Automatic shift generation device, automatic shift generation method, and automatic shift generation program

Abstract

To provide an automatic shift generation device, an automatic shift generation method and an automatic shift generation program capable of shortening a time required for convergence to an optimum solution.SOLUTION: The automatic shift generation device 200 includes a rule acquisition unit 211, an individual generation unit 212, and an individual overwriting unit 213. The rule acquisition unit 211 acquires a rule related to a shift. The individual generation section 212 generates a plurality of shift tables of the next generation by evolving a plurality of shift tables as a plurality of individuals of the current generation into a plurality of shift tables of the next generation using a genetic algorithm for evolving a plurality of individuals of the current generation into a plurality of individuals of the next generation. The individual overwriting unit 213 overwrites all or a part of the plurality of shift tables generated by the individual generation unit 212 so as to satisfy the rule.SELECTED DRAWING: Figure 6

Description

[Technical Field]

[0001] The present invention relates to an automatic shift generation device, an automatic shift generation method, and an automatic shift generation program. [Background technology]

[0002] Genetic algorithms have traditionally been used to generate operation schedules (diagrams) for buses, trains, and the like, as well as machine operation schedules (see, for example, Patent Document 1 below). Patent Document 1 discloses a schedule creation device that applies a genetic algorithm. This schedule creation device has a crossover means, a mutation means, a bias means, an evaluation means, and a selection means. The crossover means generates a child chromosome by rearranging the genetic sequence of a selected parent chromosome according to the genetic sequences of other parent chromosomes. The mutation means rearranges the genetic sequence of the child chromosome with a predetermined probability. The bias means manipulates the genetic sequence of the child chromosome every predetermined number of times according to a predetermined rule. The evaluation means evaluates to what extent the child chromosome satisfies preset conditions. The selection means selects parent chromosomes or child chromosomes with poor evaluation values ​​according to evaluation values ​​set for each parent chromosome and child chromosome.

[0003] In recent years, genetic algorithms are not limited to diagrams; they are also used to generate shifts for staff and employees in nursing homes, hospitals, factories, etc. In automatic shift generation using genetic algorithms, multiple shifts are generated as individuals for each generation based on rules. The rules are conditions related to shift work, such as work days, work hours, public holidays, and number of shifts. An evaluation value that indicates the degree to which the rules are satisfied is then calculated, and the individuals evolve to achieve a higher evaluation value. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2001-175639 Summary of the Invention [Problem to be solved by the invention]

[0005] However, when automatically generating shifts using a genetic algorithm, the amount of calculation increases exponentially depending on the number of dimensions of the individual and the types of shift symbols that can be placed. The number of dimensions of an individual is the number of shift symbols that can be placed in the shift schedule, and is equivalent to the product of the number of staff members to be shifted and the number of days in the shift schedule to be generated. Shift symbols are unique symbols that represent shift assignments in the work shift schedule. Shift symbols include, for example, "night" for night shift, "day" for day shift, "early" for early shift, "late" for late shift, "morning" for the end of a night shift, and "public" for public holiday. If the number of dimensions of an individual is n and the number of types of shift symbols is T, then the number of patterns that an individual can take is T. n Therefore, for example, in a nursing home, if there are a large number of staff members to be shifted or if there are a large number of shift symbols, the number of patterns that the individuals can take becomes enormous, and the amount of calculation required for the genetic algorithm also becomes enormous. As a result, it may take a long time for the individuals to converge to the optimal solution.

[0006] The present invention has been made in consideration of the above circumstances, and aims to provide an automatic shift generation device, an automatic shift generation method, and an automatic shift generation program that can shorten the time it takes for an individual to converge to an optimal solution. [Means for solving the problem]

[0007] The above-mentioned problems of the present invention are solved by the following means.

[0008] (1) An automatic shift generation device having: a rule acquisition unit that acquires rules related to shifts; an individual generation unit that generates a plurality of next-generation shift tables by evolving a plurality of shift tables as a plurality of individuals of a current generation into a plurality of next-generation shift tables using a genetic algorithm that evolves a plurality of individuals of a current generation into a plurality of next-generation individuals; and an individual overwrite unit that overwrites all or some of the plurality of shift tables generated by the individual generation unit so as to satisfy the rules.

[0009] (2) The automatic shift generation device described in (1) above further includes a priority setting unit that sets priorities between the rules, and the individual overwrite unit determines the order of overwriting by the rules according to the priorities.

[0010] (3) The automatic shift generation device described in (2) above, wherein the priority setting unit sets the priorities based on user input.

[0011] (4) The shift automatic generation device described in (2) above, wherein the individual generation unit calculates an evaluation value for each individual, and the priority setting unit sets the priority for the next generation individual based on the evaluation value for the current generation individual.

[0012] (5) The automatic shift generation device according to (3) or (4) above, wherein the priority setting unit sets the priority for each individual.

[0013] (6) The automatic shift generation device according to (1) or (2) above, wherein the individual overwriting unit sets the proportion of the individuals to be overwritten among the plurality of individuals based on user input.

[0014] (7) The automatic shift generation device according to (1) or (2) above, wherein the individual overwriting unit sets the proportion of the individuals to be overwritten among the plurality of individuals based on the evaluation values ​​of the individuals of the current generation.

[0015] (8) The shift automatic generation device described in (2) above, wherein the individual generation unit calculates an evaluation value for each individual, and the priority setting unit sets the priority for the next generation of individuals based on the difference in the evaluation values ​​between generations.

[0016] (9) The automatic shift generation device described in (1) or (2) above, wherein the rule acquisition unit extracts the rules from nursing care fee standards acquired by the facility.

[0017] (10) The automatic shift generation device described in (1) or (2) above, wherein the multiple rules include at least one of the following: the number of people required for early shifts, late shifts, or night shifts; the number of public holidays; the number of shifts each employee works; fixed shifts for each employee on certain days of the week; desired work schedules; the upper limit of consecutive shifts; consecutive shift patterns; discontinuous shift patterns; and the number of consecutive shifts with the same shift symbol.

[0018] (11) A method for automatically generating shift tables, comprising: a step (a) of acquiring rules regarding shifts; a step (b) of generating a plurality of next-generation shift tables by evolving a plurality of shift tables as a plurality of individuals of the current generation into a plurality of next-generation shift tables using a genetic algorithm that evolves a plurality of individuals of the current generation into a plurality of next-generation individuals; and a step (c) of overwriting all or some of the plurality of shift tables generated in the step (b) so as to satisfy the rules.

[0019] (12) An automatic shift generation program for causing a computer to execute a process including the steps of: (a) obtaining rules regarding shifts; (b) generating a plurality of next-generation shift tables by evolving a plurality of shift tables as a plurality of individuals of the current generation into a plurality of next-generation shift tables using a genetic algorithm that evolves a plurality of individuals of the current generation into a plurality of next-generation individuals; and (c) overwriting all or some of the plurality of shift tables generated in the step (b) so as to satisfy the rules. [Effects of the Invention]

[0020] According to the present invention, the automatic shift generation device, the automatic shift generation method, and the automatic shift generation program overwrite the individual so that it satisfies the acquired rules in the process of evolving the individual using a genetic algorithm. This limits the number of patterns that the individual can take. Therefore, the amount of calculation in the genetic algorithm is reduced. As a result, the time it takes for the individual to converge to the optimal solution can be shortened. [Brief explanation of the drawings]

[0021] Advantages and features provided by one or more embodiments of the present invention will be more fully understood from the following detailed description and the accompanying drawings, which are for purposes of illustration only and are not intended to define the limits of the invention. [Figure 1] 1 is a block diagram illustrating a schematic configuration of an automatic shift generation system according to a first embodiment of the present invention. [Figure 2] 2 is a block diagram showing a schematic configuration of the information processing terminal shown in FIG. 1. FIG. [Figure 3] 2 is a schematic diagram illustrating a rule reception screen displayed on the information processing terminal shown in FIG. 1. FIG. [Figure 4] This is a schematic diagram illustrating an example of a setting screen for items related to staffing system surcharges in nursing care fees. [Figure 5] 2 is a block diagram illustrating a schematic configuration of a server shown in FIG. 1. FIG. [Figure 6] 6 is a functional block diagram illustrating the functions of the CPU shown in FIG. 5. [Figure 7] 4 is a flowchart illustrating the processing steps of an automatic shift generation method by the automatic shift generation device in the first embodiment. [Figure 8] 8 is a subroutine flowchart illustrating an example of genetic manipulation processing (step S105) in the flowchart shown in FIG. 7. [Figure 9] FIG. 10 is a schematic diagram illustrating an example of a result of generating an initial population. [Figure 10] FIG. 10 is a schematic diagram illustrating an example of an evaluation result for each individual in the initial population of FIG. 9. [Figure 11] FIG. 1 is a schematic diagram illustrating the results of natural selection / crossover / mutation performed by a genetic algorithm. [Figure 12] FIG. 10 is a schematic diagram illustrating overwriting of an individual. [Figure 13] FIG. 13 is a schematic diagram showing an example of the evaluation result (second time) for each individual after overwriting in FIG. 12. [Figure 14] FIG. 10 is a schematic diagram illustrating the number of combination patterns of shift symbols when a rule is applied. [Figure 15] 10 is a graph illustrating an example of evaluation values ​​for each generation in the cases where overwriting is performed and where overwriting is not performed. [Figure 16] 16 is a diagram illustrating an example of an actual measurement value of the end time of the genetic algorithm in FIG. 15 and an evaluation value at the end time. [Figure 17] 10 is a flowchart illustrating a processing procedure of an automatic shift generation method according to the second embodiment. [Figure 18] FIG. 10 is a functional block diagram illustrating the functions of a CPU of a server according to a second embodiment. [Figure 19] FIG. 10 is a schematic diagram illustrating an example of an overwriting order in the process of overwriting an individual with a shift symbol (step S106). [Figure 20] FIG. 10 is a schematic diagram illustrating an example of setting priority levels by user input. [Figure 21] FIG. 10 is a schematic diagram illustrating an example of an overwriting order when priorities are set by user input and shift symbols are overwritten onto individuals. [Figure 22] FIG. 10 is a schematic diagram illustrating an example of an overwriting order when priorities are set based on evaluation values ​​and shift symbols are overwritten on individuals. [Figure 23] FIG. 10 is a schematic diagram illustrating an example of an overwriting order when a priority is set for each individual and a shift symbol is overwritten on the individual. [Figure 24] FIG. 10 is a schematic diagram illustrating a case where the ratio of individuals to be overwritten is set by user input and shift symbols are overwritten onto individuals. [Figure 25]FIG. 10 is a schematic diagram illustrating a case where the proportion of individuals to be overwritten is set based on an evaluation value, and shift symbols are overwritten onto individuals. DETAILED DESCRIPTION OF THE INVENTION

[0022] Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the scope of the present invention is not limited to the disclosed embodiments. In the drawings, the same elements are denoted by the same reference numerals, and duplicate explanations will be omitted. Furthermore, the dimensional proportions in the drawings are exaggerated for the sake of explanation and may differ from the actual proportions.

[0023] (First embodiment) FIG. 1 is a block diagram illustrating the schematic configuration of an automatic shift generation system 10 according to a first embodiment of the present invention. The automatic shift generation system 10 has an information processing terminal 100 and a server 200. The information processing terminal 100 and the server 200 are configured to be connectable to each other via a network. The server 200 is, for example, a cloud server, but may also be an on-premise server (server device). The server 200 constitutes an automatic shift generation device.

[0024] When the server 200 is a cloud server, the automatic shift generation system 10 may be configured with an information processing terminal 100 installed within a facility such as a nursing home, hospital, or factory, and a cloud server installed outside the facility. When the server 200 is an on-premise server, the automatic shift generation system 10 may be installed within a facility such as a nursing home, hospital, or factory. Alternatively, some or all of the components of the automatic shift generation system 10 may be installed outside the facility (for example, in the headquarters building of a corporation that operates multiple facilities). A corporation includes an organization consisting of multiple members (personnel). Members include all personnel who work in accordance with work shift rules (hereinafter simply referred to as "rules").

[0025] The configuration of the automatic shift generation system 10 and the method for automatically generating shifts using the automatic shift generation system 10 will be described below, taking as an example a case where shifts are generated for care staff at a care facility run by a corporation.

[0026] <Information processing terminal 100> The information processing terminal 100 is a terminal used by a user such as a facility director, manager, or administrator of a nursing care facility, and may be a desktop PC (Personal Computer), a notebook PC, a tablet PC, a smartphone, or the like.

[0027] Fig. 2 is a block diagram illustrating a schematic configuration of the information processing terminal 100 shown in Fig. 1. The information processing terminal 100 has a CPU (Central Processing Unit) 110, a ROM (Read Only Memory) 120, a RAM (Random Access Memory) 130, a storage 140, a communication interface 150, and an operation display unit 160. Each component is connected to each other via a bus so that they can communicate with each other.

[0028] The CPU 110 controls the above components and performs various arithmetic processing in accordance with programs recorded in the ROM 120 and storage 140. The ROM 120 stores various programs and various data. The RAM 130 temporarily stores programs and data as a working area.

[0029] Storage 140 is configured with a hard disk drive (HDD), a solid state drive (SDD), etc., and stores various programs including an operating system and various data. For example, an application is installed in storage 140 for receiving input (including instructions, etc.) from a user and transmitting and receiving various information to and from other devices such as server 200. In addition, an application is installed in storage 140 for outputting (displaying) information on operation display unit 160.

[0030] The communication interface 150 is an interface for communicating with other devices. As the communication interface 150, a wired or wireless communication interface conforming to various standards is used.

[0031] The operation display unit 160 is, for example, a touch screen display, and displays various information such as rules, which are conditions that the generated shift must satisfy, individuals, evaluation values, etc. The operation display unit 160 also accepts various inputs from the user, including rules, etc. Details of the evaluation values ​​will be described later.

[0032] The terms "rules" used herein may primarily refer to regulations, practices, and the like regarding work shifts at a facility. Rules may be established for each facility, taking into consideration factors such as the number of staff required for each caregiving task, laws and regulations such as the Labor Standards Act, staff preferences, and facility practices. For example, rules may include the number of staff required for early, late, or night shifts, the number of shifts each staff member works, fixed shift days for staff members, the maximum number of consecutive shifts, consecutive shift patterns, non-consecutive shift patterns, the number of consecutive shifts with the same symbol, the number of consecutive days of leave allowed, the number of public holidays (per week, per month) for each staff member, working hours (per week, per month) for each staff member, the number of working days (per week, per month) for each staff member, and the shifts each staff member can set. Rules may also include the number of night shifts each staff member desires (per week, per month) and the desired working days for each staff member. Staff members include full-time and part-time caregiving staff, nurses, and registered dietitians.

[0033] <Rule acquisition example 1: User input> The rules can be input by the user via, for example, the operation and display unit 160. FIG. 3 is a schematic diagram illustrating a rule reception screen displayed on the information processing terminal 100 shown in FIG. 1. The rule reception screen is displayed on the display of the operation and display unit 160. The operation and display unit 160 receives the rules, etc. input by the user on the rule reception screen and transmits them to the CPU 110. The CPU 110 transmits the received rules, etc. to the server 200. In the server 200, the received rules, etc. are stored in the RAM 230 or the storage 240 (see FIG. 4).

[0034] In this embodiment, for the sake of simplicity, it is assumed that, for example, four rules A to C shown in Table 1 below are stored in storage 240. Note that the rules exemplified in Table 1 are only a portion of the rules stored in storage 240, and the rules applicable to the present invention are not limited to these rules.

[0035] [Table 1]

[0036] In the example shown in Figure 3, the number of people required for each shift is entered as a rule on the rule reception screen. That is, the rule entered is that night shifts (shift symbol "night") are to be assigned to one staff member belonging to a unit called "Unit B" every day every month. The importance of night shift assignment is set to "Required."

[0037] Shift symbols are unique symbols that represent shift assignments in a shift schedule. Examples of shift symbols include "night" (night), "day" (day), "early," "late," "after," "after a night shift," and "public" (public holiday). Instead of the Chinese characters "night," "day," "early," "late," and "after," alphabets, numbers, and other characters may be used as shift symbols.

[0038] A shift schedule (work shift schedule) is a list showing the shift assignments for each employee for each day during a specified period, and is generated by arranging shift symbols based on rules. The specified period is usually 30 days, but is not limited to this. In other words, a shift schedule is a table that lists shift symbols for each employee for a specified period.

[0039] <Rule acquisition example 2: Extraction from nursing care fee standards> Figure 4 is a schematic diagram illustrating an example of a setting screen for items related to staffing system surcharges in nursing care fees. Staffing system surcharges are calculated when a nursing care facility has a more generous staffing system than the standard staffing standards. For example, the figure illustrates staffing standards (caregivers and nurses), staffing standards (care managers), staffing standards (life counselors), and staffing standards (nutritionists).

[0040] - Regarding personnel standards (caregiver / nursing staff), at least 14 full-time caregiver or nursing staff must be deployed.

[0041] - Regarding staffing standards (care manager), at least one person will be assigned full-time.

[0042] - Regarding staffing standards (life counselors), at least one person will be assigned full-time.

[0043] - Regarding staffing standards (nutritionists), at least one person will be assigned, either full-time or part-time.

[0044] The rule acquisition unit 211 can extract rules from the staffing system premium (nursing care fee standard) acquired by the facility. More specifically, the rule acquisition unit 211 acquires at least one of the staffing standards set in the staffing system premium as a rule.

[0045] <Server 200> Fig. 5 is a block diagram illustrating a schematic configuration of server 200. Fig. 6 is a functional block diagram illustrating the functions of CPU 210 shown in Fig. 5. Server 200 has CPU 210, ROM 220, RAM 230, storage 240, and communication interface 250. Each component is connected to each other via a network or a bus so that they can communicate with each other.

[0046] The configurations and basic functions of the CPU 210, ROM 220, RAM 230, storage 240, and communication interface 250 are similar to those of the corresponding elements of the information processing terminal 100, and therefore a duplicated description will be omitted.

[0047] Various computer programs including an automatic shift generation program for the CPU 210 to execute each process are installed in the storage 240. Rules received from the information processing terminal 100, etc., are also stored in the storage 240. The rules in the storage 240 can be added or updated by input from the information processing terminal 100, etc.

[0048] The CPU 210 executes the automatic shift generation program to realize various functions. As shown in Fig. 6, for example, the CPU 210 executes the automatic shift generation program to function as a rule acquisition unit 211, an individual generation unit 212, and an individual overwrite unit 213. Details of these functions will be described below.

[0049] <Method for automatically generating shifts using an automatic shift generation device> FIG. 7 is a flowchart illustrating the processing procedure of the automatic shift generation method by the automatic shift generation device in this embodiment. Also, FIG. 8 is a subroutine flowchart illustrating an example of the genetic operation processing (step S105) in FIG. 7. FIG. 9 is a schematic diagram illustrating the generation result of the initial population (first generation), and FIG. 10 is a schematic diagram illustrating the evaluation result of the individuals in the initial population of FIG. 9. FIG. 11 is a schematic diagram illustrating the execution result of natural selection / crossover / mutation by a genetic algorithm. FIG. 12 is a schematic diagram illustrating overwriting of individuals, and FIG. 13 is a schematic diagram showing an example of the evaluation result (second time) for each individual after overwriting in FIG. 12.

[0050] [Getting rules (step S101)] The rule acquisition unit 211 acquires rules related to shifts from the storage 240. For example, the rule acquisition unit 211 acquires rule A that there must be exactly one "night" shift symbol for each person each day, and rule B that there must be 10 or more "public" shift symbols for each employee.

[0051] [Generate shift table] In this embodiment, generating a shift table includes the following steps: generating an initial population (step S102), calculating an evaluation value for each individual (step S103), determining whether the evaluation value for each individual satisfies a predetermined standard (step S104), and performing genetic operations (step S105).

[0052] The individual generation unit 212 generates a shift table. The individual generation unit 212 generates a plurality of next-generation shift tables by evolving a plurality of shift tables as a plurality of individuals of the current generation into a plurality of next-generation shift tables using a genetic algorithm. The genetic algorithm is an algorithm that evolves a plurality of individuals of the current generation into a plurality of individuals of the next generation. Since the genetic algorithm itself is a known algorithm, an outline of the evolution of individuals by the genetic algorithm will be described below in association with the automatic shift generation method of this embodiment.

[0053] [Generation of initial population (step S102)] As shown in FIG. 9, the individual generation unit 212 randomly generates an initial population (first generation) consisting of multiple individuals. In this embodiment, the initial population consists of, for example, 100 individuals. One individual is, for example, a 30-day shift schedule for two staff members, staff member 1 and staff member 2. Therefore, the number of dimensions n of one individual is 2 × 30 = 60. Shift symbols are randomly allocated in the shift schedule. In this embodiment, an example is shown in which there are five assignable shift symbols: "early," "day," "late," "night," and "public." In other words, the number of types T of shift symbols is five. Note that the number of assignable shift symbols is not limited to the above five cases, and may be six or more.

[0054] [Calculation of evaluation value for each individual (step S103)] The individual generation unit 212 calculates an evaluation value for each individual. The evaluation value is a value that indicates the degree to which each individual has achieved (satisfied) each rule; in this embodiment, the closer the evaluation value is to 1.0 (100%), the higher the degree to which the rule has been achieved. The evaluation value E for each individual is calculated, for example, by averaging the sum of the evaluation values ​​Er for each rule. For example, if the evaluation values ​​for rule A and rule B are Era and Erb, respectively, the evaluation value E for each individual can be calculated by (Era + Erb) / 2. The evaluation value Era for rule A can be, for example, the percentage of days when rule A is satisfied. Furthermore, the evaluation value Erb for rule B can be the average for each employee of the percentage of "public" employees who satisfy rule B.

[0055] For example, as shown in Figure 10, for individual 1 of the first generation, rule A is satisfied 5 days out of 30, so the evaluation value Era for rule A is 5 / 30 = 0.16. Furthermore, since "public" is assigned twice to the first staff member and "public" is also assigned twice to the second staff member, the evaluation value Erb for rule B is (0.20 + 0.20) / 2 = 0.20. Therefore, the evaluation value E1 for individual 1 is (0.16 + 0.20) / 2 = 0.18. For individual 2, the evaluation value Era for rule A, the evaluation value Erb for rule B, and the evaluation value E2 for individual 2 are calculated in the same way as for individual 1.

[0056] The calculation method of the evaluation value Er for each rule and the evaluation value E for each individual is not limited to the above-mentioned calculation method. For example, the evaluation value E for each individual may be the maximum value, minimum value, or median value of the evaluation values ​​Er for each rule.

[0057] The individual generation unit 212 can output (display) the evaluation value E for each individual and the evaluation value Er for each rule via the operation display unit 160, for example.

[0058] [Determining whether the evaluation value for each individual satisfies a predetermined standard (Step S104)] The individual generation unit 212 determines whether the evaluation value for each individual satisfies a predetermined criterion as a termination condition for the genetic algorithm. For example, the individual generation unit 212 calculates a representative evaluation value based on the evaluation value for each individual, and determines whether the representative evaluation value is equal to or greater than a predetermined first threshold. That is, if the representative evaluation value is equal to or greater than the predetermined first threshold, the evaluation value for each individual satisfies the predetermined criterion, and if the representative evaluation value is less than the predetermined first threshold, the evaluation value for each individual does not satisfy the predetermined criterion. The representative evaluation value may be, for example, the average, maximum, minimum, or median of the evaluation values ​​of all individuals.

[0059] If the evaluation value for each individual satisfies a predetermined standard, that is, if the representative evaluation value is equal to or greater than a predetermined first threshold (step S104: YES), the individual generation unit 212 determines that the individuals have converged to a solution that satisfies the termination condition, and terminates the process (END). This stops the evolution of individuals using the genetic algorithm.

[0060] On the other hand, if the evaluation value for each individual does not satisfy the predetermined standard, i.e., if the representative evaluation value is less than the predetermined first threshold (step S104: NO), the individual generation unit 212 proceeds to the process of performing genetic operations (step S105). The individual generation unit 212 controls so that the evolution of the individuals continues using the genetic algorithm until the representative evaluation value becomes sufficiently large. In other words, the individual generation process continues until the individuals converge to a solution that satisfies the termination condition.

[0061] Alternatively, the individual generation unit 212 may be configured to determine whether the rate of change (rate of increase) ε of the representative evaluation value between generations for individuals of the second generation and later is equal to or less than a predetermined second threshold. That is, if the rate of change of the representative evaluation value is equal to or less than the predetermined second threshold, the evaluation value for each individual is deemed to satisfy the predetermined criterion, and if the representative evaluation value exceeds the predetermined second threshold, the evaluation value for each individual is deemed not to satisfy the predetermined criterion. For example, the rate of change ε ​​of the representative evaluation value between the kth generation and the (k-1)th generation may be defined as (representative evaluation value of the kth generation - representative evaluation value of the (k-1)th generation) / representative evaluation value of the kth generation.

[0062] If the evaluation value for each individual satisfies a predetermined standard, that is, if the rate of change ε ​​of the representative evaluation value is equal to or less than a predetermined second threshold (step S104: YES), the individual generation unit 212 determines that the individuals have converged to a solution that satisfies the termination condition, and terminates the process (END). This stops the evolution of individuals using the genetic algorithm.

[0063] On the other hand, if the evaluation value for each individual does not satisfy the predetermined standard, i.e., if the rate of change ε ​​of the representative evaluation value exceeds a predetermined second threshold (step S104: NO), the individual generation unit 212 proceeds to the process of performing genetic operations (step S105). The individual generation unit 212 controls so that evolution is repeated using the genetic algorithm until the change in the representative evaluation value between generations becomes sufficiently small. In other words, the individual generation process continues until the individuals converge to a solution that satisfies the termination condition.

[0064] The termination condition may be a combination of determining the magnitude of the representative evaluation value and determining the rate of change of the representative evaluation value between generations. Furthermore, the termination condition may be configured to limit the number of iterations of evolution by the genetic algorithm to a predetermined number.

[0065] The individual generation unit 212 may be configured to output (display) the converged individuals. For example, the individual generation unit 212 may display all of the converged individuals or the top m highest-rated individuals among the converged individuals.

[0066] [Performing Genetic Operations (Step S105)] The individual generation unit 212 performs at least one of the genetic operations of natural selection, crossover, and mutation on the multiple individuals in the initial population or the multiple individuals overwritten in the overwriting of the shift symbols on the individuals (step S106). Fig. 8 shows an example of a typical procedure for the genetic operation, but the procedure is not limited to this example.

[0067] 8, as an example, the individual generation unit 212 may serially perform the genetic operations of natural selection, crossover, and mutation. That is, the individual generation unit 212 performs natural selection on multiple individuals in the initial population or multiple overwritten individuals, performs crossover on highly evaluated individuals to generate multiple new individuals (next generation), and then performs mutation on these new multiple individuals. Specific examples of the genetic operations of natural selection, crossover, and mutation are described below.

[0068] [Implementation of natural selection (Step S201)] As shown in FIG. 11 , the individual generation unit 212 performs natural selection on multiple individuals in the initial population or multiple individuals overwritten with shift symbols in the overwriting of individuals (step S106). Natural selection, also called selection, selects individuals whose shift symbol patterns will remain in the next generation from multiple individuals in the current generation based on the evaluation value of each individual in the current generation. That is, natural selection transfers the shift symbol patterns of selected individuals from the multiple individuals in the current generation to individuals in the next generation. Generally, in natural selection, the shift symbol patterns of highly evaluated individuals are selected to remain in the next generation. The automatic shift symbol generation device of this embodiment can use methods such as roulette selection, ranking selection, tournament selection, and elite selection as selection methods based on evaluation values. Note that these selection methods are well known and will not be described in detail.

[0069] [Crossover execution (step S202)] The individual generation unit 212 performs crossover on a plurality of individuals that have been subjected to natural selection. Crossover is a genetic operation that generates new shift symbol patterns in individuals of the next generation by rearranging shift symbols between selected individuals of the current generation with a predetermined crossover probability. The automatic shift generation device of this embodiment can employ one-point, two-point, multi-point crossover, and uniform crossover. Note that crossover methods are well known, so detailed explanations will be omitted.

[0070] [Implementing mutation (step S203)] The individual generation unit 212 performs mutation on the new individuals obtained by the crossover. Mutation is a genetic operation that changes some shift symbols in the selected shift with a predetermined mutation probability.

[0071] [Overwriting shift symbols onto individuals (step S106)] Returning to FIG. 7 again, the individual overwriting unit 213 overwrites shift symbols for all or some of the multiple shift tables generated by the individual generation unit 212 so as to satisfy the rules acquired by the rule acquisition unit 211. In this embodiment, for simplicity of explanation, a case will be described in which, of rules A and B acquired by the rule acquisition unit 211, only rule A is the target for overwriting individuals. As shown in FIG. 12, in individuals 1 to 100 before overwriting, exactly one "night" shift symbol is not necessarily allocated for each day. After overwriting in which rule A, which requires exactly one "night" shift symbol for each day, is applied, exactly one "night" shift symbol is allocated for each day. After overwriting with shift symbols for each individual as the individual overwriting unit 213, the CPU 210 proceeds to the processing of step S103 and functions again as the individual generation unit 212.

[0072] [Calculation of evaluation value for each individual (2nd time)] The individual generation unit 212 calculates an evaluation value for each individual (step S103). For example, as shown in FIG. 13, for individual 1 of the first generation, rule A is satisfied 30 days out of 30, so the evaluation value Era for rule A is 30 / 30 = 1.0. Furthermore, since "public" is assigned once for employee 1 and "public" is assigned twice for employee 2, the evaluation value Erb for rule B is (0.10 + 0.20) / 2 = 0.15. Therefore, the evaluation value E1 for individual 1 is (1.0 + 0.15) / 2 = 0.58. For individual 2, the evaluation value Era for rule A, the evaluation value Erb for rule B, and the evaluation value E2 for individual 2 are calculated in the same way as for individual 1.

[0073] In this way, the evaluation value Era of rule A becomes 1 in each of the individual 1 and the individual 2, and therefore the evaluation value E of the individual improves.

[0074] In the above, only rule A is the target for overwriting individuals, but it is also possible to configure rule B to be overwritten after rule A is overwritten. This causes the evaluation value Erb of rule B to become 1 in each individual, further improving the evaluation value E of the individual.

[0075] As described above, in the processing of the flowchart shown in Fig. 7, the rule acquisition unit 211 acquires rules related to shifts. Subsequently, the individual generation unit 212 uses a genetic algorithm to evolve multiple shift tables as multiple individuals of the current generation into multiple shift tables of the next generation based on the rules, thereby generating multiple shift tables of the next generation. Then, the individual overwriting unit 213 overwrites all or some of the multiple shift tables generated so as to satisfy the rules.

[0076] <Reducing the number of possible combination patterns of an individual by overwriting> FIG. 14 is a schematic diagram illustrating the number of combination patterns of shift symbols when the rules are applied.

[0077] As described above, in a genetic algorithm, the process of generating individuals continues until the process converges to a solution that satisfies the termination condition. The time it takes for the genetic algorithm to converge to a solution that satisfies the termination condition (hereinafter referred to as the "termination time of the genetic algorithm" or simply "termination time") depends on the number of combination patterns that the individuals can take (hereinafter also referred to as the "number of combination patterns"). Since the amount of calculation increases as the number of combination patterns increases, the termination time of the genetic algorithm increases as the number of combination patterns increases.

[0078] If no overwriting is performed, as mentioned above, the number of dimensions n of one individual is 30 × 2 = 60, and the number of types of shift symbols T is 5, so the number of combination patterns is 5. 60 It's a pattern.

[0079] On the other hand, if we overwrite the shift code for employee 1 and employee 2 according to rule A, which requires exactly one "night" shift code for each day, the number of shift code combinations for employee 1 and employee 2 is narrowed down to eight patterns per day. Therefore, the number of combination patterns for 30 days is 8. 30 It's a pattern.

[0080] When comparing the number of combination patterns between not overwriting and overwriting, the number is 5. 60 Pattern / 8 30 Pattern ≒ 10 14.85 Therefore, the number of combination patterns is 1 / 10 14 Therefore, by overwriting, the completion time is significantly reduced.

[0081] <Measurement results of individual generation using genetic algorithms> Fig. 15 is a graph illustrating the evaluation values ​​for each generation in the cases with and without overwriting. Fig. 16 is a diagram illustrating the measured values ​​of the end time of the genetic algorithm in Fig. 15 and the evaluation values ​​at the end time.

[0082] The automatic shift generation device compared the evaluation value of the current generation with the evaluation value of the 1000th generation every 1000 generations, and if there was a change, continued generating individuals.If there was no change or the change was very small, it was assumed that the solution had converged to satisfy the termination conditions and the process was terminated.

[0083] In generating the initial population (step S102), the shift symbols of the individuals in the initial population (first generation) are randomly assigned. As shown in FIG. 15, in the case of overwriting, the individuals in the second generation, in which the genetic operation was performed on the first generation, were overwritten according to rule A, which requires exactly one "night" shift symbol per day. Therefore, the evaluation values ​​of the individuals in the second generation were high, exceeding 0.9. On the other hand, in the case of not overwriting, the evaluation values ​​of the individuals in the second generation, in which the genetic operation was performed on the first generation, were approximately 0.7. The evaluation value of 0.7 in the case of not overwriting is smaller than the evaluation value of 0.9 in the case of overwriting. The evaluation values ​​in both the case of overwriting and the case of not overwriting increased as the generational change progressed, and converged when the termination condition was met. However, the rate of increase in the evaluation value with respect to the generational change was faster with overwriting than without overwriting, and the time to convergence was shorter.

[0084] As shown in Figure 16, without overwriting, the completion time was 3514.6 seconds. The number of generations at the completion time was 21000, and the evaluation value was 0.9815. On the other hand, with overwriting, the completion time was 971.5 seconds. The number of generations at the completion time was 5000, and the evaluation value was 0.9938.

[0085] In this way, it was confirmed that the completion time can be significantly shortened by overwriting individuals in the actual measurement results of individual generation using a genetic algorithm.

[0086] The automatic shift generation device, automatic shift generation method, and automatic shift generation program of the present embodiment described above overwrite the individual so that it satisfies the acquired rules in the process of evolving the individual using a genetic algorithm. This limits the number of patterns that the individual can take. This reduces the amount of calculation required in the genetic algorithm. As a result, the time it takes for the individual to converge to an optimal solution can be shortened.

[0087] (Second embodiment) In the second embodiment, a case will be described in which rule priorities are further set and shift symbols are overwritten on individuals in an order according to the rule priorities. The rule priorities represent, for example, the relative priority relationships among multiple rules. In the second embodiment, detailed descriptions of the same processes and functions as those in the first embodiment will be omitted to avoid duplication.

[0088] Fig. 17 is a flowchart illustrating the processing steps of the automatic shift generation method in the second embodiment. Each process in the diagram is realized by CPU 210 executing the automatic shift generation program. Steps S101 to S106 in Fig. 17 are the same as the processes S101 to S106 in the flowchart of Fig. 7 of the first embodiment. In the second embodiment, compared to the first embodiment, a process (step S107) for setting the priority of the rules is added between steps S101 and S102.

[0089] 18 is a functional block diagram illustrating the functions of the CPU 210 of the server 200 in the second embodiment. Each function in the diagram is realized by the CPU 210 executing an automatic shift generation program. The CPU 210 functions as a rule acquisition unit 211, an individual generation unit 212, an individual overwriting unit 213, and a priority setting unit 214. The functions of the rule acquisition unit 211, the individual generation unit 212, and the individual overwriting unit 213 are the same as the functions of the rule acquisition unit 211, the individual generation unit 212, and the individual overwriting unit 213 in FIG. 6 of the first embodiment.

[0090] [Setting rule priority (Step S107)] The priority setting unit 214 sets priorities among a plurality of rules. For example, default values ​​for the priority of each rule are stored in advance in the storage 240. The priority of a rule is expressed by a number, for example, from 1 to 10, depending on the degree of priority desired by the user. In this embodiment, for example, a rule with a smaller number is given priority over a rule with a larger number, but conversely, a rule with a larger number may be configured to be given priority over a rule with a smaller number.

[0091] The rule acquisition unit 211 acquires, for example, the "number of public holidays" rule, the "number of people required per shift" rule, the "upper limit of consecutive work hours" rule, and the "desired work hours" rule.

[0092] The priority setting unit 214 reads from the storage 240 the priorities corresponding to the multiple rules acquired by the rule acquisition unit 211. For example, the priority setting unit 214 reads from the storage 240 priorities 4 to 1 for the "number of public holidays" rule, the "number of people required per shift" rule, the "maximum number of consecutive work days" rule, and the "desired work" rule, and sets these as the priorities among the multiple rules.

[0093] [Overwriting shift symbols onto individuals (step S106)] 19 is a schematic diagram illustrating an example of the overwriting order in the process of overwriting shift symbols to individuals. For example, in this embodiment, a case where overwriting is performed according to the "number of public holidays" rule, the "number of people required for a shift" rule, the "maximum number of consecutive work shifts" rule, and the "desired work schedule" rule will be described.

[0094] Since the later overwrite takes precedence over the earlier overwrite, the higher the priority of the rule, the later the overwrite will be performed. In the example shown in Figure 19, the overwrite is performed in the order of the "number of public holidays" rule, the "number of people required for a shift" rule, the "limit number of consecutive work shifts" rule, and the "desired work" rule.

[0095] <User-input priority setting> Fig. 20 is a schematic diagram illustrating an example of setting priority levels by user input. Fig. 21 is a schematic diagram illustrating an example of an overwriting order when priority levels are set by user input and shift symbols are overwritten on individuals.

[0096] As shown in FIG. 20, the user can set rule names corresponding to priorities via the operation and display unit 160 on the "Priority Setting" screen displayed on the information processing terminal 100. For example, the figure shows an example in which a "Number of people required for a shift" rule is set for priority 1, a "Maximum number of consecutive shifts" rule is set for priority 2, a "Desired work" rule is set for priority 3, and a "Number of public holidays" rule is set for priority 4. Rule names can also be set for priorities 5 and above as needed. The user can also set priorities for each individual individual via the operation and display unit 160.

[0097] In the example shown in Figure 20, the priority is highest for the "number of people required for a shift" rule, "limit number of consecutive work days" rule, "desired work" rule, and "number of public holidays" rule. Therefore, as shown in Figure 21, overwriting of shift symbols onto individuals is performed in the order of the "number of public holidays" rule, the "desired work" rule, the "limit number of consecutive work days" rule, and the "number of people required for a shift" rule.

[0098] In this way, the automatic shift generation device of this embodiment is configured so that priorities can be set by user input, so the user can set priorities as desired.

[0099] <Setting priority based on evaluation value> 22 is a schematic diagram illustrating an example of an overwriting order when priorities are set based on evaluation values ​​and shift symbols are overwritten on individuals. For example, the priority setting unit 214 can set priorities for individuals of the next generation based on evaluation values ​​for individuals of the current generation.

[0100] For example, in the example shown in the figure, for the k-th generation individuals, the priorities of the "desired work schedule", "maximum number of consecutive work shifts", "number of people required for a shift", and "number of public holidays" rules are each set to 1 to 4. Also, assume that the evaluation values ​​for the "desired work schedule", "maximum number of consecutive work shifts", "number of people required for a shift", and "number of public holidays" rules are 0.9, 0.9, 0.5, and 0.1, respectively.

[0101] The priority setting unit 214 may, for example, increase the priority of a rule with a low evaluation value and change the priorities so as to improve the balance of the evaluation values ​​between the rules. For example, since the evaluation value of the "number of public holidays" rule is lower than the evaluation value of the "number of employees required for a shift" rule, the priority setting unit 214 may swap the priorities of these rules, setting the priority of the "number of public holidays" rule to 3 and the priority of the "number of employees required for a shift" rule to 4.

[0102] The individual overwrite unit 213 performs overwriting in the order of the "number of people required for a shift" rule, the "number of public holidays" rule, the "limit number of consecutive work days" rule, and the "desired work" rule, according to the priority changed by the priority setting unit 214.

[0103] In this way, by increasing the priority of rules with low evaluation values ​​and improving the balance of evaluation values ​​between each rule before overwriting the individual, the overall evaluation value of the individual for each rule can be improved as generations change.

[0104] <Setting priority for each individual> 23 is a schematic diagram illustrating an example of an overwriting order when a priority is set for each individual and a shift symbol is overwritten on the individual. The priority setting unit 214 can set a priority for each individual based on the evaluation value of each rule.

[0105] For example, in the example shown in the figure, rule A and rule B are acquired by the rule acquisition unit 211 for individual 1 and individual 2 of the current generation, respectively. Also, assume that for individual 1, the evaluation value of rule A is 0.2 and the evaluation value of rule B is 0.8, and for individual 2, the evaluation value of rule A is 1.00 and the evaluation value of rule B is 0.15.

[0106] The priority setting unit 214 may, for example, increase the priority of a rule with a low evaluation value and change the priorities so that the evaluation values ​​of the rules are well balanced. For example, for individual 1, the evaluation of rule A is lower than the evaluation of rule B, so the priority setting unit 214 changes the priority so that the priority of rule A is higher than the priority of rule B. Furthermore, for example, for individual 2, the evaluation of rule B is lower than the evaluation of rule A, so the priority setting unit 214 changes the priority so that the priority of rule B is higher than the priority of rule A.

[0107] In this way, by setting a priority for each individual and overwriting the individual with a shift symbol, it is possible to improve the evaluation value for each individual.

[0108] <Setting priorities based on evaluation values ​​of multiple generations> The priority setting unit 214 can set the priority of each rule in the next generation based on, for example, the evaluation value of the generation before the current generation. For example, the priority setting unit 214 sets the priority for individuals in the next generation based on the difference in evaluation value between generations. For example, the priority setting unit 214 can set the priority of each rule in the next generation based on the difference between the evaluation value of the current generation and the evaluation value of the generation before the current generation.

[0109] <User input determines the percentage of individuals to be overwritten> 24 is a schematic diagram illustrating a case where the ratio of individuals to be overwritten is set by user input and shift symbols are overwritten onto individuals. The individual overwriting unit 213 can set the ratio of individuals to be overwritten among multiple individuals based on user input.

[0110] For example, if an individual has a low evaluation value, overwriting more individuals can raise the overall evaluation value of the individuals. On the other hand, the greater the proportion of individuals overwritten, the greater the load of the process of overwriting the individuals. Furthermore, a large proportion of individuals overwritten increases the likelihood of falling into a local solution. Therefore, it cannot be said that a large proportion of individuals overwritten is always good. The optimal proportion is set by taking into consideration the benefits of improving the evaluation value of the individuals, the increased load of the process of overwriting the individuals, and the risk of falling into a local solution.

[0111] <Set the percentage of individuals to overwrite based on the evaluation value> 25 is a schematic diagram illustrating a case where the proportion of individuals to be overwritten is set based on the evaluation value, and shift symbols are overwritten onto individuals. The individual overwriting unit 213 can set the proportion of individuals to be overwritten with shift symbols out of multiple individuals based on, for example, the evaluation value of individuals in the current generation.

[0112] For example, if the evaluation value of an individual in the current generation is low, the individual overwriting unit 213 increases the proportion of individuals to be overwritten in the next generation, and if the evaluation value of an individual in the current generation is high, the individual overwriting unit 213 decreases the proportion of individuals to be overwritten in the next generation. By overwriting more individuals in the next generation, the evaluation value of the individuals can be raised overall.

[0113] For example, suppose the rate at which overwriting processing is performed for 100 individuals of the first generation is initially set to 0.3. If the average evaluation value of the 100 individuals of the first generation is a relatively low 0.3, the individual overwriting unit 213 increases the rate at which overwriting processing is performed for individuals of the second generation, setting it to 0.7. On the other hand, if the average evaluation value of the 100 individuals of the second generation improves to 0.8, the individual overwriting unit 213 decreases the rate at which overwriting processing is performed for individuals of the third generation, setting it to 0.2. Furthermore, if the average evaluation value of the 100 individuals of the third generation drops to 0.6, the individual overwriting unit 213 increases the rate at which overwriting processing is performed for individuals of the fourth generation, setting it to 0.4.

[0114] In this way, the individual overwriting unit 213 can be configured to control the proportion of individuals to be overwritten in the next generation according to the evaluation value of the individuals in the current generation, thereby improving the evaluation value while suppressing an increase in the amount of calculation in the genetic algorithm.

[0115] The automatic shift generation device, the automatic shift generation method, and the automatic shift generation program of the present embodiment described above can further achieve the following effects in addition to the effects of the first embodiment.

[0116] In the process of evolving an individual using a genetic algorithm, the individual is overwritten by a rule with a higher priority. As a result, the shift symbol arrangement corresponding to the rule with a higher priority is included in the overwritten individual, and the number of patterns that the individual can take is more effectively limited. Therefore, the amount of calculation in the genetic algorithm is effectively reduced. As a result, the automatic shift generation device can further shorten the time it takes for the individual to converge to the optimal solution.

[0117] The configurations of the automatic shift generation device, automatic shift generation method, and automatic shift generation program described above are the main configurations described in explaining the features of the above-mentioned embodiments, and are not limited to the above-mentioned configurations, and can be modified in various ways within the scope of the claims.

[0118] For example, the information processing terminal 100 and the server 200 may each include components other than those described above, or may not include some of the components described above.

[0119] Furthermore, the information processing terminal 100 and the server 200 may be configured as separate devices, or may be configured as a single device.

[0120] Furthermore, the functions of each component may be realized by other components. For example, at least a part of the processing described as being executed by the server 200 may be executed by the information processing terminal 100 or another device. Furthermore, by incorporating the functions of the information processing terminal 100 into the server 200, the information processing terminal 100 may be omitted.

[0121] Furthermore, the processing units of the flowcharts in the above embodiments are divided according to the main processing content to make each process easier to understand. The method of dividing the processing steps does not limit the present invention. Each process can be divided into more processing steps. One processing step may execute more processes. Furthermore, multiple steps may be executed simultaneously or in combination.

[0122] The means and methods for performing various processes in the systems according to the above-described embodiments can be realized by either dedicated hardware circuits or a programmed computer. The programs may be provided, for example, on a computer-readable recording medium such as a flexible disk or CD-ROM, or online via a network such as the Internet. In this case, the programs recorded on the computer-readable recording medium are typically transferred to and stored in a storage unit such as a hard disk. The programs may also be provided as standalone application software or may be incorporated into the software of the device as a function of the system.

[0123] While embodiments of the present invention have been described and illustrated in detail, the disclosed embodiments are made for purposes of illustration and example only, and not limitation, and the scope of the present invention should be construed by the language of the appended claims. [Explanation of symbols]

[0124] 100 information processing terminal, 110 CPUs, 120 ROM, 130 RAM, 140 storage, 150 communication interface, 160 Operation display section, 200 servers, 210 CPUs, 211 Rule Acquisition Department, 212 Individual generation section, 213 Individual Supervisory Section, 214 Priority setting section, 220 ROM, 230 RAM, 240 storage, 250 communication interface.

Claims

1. a rule acquisition unit that acquires rules regarding shifts; an individual generation unit that generates a plurality of next-generation shift tables by evolving a plurality of shift tables as a plurality of current-generation individuals into a plurality of next-generation shift tables using a genetic algorithm that evolves a plurality of current-generation individuals into a plurality of next-generation individuals; an individual overwrite unit that overwrites all or some of the plurality of shift tables generated by the individual generation unit so that the rules are satisfied;

2. a priority setting unit that sets priorities among the rules; The automatic shift generation device according to claim 1 , wherein the individual overwriting unit determines an overwriting order according to the rule in accordance with the priority.

3. The automatic shift generation device according to claim 2 , wherein the priority setting unit sets the priorities based on a user input.

4. the individual generation unit calculates an evaluation value for each individual; The automatic shift generation device according to claim 2 , wherein the priority setting unit sets the priority for the individuals of the next generation based on the evaluation value for the individuals of the current generation.

5. The automatic shift generation device according to claim 3 or 4, wherein the priority setting unit sets the priority for each of the individuals.

6. The automatic shift generation device according to claim 1 or 2, wherein the individual overwriting unit sets a ratio of the individuals to be overwritten among the plurality of individuals based on a user input.

7. 3. The automatic shift generation device according to claim 1, wherein the individual overwriting unit sets a ratio of the individuals to be overwritten among the plurality of individuals based on evaluation values ​​of the individuals of the current generation.

8. the individual generation unit calculates an evaluation value for each individual; The automatic shift generation device according to claim 2 , wherein the priority setting unit sets the priority for the individuals of the next generation based on a difference in the evaluation value between generations.

9. The automatic shift generation device according to claim 1 or 2, wherein the rule acquisition unit extracts the rules from nursing care remuneration standards acquired by the facility.

10. The automatic shift generation device according to claim 1 or 2, wherein the plurality of rules include at least any of the following: the number of people required for early shifts, late shifts, or night shifts; the number of public holidays; the number of shifts each employee works; fixed shifts for each employee on a day of the week; desired work schedules; an upper limit on the number of consecutive shifts; consecutive shift patterns; discontinuous shift patterns; and the number of consecutive shifts with the same shift symbol.

11. (a) obtaining rules for shifts; (b) generating a plurality of next-generation shift tables by evolving a plurality of shift tables as a plurality of current-generation individuals into a plurality of next-generation individuals using a genetic algorithm that evolves a plurality of current-generation individuals into a plurality of next-generation individuals; and (c) overwriting all or some of the shift tables generated in the step (b) so as to satisfy the rule.

12. (a) a step of obtaining rules regarding shifts; a step (b) of generating a plurality of next-generation shift tables by evolving a plurality of current-generation individuals into a plurality of next-generation individuals using a genetic algorithm that evolves a plurality of current-generation individuals into a plurality of next-generation individuals; and (c) overwriting all or some of the shift schedules generated in the step (b) so as to satisfy the rule. An automatic shift generation program for causing a computer to execute a process including the steps.

Images