring to the geographic matrix data MX to calculate the two combinations of a base and a city, ward, town, or village that satisfy the service target time X (minutes) as optimal bases, which have the largest total Gsum of the workload G of the city, ward, town, or village. Next, a method for calculating the optimal base from the total Gsum of the workload G will be described using a simple example.

[0076] Specifically, for example, as shown in Fig. 9(A), consider a case where city α where a demand point is located can be reached in 30 minutes from base B, city β where a demand point is located can be reached in 30 minutes from base A, and city γ where a demand point is located can be reached in 30 minutes from base C. In this case, the circle indicates the range of demand points that can be reached from the bases in 30 minutes.

[0077] As shown in FIG. 9B, instead of the above-mentioned geographic matrix data MX, simple geographic matrix data mx targeting bases A to C and cities α, β, and γ is used. Furthermore, as shown in FIG. 9C, in this case, the workload Wi (i=1) of city α is "40", the workload Wi (i=2) of city β is "30", and the workload Wi (i=3) of city γ is "20". Here, the workload Wi corresponds to the above-mentioned workload G. Also, in this case, for the sake of convenience, an example is shown in which it is calculated which two bases are optimal to keep among bases A to C, targeting three municipalities, α, β, and γ.

[0078] In the geographic matrix data mx, the vertical axis i (i = 1 to 3) indicates the cities α, β, and γ, which are the demand points for the service business, and the horizontal axis j (j = 1 to 3) indicates the bases A to C. In this case, the flag "1" at the position [i, j] = [1, 2] indicates that the city α can be reached from the base B in 30 minutes of travel time (meaning that the base B can cover the city α). Also, the flag "1" at the positions [i, j] = [2, 1] and [2, 3] indicates that the city β can be reached from both the base A and the base C in 30 minutes of travel time. Furthermore, the flag "1" at the position [i, j] = [3, 3] indicates that the city γ can be reached from the base C in 30 minutes of travel time.

[0079] In such a case, let us consider the case where we need to calculate which two locations are optimal under the constraint that only two locations are to be left out of the three locations A through C. First, a mathematical engineering model performs trial calculations for three cities, α, β, and γ, and determines the cover variable Zi (i = 1 or 0) when two locations, A and B, are left out. Here, the cover variable Zi is a variable that indicates "1" when each city, ward, or village is covered by either A, B, or C, and "0" when it is not covered by any of them.

[0080] Here, if two locations, A and B, are left, based on the geographic matrix data mx, city α is covered by location B, so it is set to "1," city β is covered by location A, so it is set to "1," and city γ is not covered by either location A or B, so it is set to "0." Therefore, when selecting locations A and B, the coverage variable Zi = (1, 1, 0).

[0081] Next, the mathematical and engineering model performs further calculations for the three municipalities of α, β, and γ, and if base A and C are left, based on the geographic matrix data mx, α is not covered by either base A or C, so it is set to "0", β is covered by both base A and B, so it is set to "1", and γ is covered by base C, so it is set to "1". Therefore, when bases A and C are selected, the cover variable Zi = (0, 1, 1). Note that the number of calculations that the mathematical and engineering model performs can be set arbitrarily before the calculation, and the number of calculations in this case is "2". Therefore, when the number of calculations is set to "2", the mathematical and engineering model performs calculations for the cases where bases A and B are selected, and the cases where bases A and C are selected.

[0082] Now that the cover variable Zi has been determined, the mathematical engineering model then calculates the total values ​​Gsum(AB) and Gsum(AC) of the workload G for each of the two locations based on the cover variable Zi and the workload Wi. These total values ​​Gsum(AB) and Gsum(AC) are calculated using the following formula (3). Here, i is an element of set I, Wi indicates the workload G for cities α, β, and γ, and Zi indicates the cover variable. TIFF0007583358000001.tif21148

[0083] Therefore, when selecting locations A and B, the cover variable Zi = (1, 1, 0), the workload of City α Wi (i = 1) is "40", the workload of City β Wi (i = 2) is "30", and the workload of City γ Wi (i = 3) is "20", so the total value Gsum(AB) = 40 x 1 + 30 x 1 + 20 x 0 = 70.

[0084] In addition, when selecting locations A and C, the cover variable Zi = (0, 1, 1), the workload of City α (i = 1) is “40”, the workload of City β (i = 2) is “30”, and the workload of City γ (i = 3) is “20”, so the total value Gsum(AC) = 40 × 0 + 30 × 1 + 20 × 1 = 50.

[0085] The total values ​​Gsum(AB) and Gsum(AC) of the workload G for each of the two bases calculated in this way are compared, and it can be determined that it is optimal to maintain and keep bases A and C, which have the largest total value Gsum(AC) = 70. Using this method, the optimal solution calculation unit 52 can determine the two bases.

[0086] When multiple service target times X (minutes) are set, the optimal solution calculation unit 52 refers to the geographic matrix data MX created for all service target times X (minutes), calculates the workload Gsum for each pair of bases, and calculates the combination of two bases with the largest total value Gsum of the workload G as the optimal solution. That is, the optimal solution calculation unit 52 compares the total values ​​Gsum of the workload G of the cities, wards, towns, and villages covered by the two bases, and calculates the two bases with the largest total value Gsum. In this way, the optimal solution calculation unit 52 calculates the two bases using a so-called grid search method that calculates a solution set based on multiple service target times X.

[0087] In this case, the optimal solution calculation unit 52 performs trials for a plurality of service target times X, and then creates an output screen Ans showing the calculation results of the trial evaluation, for example, as shown in Fig. 10, and displays this on the display unit 70, or transmits it to the client terminal devices 5a to 5d via the communication unit 30 and the network 4. This output screen Ans shows that in the case of the service target time X (30 minutes) of trial 1, the combination of site A and site C has been selected.

[0088] In the trials, a service target time X was set, and the service area within the service target time X from the base was used as the target for accumulating the workload G. For example, in trial 3, a base was selected that prioritized a service area within 10 minutes from the base.

[0089] Next, the optimal solution calculation unit 52 aims to evaluate the service levels of all service areas corresponding to the service level evaluation time X' of the base selected by trial, and displays the evaluation result via the output screen Ans. Here, the service level evaluation time X' can be set to a time different from the service target time X, and can be set to, for example, 40 minutes. Of course, it is also possible to set any service level evaluation time X', such as 50 minutes or 60 minutes.

[0090] In calculating the optimal solution before displaying this output screen Ans, the optimal solution calculation unit 52 generates a service evaluation geographical matrix MX' corresponding to the service level evaluation time X' (40 minutes in this case) instead of the geographical matrix MX for the location combination A and C in trial 1 in the case of the service target time X (30 minutes), in order to evaluate all service areas corresponding to the service level evaluation time X' (40 minutes in this case), determines a cover variable Zi' corresponding to the service level evaluation time X', and calculates the total value G' of the workload G' (hereinafter also referred to as the "service covering workload") G'sum.

[0091] In Fig. 10, the ratio of the service coverage workload G'sum covered by sites A and C to the total workload of all service areas corresponding to the service level evaluation time X' (40 minutes in this case) (hereinafter, this is also referred to as the "service level") is equivalent to 96%. Furthermore, Fig. 10 also shows an example of the site combination B and C in trial 2, allowing the service levels of trial 1 and trial 2 to be compared.

[0092] The process flow of the optimal solution calculation unit 52 in the optimal base estimation unit 50 will now be described with reference to a flowchart. Fig. 11 is a flowchart showing the process flow by the optimal solution calculation unit 52. When the optimal solution calculation unit 52 receives in step SP1 that the calculation object parameters (number of bases N, number of cities, wards, towns, and villages M, and target service time X) have been set or changed via the data input / output unit 30, it proceeds to the next step SP2.

[0093] In step SP2, the optimal solution calculation unit 52 generates N×M geographic matrix data MX (FIG. 8) showing in a list whether or not it is possible to reach a city, ward, town, or village from a base within a service target time X (minutes), and then proceeds to the next step SP3.

[0094] The optimal solution calculation unit 52, while referring to the geographic matrix data MX, calculates the combination of two bases with the largest sum Gsum of the workload G among the combinations of bases and cities / ward / towns / villages that satisfy the service target time X (minutes) as the optimal base candidate, and proceeds to the next step SP4.

[0095] In step SP4, the optimal solution calculation unit 52 judges whether or not the calculation of base candidates has been completed for all the service target times X (30 minutes, 20 minutes, 10 minutes) set as the calculation objective parameters, and if a negative result is obtained, the process returns to step SP1 and repeats the above-mentioned processes of steps SP1 to SP3. In contrast, if a positive result is obtained in step SP4, the optimal solution calculation unit 52 proceeds to step SP5.

[0096] In step SP5, the optimal solution calculation unit 52 calculates the service level of the service covering workload G'sum for the service level evaluation time X' of, for example, 40 minutes for the candidate base for the service target time X (30 minutes, 20 minutes, 10 minutes), outputs the service level to the display unit 70 via the output screen Ans, and ends the processing flow.

[0097] According to the above embodiment, when reducing three locations, locations A to C, to two locations, the location optimization device 2 can calculate the two optimal locations that can handle multiple cities, wards, towns, and villages equivalent to the three locations with a high service level. In particular, the location optimization device 2 can estimate the two optimal locations that can handle the set times of 30 minutes, 20 minutes, and 10 minutes of the service target time X, respectively, and can also present the service levels of the two locations to the operator via the output screen Ans (FIG. 10). This allows the operator to easily recognize the service levels of the two locations for the service level evaluation time X' from the extraction result screen Ans.

[0098] In addition, the base optimization device 2 converts the number of services provided by city, ward, town, or village into a workload G that corresponds to the actual situation, and calculates candidate bases using that workload G, thereby making it possible to provide optimal bases and the service levels of those bases.

[0099] [3] Other embodiments In the above embodiment, the above-mentioned processing is performed by the site optimization device 2, but the present invention is not limited to this, and the processing may be distributed between the site optimization device 2 and the client terminal devices 5a to 5d. In this case, the above-mentioned processing is performed by the entire site optimization system 1.

[0100] Furthermore, the above configurations of the site optimization system 1, the site optimization device 2, and the client terminal devices 5a to 5d are merely examples, and the specific configurations are not limited to these.

[0101] Furthermore, the above-mentioned flowchart by the optimal solution calculation unit 52 shows an example for explaining the flow of processing in the optimal solution calculation unit 52, and is not limited to this. That is, each step of the flowchart by the optimal solution calculation unit 52 is a specific example, and is not limited to this flow. For example, the order of some steps may be changed, other processing may be inserted between the processing of each step, and in some cases, the processing of some steps may be performed in parallel. [Explanation of symbols]

[0102] 1...location optimization system, 2...location optimization device, 4...network, 5a-5d...client terminal device, 20...communication unit, 30...data input / output unit, 40...business volume estimation unit, 41...prediction model generation unit, 42...service number prediction unit by area, 43...business volume calculation unit, 50...optimum location estimation unit, 51...mathematical engineering model modeling unit, 52...optimum solution calculation unit, 60...memory unit, 61...city / ward / town / village basic information memory unit, 62...location basic information memory unit, 63...route memory unit, 64...city / ward / town / village service number memory unit, 65...location number memory unit, 70...display unit, 101...arithmetic unit, 102...memory device, 111...program, 112...data, 103...input device, 104...I / F device, 105...output device, 106...bus.

Claims

1. a storage unit for storing the number of service cases by location, which is the number of service cases by location that occurred in the service target area of ​​the location that is the base when providing a predetermined service; a prediction model generation unit that generates a prediction model by machine learning the attribute information of the service target area of ​​the base as an explanatory variable and the number of services by base as an objective variable using the attribute information and the number of services by base as training data; a region-specific service number prediction unit that uses the prediction model to calculate a region-specific service number of the service region different from the service target area based on the attribute information that represents an attribute of the service region; A workload calculation unit that estimates a workload of the service area based on the number of services by area; an optimal solution calculation unit that receives as input a cover variable, which is an index indicating whether or not it is possible to reach the service area from the base within a target arrival time, and a workload of the service area, calculates a total value of the workload based on a mathematical engineering model that is modeled so as to multiply the cover variable by the workload of the service area, add up the obtained value, and calculates a combination of the bases with the largest total value of the workload as an optimal solution; A location optimization device equipped with the above.

2. The workload calculation unit estimates a workload of the service area based on the number of services by area and a travel time from the base to the service area. The location optimization device according to claim 1 .

3. A step of storing the number of service cases by location, which is the number of service cases by location that occurred in the service target area of ​​the location that is the base when the computer provides a predetermined service; a prediction model generation step in which a computer generates a prediction model by machine learning the attribute information of the service target area of ​​the base as an explanatory variable and the number of services by base as an objective variable using the attribute information and the number of services by base as training data; a prediction step of calculating the number of services by area in a service area different from the service target area based on the attribute information representing attributes of the service area by using the prediction model; A business volume estimation step in which a computer estimates a business volume of the service area based on the number of services by area; an optimal solution calculation step in which a computer inputs a cover variable, which is an index indicating whether or not it is possible to reach the service area from the base within a target arrival time, and the workload of the service area, calculates a total value of the workload for each base based on a mathematical engineering model that is modeled to add up the values ​​obtained by multiplying the cover variable and the workload of the service area, and calculates the combination of bases with the largest total value of the workload as the optimal solution; A location optimization method comprising:

4. For computers, a storage step of storing the number of service cases by location, which is the number of service cases by location that occurred in the service target area of ​​the location that is the base when providing a predetermined service; a prediction model generation step of generating a prediction model in which attribute information of the service target area of ​​the base is used as an explanatory variable and the number of services by base is used as an objective variable by machine learning the attribute information and the number of services by base as training data; a region-by-region service number prediction step of calculating a region-by-region service number of the service region different from the service target area based on the attribute information representing attributes of the service region by using the prediction model; A business volume estimation step of estimating a business volume of the service area based on the number of services by area; an optimal solution calculation step in which a cover variable, which is an index indicating whether or not it is possible to reach the service area from the base within a target arrival time, and the workload of the service area are input, the cover variable is multiplied by the workload of the service area, and the resulting value is added up to calculate a total value of the workload for each base based on a mathematical engineering model, and the combination of bases with the largest total value of the workload is calculated as an optimal solution; A site optimization program that executes the following:

Citation Information

Patent Citations

  • Arrangement optimization system

    JP2006155441A

  • Vehicle arrangement method determination system

    JP2011204099A

  • Site analysis system

    JP2011227794A

  • Emergency demand prediction device, emergency demand prediction method and program

    JP2020086650A