Schedule creation support device and schedule creation support method

Abstract

To support creating a desirable schedule under constraint conditions.SOLUTION: A schedule creation support device 100 is provided with: a reception unit 111 that receives a plurality of tasks, a constraint condition for generating a schedule that assigns execution dates and times to the plurality of tasks, metrics for evaluating the generated schedule, and a KPI; a constraint condition generation unit 112 that generates a plurality of generated constraint conditions by adding deviations to the constraint condition; a schedule generation unit 113 that generates optimal schedules that satisfy the generated constraint conditions; and a constraint evaluation unit 114 that outputs the generated constraint condition corresponding to the schedule out of the generated schedules for which the KPI is optimized.SELECTED DRAWING: Figure 1

Description

[Technical field]

[0001] The present invention relates to a schedule creation support device and a schedule creation support method that support the creation of a schedule for assigning execution dates and times to tasks. [Background technology]

[0002] Creating a maintenance schedule for a plant involves allocating (determining) the dates and times for performing individual tasks (work), such as inspecting and replacing each of the facilities, equipment, and parts that make up the plant, in a way that satisfies allocation conditions. Allocation conditions include the order in which a series of tasks must be performed, the deadline for a task, the number and skills of workers required to perform a task, and the equipment required for a task. In addition to these allocation conditions for individual tasks, there are other constraints for creating the schedule, such as the overall maintenance cost, the total number of workers, the deadline for completing maintenance, the working hours per day, and the number of pieces of equipment.

[0003] A schedule is created by solving an optimization problem that maximizes a schedule metric subject to allocation constraints and constraints, such as the total cost or the number of tasks that meet the allocation constraints.

[0004] Patent Documents 1 and 2 disclose technology for planning support of maintenance. The invention disclosed in Patent Document 1 relates to a technology for supporting maintenance of a plant, and in particular, it is possible to select and determine a system maintenance method based on a total effect index that weights the maintenance effect of facilities and equipment of a large-scale plant, and on the urgency of the system. The invention disclosed in Patent Document 2 makes it possible to create a maintenance plan while evaluating the trade-off between the risk of failure and the maintenance cost of the equipment to be maintained. [Prior art documents] [Patent documents]

[0005] [Patent Document 1] JP 2008-015775 A [Patent Document 2] JP 2009-217718 A Summary of the Invention [Problem to be solved by the invention]

[0006] According to the inventions of Patent Documents 1 and 2, a plan can be created that takes into account metrics such as risk and cost, but the conditions on which the plan was created remain the original constraint conditions. By changing the constraint conditions, it is possible to create a more desirable plan than before the changes. However, with the conventional technology, it is unclear how to change the constraint conditions. For this reason, the planner must repeatedly change the constraint conditions and evaluate the created plan. For this reason, it takes a lot of time to obtain a desirable plan.

[0007] The present invention has been made in view of the above background, and an object of the present invention is to provide a schedule creation support device and a schedule creation support method that enable support for creating a desirable schedule under constraint conditions. [Means for solving the problem]

[0008] In order to solve the above-mentioned problems, the schedule creation support device of the present invention includes a reception unit that receives a plurality of tasks, constraint conditions for schedule generation that assign execution dates and times to the plurality of tasks, metrics for evaluating the generated schedule, and a KPI, a constraint condition generation unit that generates a plurality of generation constraint conditions by adding deviations to the constraint conditions, a schedule generation unit that generates an optimal schedule that satisfies the generation constraint conditions, and a constraint evaluation unit that outputs, among the generated schedules, the generation constraint conditions corresponding to the schedule that provides the best KPI. Effect of the Invention

[0009] According to the present invention, it is possible to provide a schedule creation support device and a schedule creation support method that enable support for creating a desirable schedule under constraints. Problems, configurations, and effects other than those described above will become apparent from the description of the following embodiments. [Brief description of the drawings]

[0010] [Figure 1] 1 is a functional block diagram of a schedule creation support device according to an embodiment of the present invention. [Diagram 2] 11 is a flowchart of a constraint evaluation process according to the embodiment. [Diagram 3] 13 is a graph of the score S(t) according to a modified example of the present embodiment. [Figure 4] FIG. 13 is a diagram illustrating a screen configuration of a constraint evaluation screen according to a modified example of the present embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0011] A schedule creation support device in a form (embodiment) for carrying out the present invention will be described below. The schedule creation support device generates multiple generation constraint conditions by adding deviations to given constraint conditions. Next, the schedule creation support device assigns execution dates and times to tasks under the generation constraint conditions to generate a schedule. The constraint conditions include the overall cost and deadline for executing the tasks, the total number of workers, the working hours per day, the number of pieces of equipment required to execute the tasks, etc.

[0012] Next, the schedule creation support device performs a principal component analysis on the generated schedule using the metrics as variables, and performs a clustering analysis of the schedule based on the obtained principal components. Furthermore, the schedule creation support device identifies the schedule with the best KPI (Key Performance Indicator) among the schedules closest to the center of the cluster, and identifies the generation constraint conditions corresponding to the schedule.

[0013] Here, best means the best or the best range. Best means, for example, the best KPI value, and the best range means, for example, the KPI value that can be considered almost equivalent to the best KPI. The best range is, for example, the top 10% range of the KPI. Metrics will be described later.

[0014] A user of the schedule creation support device can learn the impact of the constraints on the schedule metrics by generating a schedule by changing the constraints with reference to the identified generation constraints. For example, the user can learn that by increasing the number of workers, which is a constraint, by 5%, the number of tasks that are completed by the desired deadline, which is a metric, will increase by 15%. By understanding the impact of the constraints on the metrics, the user can obtain a desired schedule by adjusting the constraints.

[0015] Configuration of the schedule creation support device 1 is a functional block diagram of a schedule creation support device 100 according to this embodiment. The schedule creation support device 100 is a computer, and includes a control unit 110, a storage unit 120, and an input / output unit 180. User interface devices such as a display, a keyboard, and a mouse are connected to the input / output unit 180. The input / output unit 180 may include a communication device and may be capable of transmitting and receiving data to and from other devices. A media drive may also be connected to the input / output unit 180, enabling data exchange using a recording medium.

[0016] <Schedule Creation Support Device: Memory Unit> The storage unit 120 includes storage devices such as a Read Only Memory (ROM), a Random Access Memory (RAM), a Solid State Drive (SSD), etc. The storage unit 120 stores a task database 130, a condition database 140, an evaluation value database 150, a schedule database 160, and a program 128. The program 128 includes a description of a constraint evaluation process (see FIG. 2) to be described later.

[0017] The task database 130 stores information related to a task, such as the name, the number of workers required to execute the task, their skills, priority, the execution cycle (interval) of a task that is repeatedly executed, and the execution deadline (the allowable time that elapses from the last execution date and time until the next execution of the task).

[0018] Condition database 140 stores conditions for allocating execution dates and times to tasks. Conditions (allocation conditions) related to individual tasks include the order in which a series of tasks are to be executed, the number and skills of workers required to execute a task, equipment required for a task, etc. In addition to such allocation conditions, there are constraints on schedule creation, such as the overall cost of executing the tasks, the number of equipment, the number of workers, and the number of working hours per day.

[0019] The evaluation value database 150 stores calculation methods (calculation formulas) for the metrics of the schedule. The metrics include the cost of performing the tasks as a whole, the number of tasks that are late to their deadlines, the number of tasks that meet the allocation conditions, the total amount of delay (risk) of tasks that are late to their deadlines, etc. The evaluation value database 150 also stores calculation methods for KPIs that are calculated based on the metrics. The schedule database 160 stores the generated schedule (task execution dates and times) in association with generation constraint conditions, which will be described later.

[0020] <Schedule creation support device: control unit> The control unit 110 includes a CPU (Central Processing Unit), and includes a reception unit 111, a constraint condition generation unit 112, a schedule generation unit 113, and a constraint evaluation unit 114. The receiving unit 111 receives tasks, allocation conditions, constraint conditions, and formulas for calculating metrics and KPIs, and stores them in a task database 130 , a condition database 140 , and an evaluation value database 150 .

[0021] The constraint condition generating unit 112 generates generated constraint conditions by adding deviations to the accepted constraint conditions. For example, assume that the number of workers is 50, the working hours per day are 7, and the total amount of delay is 260 as the accepted constraint conditions. The constraint condition generating unit 112 sets the number of workers with deviations to, for example, 48, 49, 50, 51, and 52. The constraint condition generating unit 112 also sets the working hours per day with deviations to, for example, 6.8, 6.9, 7, 7.1, and 7.2. The constraint condition generating unit 112 also sets the total amount of delay with deviations to, for example, 250, 255, 260, 265, and 270. Then, the constraint condition generating unit 112 generates a total of 125 combinations of these constraint conditions with deviations, namely (48, 6.8, 250), (48, 6.8, 255), ..., and (52, 7.2, 270), as generated constraint conditions.

[0022] In the above example, the constraints with deviations are at equal intervals, but may be at random intervals. In other words, the constraint generator 112 may generate the constraints with deviations set to random numbers.

[0023] The schedule generation unit 113 generates a schedule by allocating execution dates and times to tasks under the allocation conditions and generation constraint conditions. In more detail, the schedule generation unit 113 allocates execution dates and times to tasks by solving an optimization problem with the objective of optimizing the KPI under the allocation conditions and generation constraint conditions.

[0024] The constraint evaluation unit 114 performs a principal component analysis using the metrics of the schedule as variables, and performs a clustering analysis of the schedule based on the obtained principal components. Next, the constraint evaluation unit 114 identifies a schedule with the best KPI among the schedules closest to the center of the cluster, and identifies the generation constraint conditions corresponding to the schedule. The best KPI may be the best KPI, or may be, for example, the top 10% KPI (the best range of KPIs).

[0025] As described above, the schedule creation support device 100 includes a reception unit 111 that receives a plurality of tasks, constraint conditions for schedule generation that assign execution dates and times to the plurality of tasks, metrics for evaluating the generated schedule, and KPIs. The schedule creation support device 100 also includes a constraint condition generating unit 112 that generates a plurality of generated constraint conditions by adding deviations to the constraint conditions.

[0026] The schedule creation support device 100 includes a schedule creation unit 113 that creates an optimal schedule that satisfies creation constraint conditions. Furthermore, the schedule creation support device 100 is provided with a constraint evaluation unit 114 that performs a principal component analysis of the metrics of the generated schedule, performs a clustering analysis of the schedule based on the obtained principal components, and outputs the generation constraint conditions corresponding to the schedule with the best KPI among the schedules closest to the center of the cluster.

[0027] Constraint evaluation process 2 is a flowchart of the constraint evaluation process according to this embodiment. It is assumed that the accepting unit 111 has already accepted tasks, allocation conditions, constraint conditions, metrics, and KPIs at the time the constraint evaluation process starts. In step S11, the constraint generating unit 112 generates generated constraint conditions by combining the constraint conditions received by the receiving unit 111 with deviation-added constraint conditions by adding deviations thereto.

[0028] In step S12, the schedule generating unit 113 starts the process of repeating step S13 for each generation constraint condition generated in step S11. In step S13, schedule generating unit 113 generates a schedule by allocating execution dates and times to tasks under the allocation conditions and generation constraint conditions. Next, schedule generating unit 113 stores the generated schedule in schedule database 160 in association with the generation constraint conditions.

[0029] In step S14, the constraint evaluation unit 114 performs principal component analysis using the metrics of the schedule as variables to calculate principal components. In step S15, the constraint evaluation unit 114 performs a clustering analysis of the schedule based on the distances based on the principal components calculated in step S14, and obtains clusters.

[0030] In step S16, the constraint evaluation unit 114 starts the process of repeating step S17 for each cluster obtained in step S15. In step S17, the constraint evaluation unit 114 calculates the KPI of the schedule that is closest to the center of the cluster.

[0031] In step S18, the constraint evaluation unit 114 identifies a schedule with the best KPI among the KPIs calculated in step S17. Next, the constraint evaluation unit 114 refers to the schedule database 160, acquires and outputs the generation constraint conditions corresponding to the schedule.

[0032] <Features of the schedule creation support device> The schedule creation support device 100 creates a schedule under generation constraint conditions that are the specified constraint conditions (accepted by the accepting unit 111) plus a deviation. Next, the schedule creation support device 100 performs a clustering analysis and outputs generation constraint conditions corresponding to a schedule that will have the best KPI among schedules close to the cluster center. These generation constraint conditions are close to the specified constraint conditions and are constraint conditions that will result in a better KPI than the KPI of the schedule created under the specified constraint conditions.

[0033] A user of the schedule creation support device 100 can refer to such generated constraint conditions and modify the original constraint conditions (specified constraint conditions) to create a schedule, thereby learning the influence of the constraint conditions on the metrics of the schedule. As a result, the user can obtain a desired schedule by adjusting the constraint conditions.

[0034] Variation: Task Score Some tasks are executed periodically (at a predetermined interval). Examples of such tasks are inspections and part replacements. For such tasks, it is desirable to assign an execution date and time to the time interval following the last execution of the task, and it is undesirable to assign an execution date and time that is too early or too late.

[0035] As a metric for a schedule including such a task, a score may be set for the execution dates and times of the task, with the execution dates and times being at a predetermined interval being the highest value. The schedule generating unit 113 assigns execution dates and times close to the predetermined interval to such a task by setting an objective function for which a high score is desirable and solving an optimization problem. A method for calculating such a score will be described below. First, the variables used in the calculation will be described.

[0036] D LAST is the last (previous) execution date and time of the task. f is the desired execution interval (period) of the task. D I is the desired execution date and time of the task, and D I =D LAST +f.

[0037] t is the execution date and time of the task. S is the score. The score is a function S(t) that takes the execution date and time t as a parameter. r is the time since the last execution date and time, and t=D LAST +r.

[0038] u is a parameter that indicates the degree of urgency. The score S(t) decreases rapidly as u increases. I The value of u corresponds to the shape parameter of the Weibull distribution. λ is the maximum value of r and the time since the last acceptable execution date. λ corresponds to the scale parameter of the Weibull distribution.

[0039] w(r) is the time elapsed since the last execution date and time r(=tDLAST ) is the cumulative distribution function of the Weibull distribution, as shown in equation (1). W(t) is the execution date and time t (=D LAST + r) is the cumulative distribution function of the Weibull distribution, as shown in equation (2).

[0040]

number

[0041] Next, the constants used in calculating the score S(t) will be explained. The constants a1, b1, a2, b2, and c2, which are independent of the execution date and time t, are calculated from the following formulas (3) to (7).

[0042]

number

[0043] Equations (3) to (6) express the desired execution date and time of the task, D I (=D LAST +f) is 1 (S(D I )=1) (see equations (8) and (9) below). Equation (7) defines the allowable final execution date and time of the task, D LAST + λ score S(t) is 0 (S(D LAST +λ)=1) (see equations (9) and (10) below). a1 and b1 can be found by solving equations (3) to (4). a2, b2, and c2 can be found by solving equations (5) to (7).

[0044] Using the obtained a1, b1, a2, b2, and c2, the score S(t) is defined as shown in the following formulas (8) to (11). LAST When S(t) is 0, D LAST ≦t≦D I In this case, S(t) is calculated using the following equation (9). I ≦t≦D LAST When +λ, S(t) is calculated using the following formula (10).LAST When +λ, S(t) is set to 0.

[0045]

number

[0046] 3 is a graph of the score S(t) according to a modified example of this embodiment. The dotted line graph has a larger shape parameter u than the solid line graph, and the score S(t) is smaller than the solid line graph. I The graph slope is steep around t=D. I =D LAST +f is the best value 1, and t=D LAST +λ is 0. For such tasks that are executed periodically, the schedule generation unit 113 calculates the score and sets the objective function so that the larger the sum of the scores, the better the value. I A schedule will be generated that is assigned an execution date and time close to the current time.

[0047] Such a score S(t) may be used as a metric for tasks that have a desired execution date and time, as well as for tasks that are executed periodically. The score S(t) described above is a score based on the cumulative distribution function w(r) of the Weibull distribution, but it may also be a score that is maximum at the desired execution date and time and decreases as the execution date and time moves away from the desired execution date and time.

[0048] As described above, the tasks include periodic tasks that are executed at predetermined intervals. The schedule generation unit 113 calculates a score for the execution date and time of a periodic task, setting the highest value to a date and time that is a specified period after the previous execution date and time, and the further away the date and time is from the specified period after the date and time, the smaller the score becomes, and generates a schedule so that the sum of the scores of the periodic tasks is high. The score is calculated based on the value of the Weibull cumulative distribution function for the length of time since the previous run.

[0049] <<Modification: Constraint Evaluation Screen>> In the above embodiment, the constraint evaluation unit 114 outputs the generated constraint conditions that result in the best (highest) KPI. The constraint evaluation unit 114 may be configured to visualize a desirable schedule with a high KPI.

[0050] 4 is a screen configuration diagram of a constraint evaluation screen 610 according to a modified example of this embodiment. The horizontal axis is the first principal component, and the vertical axis is the second principal component. The circles are schedules, and the darker the hatching, the higher the KPI. The solid and dotted arrows indicate metrics. When a user places a pointer 611 on a schedule, the constraint evaluation unit 114 displays the generated constraint conditions corresponding to the schedule in a pop-up window 612. The constraint evaluation unit 114 may display metrics and KPIs in the pop-up window 612. By operating such a constraint evaluation screen 610, the user can understand the influence of the constraint conditions on the KPIs and metrics.

[0051] As described above, the constraint evaluation unit 114 displays the results of the clustering analysis, and displays the generated constraint conditions corresponding to a schedule specified among the displayed schedules.

[0052] <<Modification: Constraint Evaluation Unit>> In the above embodiment, the constraint evaluation unit 114 performs clustering analysis and outputs a generation constraint condition for a schedule that will have the best KPI among the schedules close to the cluster center. The constraint evaluation unit 114 may be configured to output a generation constraint condition for a schedule that will have the highest KPI among the generated schedules.

[0053] As described above, the schedule creation support device 100 includes the constraint evaluation unit 114 that outputs the creation constraint conditions corresponding to the schedule that provides the best KPI among the created schedules.

[0054] Other variations Although several embodiments of the present invention have been described above, these embodiments are merely illustrative and do not limit the technical scope of the present invention. For example, the constraint evaluation process (see FIG. 2) has been described on the assumption that there is one KPI, but there may be multiple KPIs. The schedule creation support device 100 may output a schedule creation constraint condition that provides the best (best range) for each KPI.

[0055] The present invention can take various other embodiments, and furthermore, various modifications such as omissions and substitutions can be made without departing from the gist of the present invention. These embodiments and modifications are included in the scope and gist of the invention described in this specification, etc., and are included in the scope of the invention described in the claims and their equivalents. [Explanation of symbols]

[0056] 100 Schedule creation support device 111 Reception 112 Constraint condition generator 113 Schedule Generation Unit 114 Constraint Evaluator 130 Task Database 140 Condition Database 150 Evaluation Value Database 160 Schedule Database 610 Constraint Evaluation Screen

Claims

1. a reception unit that receives a plurality of tasks, constraint conditions for schedule generation that assigns execution dates and times to the plurality of tasks, metrics for evaluating the generated schedule, and KPIs; a constraint condition generating unit that generates a plurality of generated constraint conditions by adding deviations to the constraint conditions; a schedule generation unit for generating an optimal schedule that satisfies the generation constraint conditions; and a constraint evaluation unit that outputs a generation constraint condition corresponding to a schedule that optimizes the KPI among the generated schedules. Schedule creation support device.

2. The constraint evaluation unit: performing a principal component analysis of the metrics of the generated schedule; A clustering analysis of the schedule is performed based on the obtained principal components. Among the schedules closest to the center of the cluster, output the generation constraints corresponding to the schedule that provides the best KPI.

2. The schedule creation support device according to claim 1.

3. The constraint evaluation unit: Displaying the results of the clustering analysis; Display the generation constraints corresponding to the specified schedule in the displayed schedule 3. The schedule creation support device according to claim 2.

4. The tasks include periodic tasks that are executed at predetermined intervals; The schedule generation unit a score is calculated for the execution date and time of the periodic task, the maximum value being the date and time that is the predetermined period after the previous execution date and time, and the score decreases as the date and time are farther away from the date and time that is the predetermined period after the previous execution date and time; The schedule is generated so that the sum of the scores of the periodic tasks is high.

2. The schedule creation support device according to claim 1.

5. The score is calculated based on the value of the Weibull cumulative distribution function for the period from the previous execution.

5. The schedule creation support device according to claim 4.

6. A schedule creation support device, receiving a plurality of tasks, constraint conditions for schedule generation for allocating execution dates and times to the plurality of tasks, metrics for evaluating the generated schedule, and KPIs; generating a plurality of generated constraint conditions by adding deviations to the constraint conditions; generating an optimal schedule that satisfies the generation constraints; outputting a generation constraint condition corresponding to the schedule that optimizes the KPI from among the generated schedules; How to help with schedule creation.

Images