JPH04182773A - Mission scheduling device - Google Patents

Abstract

PURPOSE:To discover a quasi-optimum resolution within a practical range of time and memory by executing mission scheduling while calculating the time range in which scheduling is possible and deciding the scheduling time in a distribution or integration method. CONSTITUTION:A possible scheduling time range calculation means 4 calculates the time range in which scheduling is possible. When it cannot calculate it, a scheduling fine adjustment means 5 moves the mission scheduling to calculate the possible scheduling time range. A scheduling time deciding means 6 calculates the optimum scheduling time corresponding to distribution or integration from the time range in which scheduling is possible and inquires the resource supply quantity when executed according the scheduling time. These operations are executed in terms of the next mission and the mission scheduling time is decided. Thus, the quasi-optimum resolution is discovered within a practical range of time and memory.

Description

[Detailed Description of the Invention] [Summary] This invention relates to a mission scheduling device that determines a mission schedule under constraint conditions including resource supply capacity and resource consumption by a mission. Instead of adopting the hack trunk mechanism of selecting missions to be scheduled by trial and error, we have developed original methods for selecting trial missions for scheduling and setting the scheduling time, and have overcome combinatorial explosion while achieving practical implementation. The purpose of the present invention is to find a sub-optimal solution for mission scheduling within a reasonable time and memory range, and to determine a scheduling trial order for a plurality of missions to be scheduled. A schedulable time slot calculation means calculates a schedulable time slot based on resource consumption and resource supply for the missions in the order determined by the schedulable time slot calculation means, and a schedulable time slot calculated by the schedulable time slot calculation means. and a scheduling time determining means for calculating the best scheduling time by decentralization or centralization based on the time zone, and for the mission of trial 1115 determined by the scheduling trial mission determining means for a plurality of missions to be scheduled. , the schedulable time slot calculating means calculates a schedulable time slot, the scheduling time determining means calculates the best scheduling time corresponding to decentralization or centralization from among these schedulable time slots, and calculates the best scheduling time corresponding to decentralization or centralization, A new resource supply amount is obtained by subtracting the resource consumption amount when executed in , and the mission is repeatedly performed for the next trial order to determine the mission scheduling time.

[Industrial application field]

The present invention relates to a mission scheduling device that determines a mission schedule under constraint conditions including resource supply capacity and resource consumption by the mission. For example, various experiments such as materials experiments, life sciences, and scientific observations will be carried out on the Japan Experimental Module built on the U.S. space station (hereinafter, the same experiment may be repeated several times, and these will be collectively referred to as one mission). ) The present invention relates to a mission scheduling device for scheduling missions related to resource allocation.

[Conventional technology]

In the mid-1990s, the Japan Experiment Module (hereinafter referred to as JEM) will be built on the U.S. space station, and it is planned to carry out various missions by effectively utilizing the zero gravity, high vacuum, and high-energy particle environment of outer space. It is.

The operational plan for the implementation of Minnyong is to create a long-term plan over a five-year time span through international coordination, and then to develop a medium-term plan (two years) to a quarterly plan (three years).
The structure is such that the plans are created step by step in the following order: ``Rikitsuki'' - Weekly Plan (1 week) - Daily Plan (1 day).

As shown in Figure 15, the operation plan configured in this way is created by the ground operation control system, and the weekly plan (base schedule), which corresponds to 7 days of the daily plan, is uploaded to the onboard space station JEM system. I will do it.

The onboard crew will then carry out the mission based on this uploaded base scheduling. Additionally, if the base schedule needs to be reviewed based on the progress of space station and JEM operations, the daily plan will be reconstructed using the ground operation control system on the ground, and the base schedule or quarterly plan will also be changed. It happens.

In this way, you will be tasked with the daily task of creating daily plans and quarterly plans for mission operations as necessary. Resource consumption such as power consumption and fluid consumption required for JE
The start times of the multiple missions must be determined while keeping them within the resource supply amount of M. FIG. 16 shows an example of constraint conditions for implementing the mission. As shown in this Figure 16,
There are a considerable number of resource constraints related to the implementation of one mission, and the number of missions scheduled to be implemented in a day is quite large, in the dozens, so it is important to create a schedule for the implementation of the mission. It is expected that the load will be quite large. For example, FM
In the case of PT (primary materials experiment), it takes about a week to manually create a similar experiment schedule for one day.

If this also involves creating a quarterly plan, it is clear that it cannot be done manually.

For these reasons, a need has arisen for the development of new mission scheduling devices that can automatically create mission schedules through interaction with planners. However, participation in an experiment conducted on a space station such as JEM has never been seen before in Japan, and the current situation is that there is no prior art that is compatible with the mission scheduling device of the present invention.

[Problem to be solved by the invention]

Conventionally, the problem of calculating the combinations and scheduling times of multiple missions under predetermined constraint conditions has been solved as follows:
If you try to solve the problem blindly or by trial and error using a computer, you will search for "the number of combinations of missions" and find that "there are countless solutions within the schedulable time period" regarding the scheduling of one mission. There is a problem that searching for these results in what is called a ``combinatorial explosion,'' making it extremely difficult to find an appropriate scheduling solution within a practical amount of time. In order to solve this problem, the present invention utilizes a heuristic method, that is, human knowledge, and does not employ a hack track mechanism of trial and error in selecting missions to schedule.
The purpose is to overcome combinatorial explosion and discover semi-optimal solutions for mission scheduling within practical time and memory by using unique methods such as selecting scheduling trial missions and setting scheduling times. It is said that

[Means to solve the problem]

The means for solving the problem will be explained with reference to the first leap.

In FIG. 1, a scheduling trial mission determining means 3 determines the order in which scheduling is attempted for a plurality of missions to be scheduled.

The schedulable time period calculation means 4 calculates the schedulable time period for the missions in the order determined by the scheduling trial mission determination means 3 based on the resource consumption amount and the resource supply amount.

When a possible time slot cannot be calculated by the schedulable time slot calculation means 4, the scheduling m adjustment means 5 calculates a tentative schedulable time slot by ignoring a certain resource, and calculates a tentative schedulable time slot that already exists in this tentative schedulable time slot. When it is determined that the mission can be moved, the provisional schedulable time slot is determined as the schedulable time slot.

The scheduling time determining means 6 calculates the best scheduling time by decentralization or centralization based on the schedulable time periods calculated by the schedulable time period calculation means 4 and the scheduling fine adjustment means 5.

[Effect]

As shown in FIG. 1, the present invention includes a schedulable time slot calculation means 4 that calculates a schedulable time slot for a plurality of missions to be scheduled in a trial order determined by a scheduling trial mission determination means 3; If the calculation is not possible, the scheduling m adjustment means 5 moves the scheduling of other missions to calculate a possible scheduling time period, and the scheduling time determining means 6 selects the best scheduling time period corresponding to decentralization or centralization from among these possible scheduling time periods. The scheduling time of the mission is set in the summer, the amount of resource supply is determined by calculating the amount of resource consumed when the mission is executed at this scheduling time, and the resource supply amount is determined by repeating the process for the next mission to determine the scheduling time of the mission.

Therefore, by decentralizing or centralizing the selection of trial missions for scheduling and the determination of scheduling time, instead of adopting a bank trunk mechanism in which the selection of missions to be scheduled is carried out by trial and error, practical time and memory ranges can be achieved. It becomes possible to decide on mission scheduling within

〔Example〕

Next, the configuration and operation of the embodiment of the present invention will be explained in detail using FIGS. 1 to 14.

In FIG. 1, initial condition setting 2 includes securing a work area,
This includes setting various initial conditions.

The scheduling trial mission determining means 3 schedules a mission from among a plurality of missions to be scheduled using evaluation formula 1 and evaluation formula 2 based on the priority set in advance for the mission, the current resource usage rate, and the shape of the mission. (This will be explained in detail using Phases 0 to 3 in Figure 3).

For the mission determined by the scheduling trial mission determination means 3, the schedulable time slot calculation means 4 calculates the amount of resources consumed and the amount of resources that can be supplied to the mission to be scheduled. This is to calculate the schedulable time period within the scheduling period (details will be explained using [phase] to 0 in FIG. 4).

When the possible time slot cannot be calculated by the schedulable time slot calculation means 4, the scheduling fine adjustment means 5 calculates a tentative schedulable time slot by ignoring a certain resource, and calculates a tentative schedulable time slot that already exists in this tentative schedulable time slot. When it is determined that a mission can be moved, the temporary schedulable time zone is determined as the schedulable time zone, and the scheduling time of the already assigned mission is finely adjusted (see Figure 5 [Phase] to [Phase]). phase]).

The scheduling time determining means 6 selects from among the schedulable time periods calculated by the schedulable time period calculating means 4 or the scheduling fine adjustment means 5, as shown in FIG.
81, (bj, +c+, (d+) times are determined to calculate the best scheduling time (details will be explained using phases 0 to [phase] in FIG. 6, and see FIGS. 11 to 13).

The new environment generation means 7 includes the scheduling time determination means 6
Based on each time determined by Detailed using
(See Figure 14). Then, steps 3 to 7 are repeated, and when the amount of resource supply becomes insufficient and scheduling is no longer possible, scheduling is terminated. A detailed explanation will be given below.

FIG. 2 shows a system configuration diagram of the present invention.

In Figure 2, the mission scheduling device (
MISES) 1 is a device related to the present invention and is a device that performs mission scheduling.

ESHELL/X is an expert system construction support tool that operates under the COMMON LISP environment and constructs an inference section.

XWindow operates in a C language environment and constructs the first part of MM.

The input files include a mission file that stores mission objects to be scheduled related to this embodiment, a resource file that stores resource objects, a scheduling condition file that stores scheduling condition objects, and the like.

The output file includes a scheduling result file that stores scheduling results.

The display is used to display various data.

Next, in accordance with the order shown in the flowcharts of FIGS. 3 to 6, the process shown in FIG. The operations of the time determining means 6 and the new environment generating means 7 will be explained in detail.

In FIG. 3, ■ loads and starts the MM1 section in FIG.

■ Displays the initial screen.

0, select ``Load condition file'' from the initial screen. As a result, the mission file, resource file, and scheduling condition file are read from the input files in Figure 2.

For [Phase], select "Execute scheduling" from the initial screen. As a result, execution of mission scheduling is started based on the file read in step 0.

■ Loads and starts the inference section. this is,[
In response to selecting ``Execute scheduling'' in ``Stock'', load and start the inference section in Figure 2, and select ``Execute scheduling''.
The mission scheduling of this embodiment is performed by the process of phase].

The above processing of 0 to [phase] is shown in Figure 1 Initial condition setting 2
Do it with

In FIG. 3, [phase] groups missions based on their priority. This can be seen, for example, in Figure 7 (
c) The priority of the mission preset by the planner, shown in ■ in object A, "3" in this case, is taken out, and the plurality of missions to be scheduled are grouped according to the priority.

(2) determines whether there are multiple missions within the group. If YES, the order of the missions to be scheduled is further determined using [phase] and [phase] for the missions within the group. Then, proceed to Figure 4 ■. If NO, proceed to Figure 4 (■).

[Phase] calculates the importance of all resources using evaluation formula 1, and selects the resource with the highest value. This calculates the importance of all resources using the evaluation formula 1 below from the resource priority and resource usage rate set in advance for the resource, and the initial resource supply pattern area of the resource with the highest value.Here, the resource priority is, for example, The priority is "2" shown in (■) of the power object in FIG. 8 (c).

The maximum value of resource priority is the maximum value of the resource priorities of resources to be scheduled. The resource evaluation coefficient E4 is, for example, the resource evaluation coefficient E4 "1.55" assigned to the entire resource in the scheduling condition object example in FIG. This is the area obtained by reducing the remaining pattern area.

Note that here, the resource evaluation coefficient E5 is multiplied by the resource priority when the target resource is other than the power type.

[Phase] calculates evaluation formula 2 from the resource with the highest degree of importance and the request pattern of each mission in the same group, and determines the mission with the highest value as the mission for the scaling trial. This is the resource selected in [Phase] (the resource with the highest priority, that is, the critical resource).
For all scheduled missions in the same group, the scheduling order is determined based on evaluation formula 2, that is, the pattern length is long, the pattern height is low, the pattern area is large, the number of patterns is large,
The scheduling order is determined based on the assumption that the greater the number of experiment repetitions, the higher the scheduling order.

Here, for example, pattern length/pattern length average is normalized by focusing on the most important resource selected in [phase] of each mission and dividing the pattern length of the resources of the mission by the pattern length average. It is. The same applies below.

The above-mentioned [phase] to [phase] processing is performed by the scheduling trial mission determining means 3 in FIG. , Furthermore, when there are multiple missions in the group, select the resource with the highest importance using evaluation formula 1, and then focus on this resource with the highest importance,
Evaluation formula 2 for the relevant resources required by each mission
By determining the scheduling order in descending order of the values, it becomes possible to determine the scheduling order of the mission by simple reasoning.

Next, the operation of the schedulable time slot calculation means 4 will be explained using FIG. 4.

In FIG. 4, [phase] determines whether the maximum value of the mission request pattern is the maximum value of the resource supply pattern. This is the maximum value of the resource consumption pattern of the mission to be scheduled, for example, the 7th
Figure (C) Maximum value of consumption pattern of "Power object A""60" (Maximum value of resource supply pattern, for example, Figure 8 (C) Maximum value of supply pattern of power object "
100''. If YES, the maximum value of the mission request pattern is smaller than the maximum value of the resource supply pattern, so resources can be allocated, so proceed to ■. If NO Since the maximum value of the mission request pattern is greater than the maximum value of the resource supply pattern, the resource cannot be allocated, so the scheduling is determined to have failed in [phase], and the next scheduling trial mission is determined in [phase]. ([Phase]), proceed to Figure 4 ■ ([Phase]).

(2) determines whether the length (period) of the mission is a scheduling period or not. This means that the length of the mission, for example, the length of "100 hours" in the scheduling condition object in FIG.
2 days". If YES, proceed to 0. If NO, select [phase] as a scheduling failure, and select ■ to determine the next scheduling trial mission. ]), proceed to Figure 4 ■ ([Phase]).

0 determines the time that can be included in the mission request pattern for each current resource supply pattern (remaining resource pattern).

In (2), it is determined whether or not there is a common part of the time slots found for each resource. , YES, the scheduling time slot has been calculated, so the process proceeds to FIG. On the other hand, N
In the case of O, since the schedulable time slot could not be calculated, the process proceeds to FIG.

By the above-mentioned processing of [phase] to 0, the schedulable time period for the mission is calculated.

With reference to FIG. 5, the operation of attempting to calculate a schedulable time slot by the scheduling fine adjustment means 5 for a mission for which a schedulable time slot could not be calculated in FIG. 4 will be described. This is because the request pattern of the mission cannot be scheduled with the current remaining resource pattern, but the current request pattern may not be able to be scheduled due to scheduling of another request pattern, so move the mission that has already been scheduled. Adjust the current mission to be able to be scheduled.

In FIG. 5, [phase] calculates a common portion (temporary schedulable time slot) while ignoring resources that do not have a schedulable time slot. This is, for example, the 10th
As shown in the figure, among resources A, B, and C, resource C is ignored and the common portion of the remaining resources A and B is determined as a tentative schedulable time period as shown in the figure.

0 searches for other missions that have already been scheduled during the tentative scheduling period. This search targets only missions belonging to the same level so as not to disrupt scheduling priorities.

[Phase] determines whether the exploration mission can move. For example, as shown in FIG. 10, these O and O are obtained by searching for other missions that have already been scheduled in the tentative scheduling period for the mission to be scheduled, and finding them as shown in the figure. It is determined whether or not another mission can be scheduled by moving it as indicated by the arrow, or whether it can be scheduled at another position. If YES,
The provisional scheduling possible time zone is determined as the scheduling possible time zone, and the process proceeds to Figure 6 ■''.On the other hand, in the case of No, the scheduling possible time zone could not be calculated, so the scheduling fine adjustment is determined to have failed in [phase]. , [phase] determines the next scheduling trial mission.
~[phase]), 4th rj! Proceed to J■.

With the above [phase] or [phase], when the schedulable time slot cannot be calculated, another mission that has already been scheduled is moved to try to calculate the schedulable time slot.

Using FIG. 6, the operation of the scheduling time determining means 5 to determine the scheduling time based on the schedulable time zone determined in FIGS. 4 and 5 will be explained.

In FIG. 6, [Stock] determines the critical resource with the highest consumption rate in the current resource situation.

Here, the critical resource is the resource with the highest consumption rate among the current resources.

(2) extracts the part with the smallest width from the schedulable time slot and sets both ends of it as the scheduling candidate time (Ml).This means, for example, as shown in FIG. (possible time zone), the times at both ends of the smallest one 1..1
．． seek.

@ determines the scheduling candidate time based on the usage status of critical resources determined in ■ ((bl)
. For example, as shown in Figure 11 (bl), from the critical resource remaining pattern obtained in step (■), the request pattern of the selected mission is inscribed as indicated by diagonal lines, and the start time is in descending order of the pattern. Illustrated t3,”
It is determined as a, tS, L6, t), Ll), t.

[Phase] calculates the scheduling candidate time based on the operation model ((c1), for example, as shown in FIG. 11(c
), if the operational model is set by the planner, the mission leader can use the operational model (
Here, four times can be taken) as shown in the figure, and the times at that time are selected as tlo, tl), t1□, t).

■ Finds the times at both ends of all operational hours ((d
)). For example, as shown in FIG. 11 (dl), the times at both ends of the schedulable time zone are
．． 115, tl6, tl-, tll), tl9...
Close and ask.

[Phase] determines the scheduling time from (a), (b), (c1, fdl) by decentralizing or centralizing.

+11 Determination of time by decentralization: For example, Fig. 13 (
The times shown in ta+, (b), (c1, fdl in b),
By the processing of FIG. 6 0, ■, [phase], and [phase] described above, the results are obtained as schematically shown in FIG. 11 ia+, (1)), (c1, and (d), respectively. These (a) ,Q)),t
e), (dl), (a), (b), (c1
, 1dl so that the missions do not overlap, the result is obtained as shown in the figure.

Therefore, the time tI % t3, L4, ts, L is obtained as the scheduling time by decentralization.

(2) Determination of time by centralization: for example, Figure 13 (b)
The time shown in ta+, (b), IcI, (dl of
Figure 1 (a), (c), (dl) are respectively calculated as shown schematically.
Sort all the times found in l from earliest to earliest. If missions are selected from the beginning of the sorted time so that they do not overlap, the result shown in the figure is obtained. Therefore,
By centralization, time j is set as the scheduling time.
6, tll), t7, t), tI), t, are obtained.

■ Finds a new environment by subtracting the resource consumption of the mission from the current remaining resource amount. This is, for example, the first
Figure 4 (a) Figure 14 (b), which is obtained by subtracting the mission request pattern (resource consumption) from the current resource remaining amount
Seek resources as a new environment.

[Phase] determines the next scheduling trial mission (0 to [Phase]), and proceeds to Fig. 4 ■'. At this point, if there is no next mission, the game ends.

FIG. 7 shows an example mission object of the present invention.

FIG. 7(a) schematically represents the power consumption pattern. Here, the horizontal axis represents time, and the vertical axis represents resource consumption (KW).

FIG. 7(b) schematically represents the crew work consumption pattern. Here, the horizontal axis represents time, and the vertical axis represents resource consumption (people).

FIG. 7(c) represents mission object A. FIG. This describes, as an object, a mission (experiment) having power consumption patterns and crew work consumption patterns schematically shown in FIGS. 7(a) and 7(b). Here, mission object A has: - Priority 3 - Mission time 5 hours - Operational model October 1, 1990 to 1990-1
It consists of: October 10th, number of experiment repetitions: 10 times, electric power A', crew A'', etc.

Similarly, the power object A' describes the pattern shown in the seventh part (a), and the crew object A'' describes the pattern shown in FIG. 7 (b).

FIG. 8 shows an example resource object of the present invention.

FIG. 8(a) schematically represents the power supply pattern. Here, the horizontal axis represents time, and the vertical axis represents resource consumption (KW).

FIG. 8(b) schematically represents the crew work supply pattern. Here, the horizontal axis represents time, and the vertical axis represents resource consumption (people).

FIG. 8(C) shows a power object, which is described as an object having the power supply pattern schematically shown in FIG. 8(B). Here, the power object is composed of: ・Priority 2 ・Supply pattern, area, etc.

FIG. 8(d) represents a crew work object. This is described as an object having a crew work supply pattern schematically shown in FIG. 8(b). Here, the crew work object is composed of: ・Priority 3 ・Supply pattern, area, etc.

FIG. 9 shows an example of a scheduling condition object of the present invention. This includes schedule start time, period, mission selection coefficients E1, E2, E3, E4 as shown in the figure.
, E5, is an object for the designer to preset the maximum value of resource priority.

FIG. 10 shows an explanatory diagram of scheduling fine adjustment related to the present invention. As explained in Figure 5 [Phase] to [Phase], when there is no schedulable time period for a mission, one resource is ignored,
The remaining common part is defined as a tentative schedulable time zone, and when there is an overlap among other missions that have already been scheduled, if this other mission can be moved, the tentative schedulable time zone becomes the schedulable time. FIG. 4 is an explanatory diagram of WX adjustment of scheduling in which a band is determined.

FIG. 11 shows examples of scheduling candidate times according to the present invention. This is an explanatory diagram of a specific example of time selection based on the schedulable time zone. 6 See Figures 0 to ■.

FIG. 12 shows a scheduling time determination method according to the present invention. Here, decentralization is a method of distributing and scheduling missions as much as possible, and centralization is a method of scheduling missions as early as possible.

Both are fa in Figure 6 @, @, [phase], and ■ as described above.
), ('b), te), and (d) are found. In the case of decentralization, as mentioned above, (a), middle), f
Arrange the scheduling times in the order of c) and +d) so that the missions do not overlap (select, Figure 13 (a)
reference). In case of centralization, la), (b), (c),
After sorting all the times in (d) from earliest to earliest,
Arrange the scheduling times from the beginning so that the missions do not overlap (select, see Figure 13 (b)).

FIG. 13 shows an example of scheduling time determination according to the present invention. fa), (b), (c1, (dl are Fig. 6 0,
An example of time obtained using @, [phase], and ■ is shown.

FIG. 13(a) shows an example of determining scheduling time by decentralization. Regarding the times obtained in (a), (1)), fc), and (d), missions are selected in the order of fa), mountain), fc), and fd) so as not to overlap, and as a result, the missions are as shown in the figure. Get the scheduling result.

FIG. 13(b) shows an example of determining the scheduling time by centralization. After sorting the times obtained by (a), (b), and (c) fdl in ascending order, the missions are selected from the beginning so that they do not overlap, and as a result, the mission scheduling results are obtained as shown in the figure.

FIG. 14 shows an explanatory diagram of new environment generation according to the present invention.

Figure 14 (a) shows the current remaining resource amount, and the 14th
Figure (b) shows the remaining resource amount after subtracting the mission request pattern (consumption pattern) from the current resource remaining amount. Perform scheduling.

〔Effect of the invention〕

As explained above, according to the present invention, the selection of a scheduling trial mission, the calculation of a schedulable time period (if a schedulable time period does not exist, further fine-tuning of the scheduling is performed), and the determination of a scheduling time are decentralized. Alternatively, since a configuration is adopted in which mission scheduling is performed centrally, mission scheduling can be determined within a practical time and memory range.

In particular, because it does not use the conventional pack-track mechanism of trial and error selection when scheduling missions, it uses heuristic methods to find optimal scheduling, which makes it difficult to perform highly complex scientific observations and experiments on the space station. It becomes possible to quickly create mission schedules under constraint conditions using a common algorithm even for missions that use different types of resources.

[Brief explanation of drawings]

Figure 1 is a block diagram of the principle of the present invention. Figure 2 is a system configuration diagram of the present invention. Part 3, Figures 4, 5, and 6 are flowcharts explaining the operation of the present invention. Figure 7 is the mission of the present invention. Object example FIG. 8 is an example of a resource object according to the present invention FIG. 9 is an example of a scheduling condition object FIG. 10 is an illustration of scheduling fine adjustment according to the present invention FIG. FIG. 13 is an example of scheduling time determination according to the present invention. FIG. 14 is an explanatory diagram of new environment generation according to the present invention. FIG. 15 is a conceptual diagram of the operation of the mission scheduling device according to the present invention. indicates an explanation of the scheduling constraints. In the figure, 1: Mission scheduling device 2: Initial condition setting 3: Scheduling trial mission determining means 4: Scheduling possible time slot calculating means 5: Scheduling fine adjustment means 6: Scheduling time determining means 7: New environment generating means

Claims (4)

[Claims] (1) In a mission scheduling device that determines a mission schedule under constraint conditions including resource supply capacity and resource consumption by missions, scheduling trial mission determination that determines the scheduling trial order for multiple missions to be scheduled. means (3), and a schedulable time slot calculation means (4) that calculates a schedulable time slot based on resource consumption and resource supply for the missions in the order determined by the scheduling trial mission determination means (3). ), and a scheduling time determining means (6) for calculating the best scheduling time by decentralization or centralization based on the schedulable time zone calculated by the schedulable time zone calculating means (4), The schedulable time slot calculation means (4) calculates the schedulable time slot for the missions in the trial order determined by the scheduling trial mission determination means (3) for the plurality of target missions, and the scheduling time determination means (
6) calculates the best scheduling time corresponding to decentralization or centralization from among these available scheduling time periods, calculates the new resource supply amount by subtracting the resource consumption when executing at this scheduling time, and then A mission scheduling device characterized in that it is configured to repeatedly perform missions in a trial order and determine a mission scheduling time. (2) The above scheduling trial mission determination means (
3) Groups multiple missions to be scheduled according to preset priorities, and if there is the same group, selects the resource with the lowest current remaining resource percentage, and selects the resource with the lowest current remaining resource percentage, and then assigns the selected resource to the same group. The mission scheduling order is determined based on the higher the order of pattern length, the lower the pattern height, the larger the pattern area, the larger the number of patterns, and the greater the number of experiment repetitions. A mission scheduling device according to claim (1). (3) The above scheduling trial mission determination means (
For the mission determined by 3), when resources cannot be allocated and there is no schedulable time slot, one of the resources is ignored and a tentative schedulable time slot is found for the other resources, and the provisional schedulable time slot is Claim (1) characterized in that a scheduling fine adjustment means (5) is provided for determining the temporary scheduling possible time period as the scheduling possible time period when it is determined that an existing already allocated mission can be moved. , the mission scheduling device according to paragraph (2). (4) Means for calculating the available scheduling time (4)
However, the above scheduling trial mission determination means (3
), based on the resource consumption amount and resource supply amount, for the missions in the order determined by (c) Time based on the operational model (d) Find the times at both ends of the schedulable time range, and set the priority order of (a), (b), (c), and (d) from the beginning to avoid overlapping missions. Determine the scheduling time (distributed), or sort the times in (a), (b), (c), and (d) from earliest to lowest, and then determine the scheduling time from the beginning to avoid overlapping missions (centralized). 3. The mission scheduling device according to claim 1, wherein the mission scheduling device is configured to: