JP2709010B2 - Reasoning method for scheduling - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to an inference method for scheduling by an expert system.

[0002] There is an increasing momentum to solve various scheduling problems such as production planning, personnel planning, and sales planning by an expert system. Accordingly, an inference method capable of efficiently solving a complicated schedule problem has been desired.

[0003]

2. Description of the Related Art Generally, a so-called vertical inference method is used in an expert system for solving a schedule problem. That is, each time a hypothesis is completed for one scheduling target, the hypothesis is evaluated. Based on the optimal hypothesis selected based on the evaluation result, a hypothesis is formed for the next target scheduling object. Is being done.

[0004] At that time, conventionally, the form of inference (generation of a hypothesis, evaluation of a hypothesis, and selection of a hypothesis) has been standardized for each scheduling object. In other words, for example, with regard to how to form a hypothesis, an agreement to establish all hypotheses that satisfy the constraints, an agreement to set a limit on the number of hypotheses (an agreement to terminate the generation of hypotheses when a certain number of hypotheses are obtained), If a hypothesis that satisfies is not obtained, one of a variety of agreements is selected at the system construction stage, such as an agreement to relax the constraints and form a hypothesis, and all scheduling targets are selected according to the selected agreement. Was hypothesized.

[0005]

However, for example, in a production plan which is a scheduling in the manufacturing industry,
When the work of a plurality of processes is to be scheduled, the first process has a plurality of available facilities (machines, production lines, etc.) and requires the selection of the place where the work is to be performed. In the third step, only relatively strict restrictions are imposed, such as that the equipment is singular and the place of implementation is uniquely determined, or the third step is only required to meet the delivery date. In many cases, such as being limited to daytime, there are a wide variety of constraints on hypothesis generation and evaluation related to scheduling.

In such a case, conventionally, an inference form capable of coping with the most complicated constraint is selected, and for this reason, a complicated inference process is performed even for a scheduling target whose constraint is relatively simple. However, there has been a problem that the efficiency of inference is impaired as a whole.

The present invention has been made in view of the above problems.
Therefore, schedule for a series of operations including multiple processes
When performing scheduling, it is possible to generate a schedule proposal by inferring the optimal form according to the attribute of each process.
Useless constraints and useless hypotheses.
An object of the present invention is to improve the efficiency of solving a complicated schedule problem without generating the schedule.

[0008]

According to a first aspect of the present invention, there is provided an inference method including a plurality of steps using an expert system as shown in FIG.
A reasoning method for scheduling a communication of the work, its by applying the constraint information according to the attribute of each of the steps
A plurality of hypothesis generation management units 13 that generate respective hypotheses
1, 132, and 133, corresponding to the attributes of each of the processes.
Inference Management Units 12A, 12B, 12C Used
And the plurality of inference management units 12A, 12B, and 12C.
A series of tasks to be scheduled and scheduled.
Multiple processes with individual attributes as module units
And the plurality of inference management units 12A, 12B, 12C for each of the divided processes in accordance with the attribute.
An inference engine for selecting each as inference realizing means one of the plurality of inference control section 12A, 12B, 1
2C, each inference management unit 12A, 12B, 1
A memory for storing the hypothesis of each step generated by 2C.
And a working memory 21 as a schedule unit.
The plurality of inference management units according to the attributes of the divided processes.
Select one of 12A, 12B, 12C and select
The inference form defined in the inference administration
By generating a hypothesis,
The process of generating a schedule plan is repeated for each of the above steps.
By doing so, a series of
Generate and store a proposed schedule.

[0009]

The inference method according to the second aspect of the present invention is directed to the inference method according to the third aspect of the present invention, wherein a series of operations related to the production of a product is the object of the scheduling. Correspond to different processes related to the production of one product.

[0011]

For example, in a production plan, various operations for producing a desired number of products (work in process) using predetermined equipment are to be scheduled.

The inference engine 10 sequentially pays attention to the individual tasks to be scheduled, and for the noted task, for example, “which step is the task”, “requires high precision”, “urgent task”, etc. The attribute (characteristic) of the work is determined, and the inference management unit 12 according to the determination result.
One of A, 12B and 12C is selected as the inference realizing means. As a result, different forms of inference are performed depending on the scheduling target.

When inferring each task, first, a hypothesis is generated by the hypothesis generation management unit 131 having the highest application order based on the constraint information associated therewith. And it is not possible to generate a hypothesis that satisfies the constraints,
If a hypothesis to be adopted as a schedule proposal cannot be obtained because the hypotheses generated are all rejected due to a low evaluation score, the other hypothesis generation management units 132 and 133 generate hypotheses.

The hypothesis is an example of setting the execution time of when to perform the work and the place of execution where to perform the work (which equipment to use).

[0015]

FIG. 1 is a block diagram of a main part of an expert system 1 to which an inference method according to the present invention is applied.

The expert system 1 includes a series of three processes (hereinafter referred to as process A,
This is an expert system for production planning that performs scheduling of processes B and C), and includes an inference engine 10, a working memory 11 for Gantt charts, and three inference management units 12A, 12B, and 12 provided for each process.
C, a hypothesis working memory 21, an in-process working memory 22, and the like. Such an expert system 1 is realized on a computer system such as a workstation, for example.

The inference engine 10 is a program that supervises scheduling, and includes inference managers 12A and 12A.
When the schedule plan is generated by B and 12C, predetermined display data is written to the Gantt chart working memory 11 in order to display the schedule plan on a display (not shown) in the form of a Gantt chart (schedule).

The inference management unit 12A is provided to realize an inference (processing or solving method) specific to the work of the process A, and is provided for each of the hypothesis generation ranks described later and the hypothesis in which the application order is determined. Generation management units 131, 132, 13
3, a judgment processing unit 17 for judging the suitability of the hypothesis, a scoring processing unit 18 for calculating the evaluation score of the hypothesis, and an inference management data table 20 storing information defining an inference mode.

Hypothesis generation management units 131, 132, 133
Is an execution place determination processing unit 14 for allocating equipment to work.
And an execution time search processing unit 15 for searching for an allocatable free time, and an execution time determination processing unit 16 for setting work start and end times.

Each processing unit 14 of the inference management unit 12A
Reference numerals 15, 16, 17, and 18 comprise programs corresponding to predetermined tasks, and refer to the inference management data table 20 when executing processing.

The inference management unit 12B and the inference management unit 1
2C is provided for realizing a unique inference regarding the work of the process B and the process C, respectively, and the basic configuration thereof is the same as that of the inference management unit 12A. However, in the inference management units 12A, 12B, and 12C, data values stored in the inference management data table 20 are different from each other, and the number of hypothesis generation management units is not necessarily the same.

The hypothesis working memory 21 is a hypothesis management space (search space) for storing hypotheses generated sequentially for each task in association with each other, and is shared by the inference management units 12A, 12B, and 12C.

The inter-process working memory 22 is a memory space for temporarily storing data generated in the course of various processes by the processing units 14 to 18 of the inference management units 12A, 12B, and 12C.

FIG. 2 is a diagram showing an example of the data configuration of the inference management data table 20. The inference management data table 20 includes an attribute value D0, a hypothesis generation rank number D1, an implementation location determination function D2, an implementation time determination function D3, a determination constraint D4, a scoring constraint D5, a score evaluation function and a coefficient D.
6, a score calculation function D7, and a limit value D8 of the number of hypotheses.

The attribute value D0 is stored in the inference management data table 2
This is data indicating the contents of the attribute of the work to which 0 is applied (corresponding steps in this embodiment). The number of hypothesis generation ranks D1 is
This is the number of ranks in the stepwise generation of hypotheses by ranking the constraint information, and is data indicating the number of hypothesis generation management units 131 to 133.

The execution place decision function D2 is data indicating the restriction on the setting of the execution place, and is registered for each rank of hypothesis generation. For example, "If possible, implement at the highest priority facility; if that is not possible, implement at the second priority facility; otherwise, implement at another facility."
In such a case, data indicating the constraint that the priority is to be implemented in the highest-ranking facility is registered as the decision function D2 of rank 1, and data indicating the constraint that the priority is to be implemented in the facility of the second rank is rank 2 Is registered as a decision function D2 of rank 3, and data indicating a restriction to be implemented in an arbitrary facility is registered as a decision function D2 of rank 3.

The execution time decision function D3 is data indicating restrictions on setting the execution time , and is also registered for each rank of hypothesis generation. For example, in the case of "make sure to meet the delivery date as much as possible, and if it is not possible, finish the process as soon as possible", data indicating a constraint to meet the delivery date is registered as the rank 1 decision function D3. , Data indicating a constraint to finish as soon as possible is registered as the rank 2 decision function D3.

The determination constraint D4 is data that defines the processing of the determination processing unit 17, and the scoring restriction D5 is data that defines the processing of the scoring processing unit 18. The score evaluation function and the coefficient D6 are data that define the hypothesis evaluation and the weighting of the result.

The score calculation function D7 is data indicating a formula for calculating an evaluation score, and the limit value D8 of the number of hypotheses is data indicating an upper limit of hypotheses generated for one work. Note that these data D0 to D8 are based on the condition of the equipment,
In order to obtain a more practical schedule plan in accordance with the actual situation such as the preparation status of materials and personnel, the inference management unit 12
A, 12B, and 12C (that is, processes A, B, and C) are appropriately selected and registered.

In the expert system 1 having the above configuration, scheduling is executed for each order (production order) issued with information on work requirements such as the name of the product to be manufactured, its quantity, and delivery date. , 3
It is executed for the order of several hundred in an appropriate period such as months or six months.

The inference engine 10 has three inference managers 12 for each operation to be scheduled, according to the attribute of which operation the operation is.
One of A, 12B, and 12C is selected to execute the inference.

For example, if the work of interest is the work of step A, the inference management unit 12A is selected as the inference executing means. The selected inference management unit 12A firstly performs the hypothesis generation management unit with the highest application order, that is, the strictest rank 1
A hypothesis is generated by applying the hypothesis generation management unit 131 corresponding to the constraint.

When generating a hypothesis, first, the processing unit 14 for deciding the place of execution is started, and the inference management data table 2 is set.
The execution place is set in accordance with the constraint indicated by the decision function D2 of rank 1 in 0. Subsequently, the feasible time search processing unit 15 determines the feasible time in consideration of, for example, the context of other work, and the like. The implementation time is set so as to satisfy the constraint indicated by. With the setting of the execution place and the execution time, one hypothesis for the work has been generated. The generated hypothesis is stored in the hypothesis working memory 21.

The execution time determination processing unit 16
The other hypotheses for one operation are generated in order, and the process is terminated when the constraint cannot be satisfied or when the hypothesis with the limit value D8 of the number of hypotheses is generated.

When the hypothesis is generated, the inference management unit 12A
Activates the determination processing unit 17 and the scoring processing unit 18 in order, determines whether the hypothesis is appropriate, and scores and evaluates the conforming hypothesis.

When a hypothesis satisfying the rank 1 constraint is not satisfied in the hypothesis generation process by the hypothesis generation management unit 131 or all hypotheses are rejected at the evaluation stage, that is, a hypothesis to be adopted as a schedule plan is obtained. If not, the inference management unit 12A sends the hypothesis generation management unit 1
Instead of 31, a hypothesis is generated again by applying the hypothesis generation management unit 132 corresponding to the second rank and the restriction of rank 2. Then, the rank 2 hypothesis generation management unit 13
If the hypothesis to be adopted is not obtained even after applying 2, the inference management unit 12A further sets the modest rank 3
Hypothesis generation management unit 13 corresponding to the constraint and having the third highest ranking
3 is applied to generate a hypothesis.

That is, the expert system 1 does not generate all the hypotheses that are permissible in view of the entire constraint at once, but first generates a hypothesis that satisfies the strictest constraint, and fails to obtain a hypothesis to be adopted. The hypothesis is generated in stages by the procedure of generating the hypothesis by only relaxing the constraint.

On the other hand, if the work of interest is the work of the process B, the inference management unit 12B is selected as the inference executing means. Then, the constraint information unique to the process B (the above-described data D
Based on the inference management data table 20 in the inference management unit 12B in which 0 to D8) are registered, inference is performed on the work of interest. If the work of interest is the work of the process C, the inference management unit 12C is selected, and the inference is performed based on the inference management data table 20 in which the constraint information unique to the process C is registered.

The inference management unit 12B and the inference management unit 1
In 2C, the hypothesis is generated in a stepwise manner, similarly to the inference management unit 12A. According to the above-described embodiment, only when the hypothesis to be adopted is not obtained, the constraint is gradually reduced to generate the hypothesis, so that the generation of the useless hypothesis is suppressed. The capacity of the working memory 21 can be reduced, and the speed of inference can be increased.

In the above-described embodiment, the number of ranks of the hypothesis generation constraint is arbitrary. That is, each inference management unit 12A,
Hypothesis generation management units 131 to 133 provided in 12B and 12C
Can be selected as appropriate according to the ranking of the constraints.

In the above embodiment, steps A, B, C
Although the inference management units 12A, 12B, and 12C are provided for each, that is, the form of the inference is changed according to the attribute of the process to which the work to be scheduled belongs. And an operation corresponding to an order having a relatively long delivery time, and an inference management unit may be provided for each of these groups to perform group-specific inference. In this specification, an expert system includes a system construction tool.

[0042]

According to the first to third aspects of the present invention,
Schedule for a series of operations involving multiple steps
Optimum form according to the attribute of each process when performing
Went the reasoning can be used to generate a proposed schedule,
Useless constraints are applied or useless hypotheses are generated.
It is possible to improve the efficiency of solving complicated schedule problems. Moreover, the attribute of each process
In response to sex, to mitigate the stepwise restriction can produce proposed schedule <br/>, that suppresses the generation of useless hypothesis
And the schedule generated by each inference manager
It is possible to manage plans in a unified way with a common working memory
Because it can increase the speed of and reasoning can in <br/> be made small capacity of the working memory.

[0043]

[Brief description of the drawings]

FIG. 1 is a block diagram of a main part of an expert system to which an inference method according to the present invention is applied.

FIG. 2 is a diagram illustrating an example of a data configuration of an inference management data table.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Expert system 12A, 12B, 12C Inference management part 10 Inference engine 131, 132, 133 Hypothesis generation management part

 ──────────────────────────────────────────────────続 き Continuation of the front page (72) Inventor Kazunori Fukuyama 2-15-22 Nishiki, Naka-ku, Nagoya City, Aichi Prefecture Fujitsu Tokai System Engineering Co., Ltd. (56) References JP-A-4-271467 (JP, A JP-A-1-224845 (JP, A) JP-A-1-257300 (JP, A) JP-A-3-188562 (JP, A) JP-A-64-76360 (JP, A) JP-A-62-1987 6857 (JP, A) JP-A-62-285172 (JP, A) JP-A-4-239975 (JP, A) JP-A-2-249032 (JP, A) JP-A-1-140337 (JP, A) JP-A-63-304328 (JP, A)

Claims (3)

(57) [Claims] 1. A plurality of steps using an expert system
An inference method for performing scheduling of a series of operations including: applying constraint information corresponding to the attribute of each process to each of the processes.
Comprising a plurality of hypothesis generation management unit for generating a hypothesis,
A plurality of inference managers used in accordance with the attributes of each process, and the plurality of inference managers are managed and scheduled.
A series of tasks to be performed
Divided into multiple processes with
Management of the plurality of inferences according to their attributes for each of the selected processes
An inference engine for selecting each as inference realizing means one of the parts is shared by the plurality of inference control section, each inference manager
Workin for memorizing the hypothesis of each process generated from
According to the attribute of the process divided as a schedule unit.
Selecting one of the plurality of inference managers and selecting the selected one;
According to the form of inference defined in the Inference Management Department,
By generating a hypothesis,
The process of generating a joule plan is repeated for each of the above steps
A series of schedules on the working memory.
And generating and storing a rule plan . 2. The inference method for scheduling according to claim 1, wherein a series of operations related to the production of a product is set as an object of the scheduling. 3. The inference method for scheduling according to claim 2, wherein each of said inference management sections corresponds to different processes related to production of one product.