JPH03257632A - Schedule preparing system - Google Patents

Abstract

PURPOSE:To improve the preparing efficiency of a system as a whole by providing an inference processing part and a schedule knowledge control part. CONSTITUTION:According to a command from an inference control part 11, a resource deciding means 13 activates a limit control means 24 by using the information of resources obtained by activating a resource control means 21, the information of rules for selecting the resources obtained by activating a rule control means 23 and the information of limit obtained by activating a limit control means 24. The resources are decided so as to be allocated to a task selected by a task deciding means 12, and outputted through a schedule result output means 32 to an output device 5. On the other hand, information concerning the satisfactory degree of the limit is controlled as schedule knowledge similarly to information such as the task, resources and rules. Thus, the preparing efficiency of the system can be improved as a whole.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a schedule creation method, and particularly to a schedule creation method that employs a candidate evaluation method.

[Conventional technology]

Traditionally, when creating a schedule, when determining resources for a selected task, a method has been used to select resources that satisfy the constraints that are the conditions for allocation. It was also managed together with constraint information, and when refining the scheduling system to improve the quality of the schedule, the satisfaction value was changed by modifying the constraint program.

[Problem to be solved by the invention]

In the conventional method for determining candidate evaluation values described above, when making modifications such as refining the scheduling system, the satisfaction values for evaluating candidates are written in the program, so it is necessary to check the constraints. The parts that provide satisfaction with the constraints and the parts that provide satisfaction with those constraints are managed in the same place, making the program itself difficult to read, reducing the efficiency of creating the entire system, and making it difficult to improve the system.

[Means to solve the problem]

The schedule creation method of the present invention, when allocating resources to tasks that are elements of individual combinations, selects resources based on the evaluation of satisfaction of constraints that are conditions for allocation, and combines them with tasks. , an information system that creates an entire schedule by repeating such operations, comprising: task determining means for selecting a task to be allocated; and resource determining means for selecting a resource to be allocated to the task selected by the task determining means. , a constraint applying means for applying constraints to determine a resource, an evaluation means for evaluating candidates when selecting a resource, and controlling activation and repetition of the task determining means and the resource determining means. an inference processing unit including an inference control means; a task management means for managing information about tasks; a resource management means for managing information about resources; and a rule for managing information about rules, which is information about application of constraints. A schedule knowledge management section including a management means, a constraint management means for managing information regarding constraints, and an evaluation function management means for managing an evaluation function for evaluating resource candidates.

〔Example〕

Next, the present invention will be explained with reference to the drawings.

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention.

In FIG. 1, 1 is an inference processing unit, 2 is a schedule knowledge management unit, 3 is an input/output control unit, 4 is an input device consisting of a keyboard device, a mouse (position specifying device), etc., and 5 is a CRT.
Represents an output device such as a display device.

The inference processing unit 1 includes an inference control unit 11, a task determination unit 12, a resource determination unit 13, a constraint application unit 14, and an evaluation unit 15, and the schedule knowledge management unit 2
includes a task management means 21, a resource management means 22, a rule management means 23, a constraint management means 24, and an evaluation function management means 25.

The input/output control unit 3 includes a schedule knowledge input unit 31 that inputs information necessary for scheduling information regarding tasks, information regarding resources, information regarding rules, information regarding constraints, and information regarding evaluation functions, and a task determining unit 1.
2 and resource determining means 13, schedule result output means 32 outputs information on selected tasks and resources by obtaining task and resource information from task management means 21 and resource management means 22. I'm here.

The task management means 21 includes schedule knowledge input means 31
Information about tasks, which are the refinement elements of the schedule input by
(earliest start time, latest start time, constraints, priorities, etc.).

The resource management means 22 inputs information about the resource to be allocated to the task inputted by the schedule knowledge input means 31 (schedule target start time, schedule target end time, free time, constraints, priority, assignment to the task, etc.). > Management.

The rule management means 23 includes the schedule knowledge input means 31
The system manages information on rules and know-how regarding the application of constraints entered by (constraint relaxation limits, procedures when rules are successful, etc.). Furthermore, the constraint management unit 24 manages information regarding conditions for assignment of tasks and resources inputted by the schedule knowledge input unit 31 (constraint importance, calculation cost, search efficiency, constraint function, etc.).

The evaluation function management means 25 manages evaluation functions for each constraint in order to convert the values obtained as a result of applying the constraint functions into satisfaction levels for evaluating resource candidates.

The inference control unit 11 activates and repeats the task determination means 12 and the resource determination means 13 based on commands input and accumulated in advance.

The task determining means 12 uses the task information obtained by activating the task management means 21 and the rule for selecting a task obtained by activating the rule management means 23 in response to a command from the inference control section 11. Using the information and constraint information obtained by activating the constraint management means 24, the constraint application means 24 is activated, a task to be scheduled is determined, and the output device 5 is outputted via the schedule result output means 32. Output to.

The resource determining means 13 uses the resource information obtained by activating the resource management means 21 and the rule for selecting resources obtained by activating the rule management means 23 in response to a command from the inference control unit 11. Information and constraint management means 24
Using the constraint information obtained by activating , the constraint management means 24 is activated, which determines the resources that can be allocated to the task selected by the task determination means 12 , and the resource is sent via the schedule result output means 32 . Output to output bag W5.

The constraint application means 14 uses task information obtained by starting the task management means 21, resource information obtained by starting the resource management means 22, and rule management means 23.
The constraint conditions are applied using the rule information obtained by activating the constraint management means 24 and the constraint information obtained by activating the constraint management means 24.

The evaluation means 15 calculates the evaluation value of the candidate using the evaluation function information obtained by activating the evaluation function management means 25 for the resources whose constraints are satisfied by the constraint application means 14.

Next, the processing by the constraint applying means 14 will be explained with reference to FIG. 2. First, information on a constraint function corresponding to the specified constraint is obtained by referring to the constraint management means 24 (processing 71). Next, in process 72, information regarding the instructed task or resource is obtained by activating the task management means 21 or resource management means 22, and information regarding the rule obtained by activating the rule management means 23 is used. and execute the constraint function.

As a result of executing the constraint function, values such as the remaining amount of materials for making a product and the start time of a task are returned.

The constraint application means 14 further activates the evaluation means 15.

The evaluation means 15 obtains information on the evaluation function corresponding to the specified constraint by activating the evaluation function management means 25 (processing 73).

Finally, in process 74, the value obtained in process 72 is converted by executing the evaluation function obtained in process 73, and a value of the degree of satisfaction of the constraint is obtained. In the example of FIG. 3, the evaluation function is a function that converts the value of the start time of a task into an integer value from 0 to 100. When the satisfaction value is O, it means that the constraint is not satisfied, and when it is from 1 to 100, it means that the constraint is more satisfied as the value increases.

It should be noted that the present invention is not limited to this embodiment, and in particular, regarding the domain of the evaluation function, a set of real numbers and a set other than the set of numbers can also be considered. For example, if the constraint is "brighter colors are better," and the value returned from the constraint function (the range of the evaluation function) is yellow, blue, white, or black, the evaluation function is as shown in Figure 4. It is managed in the form of a table like this.

〔Effect of the invention〕

As explained above, the present invention adjusts the satisfaction level by managing the information on the satisfaction level of constraints, which was conventionally written in the program, as schedule knowledge in the same way as information such as tasks, resources, constraints, and rules. This makes it easier to improve the system, and by separating the constraint function that checks the constraints from the evaluation function that evaluates them, the readability of the constraint functions increases, improving the efficiency of creating the entire system. I can do it.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing the configuration of a schedule creation method in which the evaluation method of an embodiment of the present invention is adopted, FIG. 2 is a flow chart showing an overview of the process of applying constraints and obtaining satisfaction, and FIG.
4 and 4 are explanatory diagrams showing examples of evaluation functions. DESCRIPTION OF SYMBOLS 1... Inference processing unit, 2... Schedule knowledge management unit, 3... Input/output control unit, 4... Input device, 5...
Output device, 11... Reasoning control means, 12... Task determining means, 13... Resource determining means, 14... Constraint application means, 15... Evaluation means, 21... Task management means, 22... Resource management means, 23... Rule management means, 24... Constraint management means, 25... Evaluation function management means, 31... Schedule knowledge input means, 32...
- Schedule knowledge output means.

Claims (1)

[Claims] When allocating resources to tasks that are elements of individual combinations, resources are selected and combined with tasks based on evaluation of satisfaction of constraints that are conditions for allocation, and such operations are performed. In an information system that creates an entire schedule by repeating the following steps, a task determining means selects a task to be allocated, a resource determining means selects a resource to be allocated to the task selected by the task determining means, and a resource determining means selects a task to be allocated. a constraint applying means for applying constraints to select a resource; an evaluation means for evaluating candidates when selecting a resource; and an inference control means for controlling activation and repetition of the task determining means and the resource determining means. an inference processing unit including; a task management means for managing information regarding tasks; a resource management means for managing information regarding resources; a rule management means for managing information about rules, which is information regarding application of constraints; A schedule creation method comprising: a schedule knowledge management section including a constraint management means for managing information regarding resource candidates; and an evaluation function management means for managing an evaluation function for evaluating resource candidates.