JP2002297848A - Attending schedule preparing device for telephone operator, schedule preparing device, wait time estimating device, methods for the above, program and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To obtain the optimum solution in preparing an attending schedule for a plurality of telephone operators and to estimate the wait time. SOLUTION: A day scheduler 30 prepares an attending schedule of the plurality of operators, and at that time, the optimizing processing for obtaining the optimum solution using SA method is conducted to obtain an attending schedule table 70. On the other hand, an evaluation value creasing part 10 performs queuing operation taking the tolerance time of a calling subscriber, talking time based on the past actual data, a call interval distribution showing the number of calls to the designated time, and the number of operators as parameters, thereby obtaining the evaluation value for evaluating the solution in the optimizing processing.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a reusable person,
Telephone operator's landing schedule creation device, method, schedule creation device, method, waiting time estimation device, method, and computer executed by each device for creating a work schedule of resources such as devices so as to obtain predetermined target values And a computer-readable recording medium on which the program is recorded.

[0002]

2. Description of the Related Art Conventionally, a plurality of operators in a call center arrive at an exchange board, for example, every 15 minutes as one frame. In this case, the total number of frames per day, the number of continuous work frames, the number of lunch break frames, the number of other break frames, etc. are determined by work rules such as the Labor Standards Law, and a plurality of daily work patterns that differ depending on each operator. (Start work /
End time, night shift, etc.). Further, there is a variation in the ability of the operator.

Therefore, when preparing a landing schedule for each operator, a predetermined target value (short waiting time of caller, small loss probability) while satisfying the above conditions is satisfied.
Needs to be obtained. For this purpose, specialized schedulers have traditionally created a manual method based on experience and intuition, enumerated all combinations using a computer, calculated each combination, and found an exact solution to find the best one. Algorithm) has been used.

[0004]

However, the method that is manually performed by a scheduler requires a great deal of time, and it is not possible to evaluate whether or not the obtained solution is an optimal solution. In many cases, the system is operated, and therefore, there has been a problem that the waiting time and the call blocking rate become lower than the target values. Further, in the case of the exact solution using a computer, there is a problem that if the number of operators increases and the scale increases, the number of combinations increases exponentially, and it becomes impossible to obtain a good quality solution. further,
The method using the GA method is the result of crossover by mutation operation,
The problem is that it takes time to find the optimal solution because there is a possibility that a child (solution) that inherits only the bad part of the parent can be created and a solution that is far from the optimal solution can be created by mutation. there were.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and has as its object to enable an optimum solution to be obtained in a short time and to be able to predict a waiting time.

[0006]

In order to achieve the above object, a telephone operator's mounting schedule creating apparatus according to the present invention performs an optimization process for finding an optimal solution using the SA method. Schedule creation means for creating an operator's arrival schedule, caller's patience time, call time based on past actual data, call interval distribution indicating the number of calls for a predetermined time, and the number of operators as parameters Evaluation value generating means for obtaining an evaluation value for evaluating a solution generated in the optimization processing by performing a queue operation;

[0007] In addition, the telephone operator's landing schedule creation method according to the present invention creates an arrival schedule for a plurality of operators and performs call processing by performing optimization processing for finding an optimal solution using the SA method. By performing a queuing operation using the patience time, call time based on past actual data, call interval distribution indicating the number of calls for a predetermined time, and the number of operators as parameters, a solution generated in the above-described optimization processing is obtained. An evaluation value for obtaining an evaluation value for performing the evaluation is created.

Further, a schedule creation device according to the present invention is a schedule creation device for each processing means in a system in which a plurality of processing means perform processing in response to processing requests arriving at random. The schedule creation unit creates a processing schedule for a plurality of processing units by performing an optimization process for determining the processing time, the processing requester's patience time, the processing time based on past actual data, and the number of requests for a predetermined time. An evaluation value generating means for obtaining an evaluation value for evaluating a solution generated in the optimization processing by performing a queue operation using the request interval distribution and the number of processing means as parameters is provided.

A schedule creation method according to the present invention is a schedule creation method for each processing means in a system in which a plurality of processing means perform processing in response to a processing request that arrives at random. By performing an optimizing process for obtaining a processing schedule, a processing schedule of a plurality of processing means is created, and a request interval distribution indicating a patience time of a processing requester, a processing time based on past actual data, and a request number for a predetermined time. By performing a queue operation using the parameter and the number of processing means as parameters, an evaluation value for evaluating a solution generated in the optimization processing is created.

[0010] Further, the waiting time prediction device according to the present invention comprises:
A waiting time estimating device for estimating the waiting time of the processing requester in a system in which a plurality of processing means performs processing in response to a processing request that arrives at random. There is provided an operation means for calculating the waiting time by performing a queue operation using the processing time based on the request interval distribution indicating the number of requests for a predetermined time and the number of processing means as parameters.

Further, the waiting time prediction method according to the present invention comprises:
A waiting time prediction method for predicting the waiting time of the processing requester in a system in which a plurality of processing units perform processing in response to processing requests arriving at random, comprising: a patience time of the processing requester; The waiting time is predicted by performing a queue operation using a processing time based on the request, a request interval distribution indicating the number of requests for a predetermined time, and the number of processing means as parameters.

In addition, the telephone operator's landing schedule creation device and the program used in the schedule creation device according to the present invention perform an optimization process for finding an optimum solution by using the SA method, so that the processing schedule of a plurality of processing means can be improved. Schedule creation processing, processing requester's patience time, processing time based on past actual data,
An evaluation value creation process for obtaining an evaluation value for evaluating a solution generated in the optimization process by performing a queue operation using a request interval distribution indicating the number of requests for a predetermined time and the number of processing means as parameters. Is a program for causing a computer to execute.

The program used in the waiting time estimating apparatus according to the present invention includes a processing requester's patience time, a processing time based on past actual data, a request interval distribution indicating the number of requests for a predetermined time, and a processing means. This is a program for causing a computer to execute an operation for calculating a waiting time of a process requester by performing a queue operation using a number as a parameter.

[0014]

Therefore, according to the present invention, there is provided a patience time of a requester such as a caller, a processing time such as a talk time, a request interval distribution indicating the number of requests such as the number of calls for a predetermined time, an operator and the like. By performing a queue operation using the number of processing means as a parameter, an evaluation value for evaluating a solution generated in the optimization process is obtained, and an optimization process for obtaining an optimum solution using the SA method is performed. At this time, by evaluating the solution using the evaluation value, a good processing schedule of the processing means such as a plurality of operators can be created in a short time. Also, the waiting time and the call blocking rate according to the time can be predicted using the evaluation value.

[0015]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration of an operator's mounting schedule creation device used in a call center according to an embodiment of the present invention. In FIG. 1, an evaluation value creation unit 10 includes a patience time (a patience wait time when a caller waits until a caller makes a call and the other party comes out),
Talk time (processing time) distribution, call interval distribution [number of calls / 1 sec], and the number of operators working on that day are input as parameters, and a queue operation is performed based on these parameters to obtain an evaluation value of 20. Create and output as a file called dat.

The daily scheduler 30 is an SA (Simula), which is an algorithm for finding an optimal solution.
By performing an optimizing process for obtaining an optimal solution in consideration of the above-described conditions using the ted Annealing method, a landing schedule for each operator is created. At this time, in order to evaluate the solution generated by the above-described optimization processing, the evaluation value creation unit 10
Is used. And, for example, FIG.
And a graph showing the average waiting time of the caller at each time of day as shown in FIG. 9 is output.

For this purpose, the above parameters, the above various conditions and the like are inputted in advance from the input unit 40 via the input / output control unit 50 and set. The landing schedule table and the graph indicating the average waiting time can be displayed on the display unit 60 via the input / output control unit 50. Further, the daily scheduler 30 may create and display a rotation table for each operator based on the arrival schedule table.

Next, the evaluation value creating section 10 will be described. First, input parameters will be described with reference to FIG. Since it has been proven that the call interval distribution can be approximated by an exponential distribution, as shown in FIG. 2A, the call interval distribution is input as an exponential distribution P (t) where the average number of calls per second based on past actual data is λ. Is done. The administrator selects which curve in FIG. 2A is to be selected according to the situation expected on that day.

As the call time, a step function P [j] based on a call time distribution modeled on the basis of past actual data as shown in FIG. 2B is used. The endurance time is constant at t0 as shown in FIG. 2 (c), and here, for example, t0 = 7 seconds. Further, the number of operators n is set. The evaluation value creation unit 10 obtains an evaluation value (risk.dat) 20 for evaluating the waiting time (and / or the call blocking rate) by a queue operation using these four parameters.

Next, the evaluation value 20 will be described with reference to FIGS. FIG. 3 is a diagram for explaining the queue operation when the number of operators is n. In the figure, t0: patience time tn: time the call arrived (n: 1 to m) τ: time until the first available operator becomes available when the call arrives (waiting time) tm: next call Tn: Talk time (processing time, n: 1 to m) τsum: Total waiting time τav: Average waiting time

In FIG. 3, when a call is received at time tm1 between T1 and Tn when all the operators are closed, for example, it is expected from the call time Tn that the operator 1 will be available first at time tm2. Τ = t
m1 to tm2 is the waiting time. Therefore, if this τ exceeds the endurance time t0, call blocking occurs.

FIG. 4 shows the evaluation value 20. As shown, the number of operators op (X-axis) and the call interval distribution (Y
3 is a graph in which the waiting time pr (or call loss) (Z-axis) with respect to (axis) is expressed in three-dimensional coordinates.

That is, the evaluation value 20 is a combination of the waiting times corresponding to the time for all combinations of the number of operators and the number of calls, of which the waiting time within the patience time t0 is obtained.

FIG. 5 is a flowchart of a queue calculation process for obtaining the evaluation value 20 of FIG. In FIG. 5, first, tm is determined at an appropriate time (S1), and it is determined whether or not any vacant operator is present at that time (S1).
2) If not, find the expected waiting time τ (S
3). Next, it is checked whether this τ is within the waiting time t0 (= 7 seconds) (S4), and if it is not within t0, it is counted as a call loss (S5). If τ is within t0, the average waiting time τav is determined. Here, 1 indicates the number of people who have been waiting until now (S6). Then, tm is newly added based on τav.
(S7), and after determining the talk time Tm (S8), S
Return to 1.

FIG. 6 shows an example of the evaluation value 20 obtained as described above. As shown in the figure, the evaluation value 20 is shown as a combination of the values of the number of calls, the waiting time, and the call blocking rate for each of the cases of 1 to n operators. In FIG. 6, for example, when there is one operator, the waiting time is 3.014967 seconds and the call loss rate is 0.204450 when the number of calls is one, and the waiting time is 6.747799 when the number of calls is four. Seconds, call blocking rate is 0.53341
0 is shown. When the number of operators is three, the waiting time is 0.5469 when the number of calls is one.
74 seconds, the call blocking rate is 0.002440, and when the number of calls is three, the waiting time is 1.348658 seconds and the call blocking rate is 0.045520. Such an evaluation value 20 is calculated by S
It is used to evaluate the solution generated by the optimization process using the method A.

Next, the daily scheduler 30 will be described. FIG. 7 is a flowchart showing the optimization process by the SA method performed by the daily scheduler 30. First, a schedule Sch is prepared as an initial value of a landing schedule table finally obtained as an optimal solution shown in FIG. 8 (S
11). The schedule Sch as the initial value is created in consideration of various conditions input from the input unit 40 in advance.

Various conditions include: the number of operators n, the number of calls (traffic value) according to the time predicted for one day. At present, the total number of frames per day, the number of consecutive frames, the number of lunch breaks, the number of breaks, the number of daily work patterns depending on the operator, the operator's ability, etc. Things that cannot be done. Multiple work patterns include, for example,
For example, FIG. 8 in which the end time, night shift time, and the like are shifted.
And patterns A, B to F as shown in FIG.

In addition, it may be set in consideration of a change in the number of calls due to, for example, a day of the week, a month, a season, or the like, or a change in the number of calls of the day due to special circumstances. These conditions are set and changed by the administrator, and are input from the input unit 40 via the input / output control unit 50.

Next, the schedule Sch is evaluated with an evaluation value of 20.
The evaluation f is obtained by performing evaluation using (S1).
2). Next, a mutation operation is performed to generate a schedule Sch ′ that is a modification of the schedule Sch (S13). This schedule Sch 'is a mutation operation A, B, C composed of three patterns for respectively modifying the schedule Sch.
Are prepared in advance, and Sch is modified based on one mutation operation arbitrarily selected among them.

The three mutation operations A, B, and C have slightly different degrees of deformation. For example, A: One frame is exchanged in the first half or the second half of the day of the operator. B: One frame in the first half and one frame in the second half are exchanged. C: Change the start time and end time.

Next, by evaluating the schedule Sch ′ using the evaluation value 20, an evaluation f ′ is obtained (S14). Then, f and f ′ are compared (S15),
When f 'is high, f' is left, but when f 'is worse than f, one of them is left with a certain probability. And
Returning to S13, any one of the patterns A, B, and C is selected again, a schedule Sch 'is created, and evaluation is performed. Hereinafter, an optimum solution can be obtained by repeating the same processing a specified number of times (for example, 100,000 times or more).

The above evaluation method combines the number of calls with the number of operators for each of the number of operators, and the evaluation value 2 shown in FIGS.
A combination corresponding to each combination is obtained from 0, and the rate at which a combination in which the waiting time is within the patience time is obtained is evaluated f and f '. Note that the rate at which a combination in which the waiting time exceeds the patience time is obtained is referred to as a blocking rate.

Based on the optimal solution, a "good mounting schedule table 70" can be created and output as a file called Sch.dat (S17, S18). here,
"Good landing schedule table" means that the waiting time is within the tolerable time t0 (7 seconds) and the average waiting time is small,
That is, the target value has been achieved, and FIG. 8 shows an example thereof. In the example of FIG. 8, the number of operators is 16 and one frame is a unit of 15 minutes. A black frame indicates a stand, and a white frame indicates a break. As described above, in the schedule table, a row of a plurality of frames is represented as a landing pattern.

A graph 80 indicating the average waiting time at each time in one day as shown in FIG. 9 can also be created as a file called wait.dat. You can also create tables.

As described above, in the optimization processing according to the present embodiment, the use of the SA method may cause crossover or mutation as in the conventional GA method when performing a mutation operation for generating a solution. Since there is no such solution, an extremely bad solution does not appear, and the optimal solution can be obtained reliably in a short time statistically with a small number of iterations. In the results of the experiment,
Scheduling could be performed in a short time of 1/30 of the conventional manual operation time.

In the optimization process, one of the three mutation operations is selected to generate a solution. In evaluating the solution, all the good ones are adopted, but those with bad evaluation values are not always rejected. Therefore, while generating bad examples several times, it is likely that the next time a better solution will be found than before, and the speed can be increased. For example, if you look for the highest mountain peak in a series of mountains, if you go only a good solution (high direction), you will go to the top of a small mountain But you can imagine that you have to go down once to get to another high mountain.

Since the optimal solution obtained by the above-described optimization processing is strictly the closest to the optimal solution, in some cases, it is predicted that the waiting time exceeds the patience time t0 in one day. Sometimes. In such a case, the replenishment operator is laid only at that time, so that the operator's useless work at other times can be eliminated.

It is to be noted that 4 to be input to the evaluation value creation unit 10
In order to obtain one parameter and various conditions set in the one-day scheduler 30, necessary data such as data such as the number of calls (traffic value) may be obtained in advance from a private branch exchange (PBX). By doing so, it is not necessary for the administrator to input and set parameters, various conditions, and the like from the input unit 40, and necessary data can be obtained automatically.

Although the embodiment described above is an apparatus for preparing a landing schedule for a telephone operator, the technical concept of this apparatus can be applied to a general schedule preparing apparatus and a waiting time estimating apparatus. In this case, the schedule creation device processes a plurality of processing means (corresponding to an operator and resources such as reusable persons and devices) in response to a randomly-arrived processing request (corresponding to a call in a call center). In a system in which the processing is terminated, a schedule in charge of each processing means can be created.

In the above-described embodiment of the schedule creation device, the four parameters to be input to the evaluation value creation unit are the request interval distribution, the processing time, the patience time of the requester (requester or device), and the number of processing means. Is used. The evaluation value creation unit creates an evaluation value by performing a queue operation using these parameters. The schedule creating means corresponding to the one-day scheduler performs an optimization process while generating a solution by selecting one of the three mutation operations A, B, and C using the SA method. By evaluating the generated solution using the above evaluation value, it is possible to create a schedule table of each processing means and a graph indicating the waiting time of the requester according to the time.

Further, in the embodiment of the waiting time prediction device, the evaluation value creating unit performs a queue operation using the four parameters of the request interval distribution, the processing time, the patience time, and the number of processing means. A waiting time (predicted waiting time) according to the time can be obtained.

Such a schedule creation device and a waiting time prediction device include, for example, a use schedule and a waiting time of a plurality of elevators, a use schedule of a seat of a restaurant and the like, a waiting time of a customer, a schedule of a plurality of receptionists and a customer. Widely applicable when multiple people or devices perform some kind of processing (work) in response to a processing request that arrives at random timing, such as when you want to know the waiting time, the use schedule and waiting time of multiple boxes in the toilet, etc. can do.

Next, a program according to an embodiment of the present invention and a recording medium for recording the program will be described. A program for executing the processing based on the above-described operation of FIG. 1 and the processing shown in each flowchart by the CPU of the computer system in the telephone operator's landing schedule creation device, schedule creation device, and waiting time prediction device according to the present invention includes: A program according to the present invention is configured.

The recording medium for recording the program constitutes the computer-readable recording medium according to the present invention. As this recording medium, a magneto-optical disk, an optical disk, a semiconductor memory, a magnetic recording medium, or the like can be used.
M, a floppy (registered trademark) disk, a memory card, or the like.

The recording medium is a fixed medium such as a volatile memory such as a RAM inside a computer system serving as a server or a client when a program is transmitted through a network such as the Internet or a communication line such as a telephone line. The one that holds the time program is also included.

The above program may be transmitted from a computer system in which the program is stored in a storage device or the like to another computer system via a transmission medium or by a transmission wave in the transmission medium.
The transmission medium refers to a medium having a function of transmitting information, such as a network (communication network) such as the Internet or a communication line (communication line) such as a telephone line.

Further, the above program may be for realizing a part of the functions described above. Furthermore, what can realize the above-described function in combination with a program already recorded in the computer system, that is, a so-called difference file (difference program) may be used.

Therefore, the program and the recording medium are used in a system or apparatus different from the system or apparatus shown in FIG. 1, and the computer of the system or apparatus executes this program, so that the functions described in each embodiment can be achieved. The same function and effect as those described above can be obtained, and the object of the present invention can be achieved.

[0049]

As described above, according to the present invention, S
A schedule is created by performing an optimization process for obtaining an optimum solution by using the method A, and an evaluation value is obtained by performing a queue operation using predetermined parameters, and the evaluation value is generated in the optimization process. Because it is used to evaluate solutions, it is possible to create an optimal schedule of processing means such as a telephone operator in a short time and to predict a waiting time of a processing requester such as a caller. Can be.

In the optimization process, when a solution is generated by a mutation operation, one of a plurality of previously prepared mutation operations is arbitrarily selected, and a solution is generated according to the selected mutation operation. As a result, the optimization process can be performed more reliably in a short time, and a more optimal schedule can be created.

Further, since the initial value of the solution and / or the parameter setting can be changed, it is possible to cope with various conditions when creating a schedule.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an apparatus for creating a landing schedule of a telephone operator according to an embodiment of the present invention.

FIG. 2 is a graph illustrating parameters used for creating an evaluation value.

FIG. 3 is a timing chart for explaining evaluation values.

FIG. 4 is a graph for explaining evaluation values.

FIG. 5 is a flowchart illustrating a queue calculation process for obtaining an evaluation value.

FIG. 6 is a configuration diagram illustrating an example of an evaluation value.

FIG. 7 is a flowchart showing a process for obtaining an optimal solution using the SA method according to the embodiment of the present invention.

FIG. 8 is a configuration diagram showing an example of a arrival schedule table created by the arrival schedule creation device of the telephone operator according to the embodiment of the present invention.

FIG. 9 is a graph showing the estimated waiting time for each time of day, which is created by the telephone operator's boarding schedule creating apparatus according to the embodiment of the present invention.

[Explanation of symbols]

 Reference Signs List 10 evaluation value creation unit (queue operation) 20 evaluation value (risk.dat) 30 1-day scheduler (SA method) 40 input unit 50 input / output control unit 60 display unit 70 arrival schedule table (sch.dat) 80 each time Average wait time (wait.dat)

 ────────────────────────────────────────────────── ─── Continued on the front page (72) Inventor Kazuyuki Yoshimura 2-3-1 Otemachi, Chiyoda-ku, Tokyo Nippon Telegraph and Telephone Corporation

Claims (24)

[Claims] 1. A schedule creation means for creating a landing schedule for a plurality of operators by performing an optimization process for finding an optimal solution using the SA method, a caller's patience time, and past actual data. Based on a call time distribution, a call interval distribution indicating the number of calls for a predetermined time based on past actual data, and a queue operation using the number of operators as parameters, the solution generated in the optimization process is evaluated. And an evaluation value generating means for obtaining an evaluation value for the telephone operator. 2. The schedule creating means, when generating a solution by a mutation operation in the optimization process, arbitrarily selects one of a plurality of mutation operations prepared in advance,
2. The device according to claim 1, wherein a solution is generated according to the selected mutation operation. 3. An apparatus according to claim 2, further comprising setting means for changing the initial value of the solution and / or setting of the parameter. 4. The apparatus according to claim 3, wherein the initial value of the solution is determined by constraints on the working hours of the operator. 5. The landing schedule of the operator,
Expressed in a mounting pattern by a predetermined array of mounting time units in one day, the plurality of mutation operations are obtained by exchanging one mounting time unit in the first half or the latter half of the daily schedule. One in which the first half of the arrival time unit is replaced with the second half of the arrival time unit, and one in which the start and end times of the daily schedule are changed.
5. The device according to claim 2, wherein the telephone operator has a schedule. 6. The telephone according to claim 2, wherein data necessary for obtaining an initial value of the solution and / or the parameter is obtained from a private branch exchange. An operator's boarding schedule creation device. 7. An optimizing process for finding an optimal solution by using the SA method creates a landing schedule of a plurality of operators, and sets a caller's patience time and a call time based on past actual data. And a call interval distribution indicating the number of calls for a predetermined time, and performing a queue operation using the number of operators as parameters to obtain an evaluation value for obtaining an evaluation value for evaluating a solution generated in the optimization process. A method of creating a boarding schedule for a telephone operator. 8. When generating a solution by a mutation operation in the optimization process, arbitrarily selecting one of a plurality of mutation operations prepared in advance and generating a solution according to the selected mutation operation. The method according to claim 7, wherein the telephone operator sets a landing schedule. 9. The method according to claim 7, wherein the initial value of the solution and / or the setting of the parameter can be changed. 10. A schedule creation device for each processing means in a system in which a plurality of processing means perform processing in response to processing requests arriving at random, wherein an optimization process for obtaining an optimum solution using an SA method is performed. A schedule creation means for creating a processing schedule for a plurality of processing means, a processing requester's patience time, a processing time based on past actual data, a request interval distribution indicating the number of requests for a predetermined time, and a number of processing means. And a queue calculation using the parameters as parameters to obtain an evaluation value for obtaining an evaluation value for evaluating a solution generated in the optimization process. 11. The schedule creating means, when generating a solution by a mutation operation in the optimization process, arbitrarily selects one of a plurality of mutation operations prepared in advance and responds to the selected mutation operation. 11. The schedule creation device according to claim 10, wherein the solution is generated by performing the following. 12. The schedule creation apparatus according to claim 10, further comprising setting means for changing the initial value of the solution and / or the setting of the parameter. 13. A method for creating a schedule for each processing means in a system in which a plurality of processing means perform processing in response to processing requests arriving at random, wherein an optimization process for obtaining an optimum solution using an SA method is performed. In addition to creating a processing schedule for a plurality of processing means, the processing requester's patience time, processing time based on past actual data, a request interval distribution indicating the number of requests for a predetermined time, and the number of processing means A schedule calculation method for creating an evaluation value for evaluating a solution generated in the optimization process by performing a queue operation as follows. 14. When a solution is generated by a mutation operation in the optimization process, one of the plurality of mutation operations prepared in advance is arbitrarily selected, and a solution is generated according to the selected mutation operation. 14. The method according to claim 13, wherein: 15. The method according to claim 13, wherein an initial value of the solution and / or a setting of the parameter can be changed. 16. A waiting time estimating apparatus for estimating a waiting time of a processing requester in a system in which a plurality of processing means perform processing in response to a processing request randomly received, comprising: a patience time of the processing requester; Arithmetic means for calculating the waiting time by performing a queue operation using a processing time based on past actual data, a request interval distribution indicating the number of requests for a predetermined time, and the number of processing means as parameters. Waiting time prediction device. 17. The waiting time estimating apparatus according to claim 16, further comprising setting means for changing the setting of the parameter. 18. A waiting time estimating method for estimating a waiting time of a processing requester in a system in which a plurality of processing means performs processing in response to a processing request that arrives at random, comprising: a patience time of the processing requester; A waiting time prediction, wherein the waiting time is calculated by performing a queue operation using a processing time based on past actual data, a request interval distribution indicating the number of requests for a predetermined time, and the number of processing means as parameters. Method. 19. The waiting time prediction method according to claim 18, wherein the setting of the parameter can be changed. 20. A schedule creation process for creating a process schedule of a plurality of processing means by performing an optimization process for finding an optimal solution using the SA method, a process requester's patience time, and past actual data. Based on a processing time, a request interval distribution indicating the number of requests for a predetermined time, and a queue operation with the number of processing means as parameters, an evaluation value for evaluating a solution generated in the optimization processing is calculated. A program for causing a computer to execute the required evaluation value creation processing. 21. Performing a queue operation by using the patience time of the processing requester, the processing time based on past actual data, the request interval distribution indicating the number of requests for a predetermined time, and the number of processing means as parameters. A program for causing a computer to execute an arithmetic process for calculating a waiting time of a process requester. 22. The schedule creation process, when generating a solution by a mutation operation in the optimization process, arbitrarily selects one of a plurality of mutation operations prepared in advance, and responds to the selected mutation operation. 22. The program according to claim 21, wherein the solution is generated by: 23. The apparatus according to claim 2, wherein an initial value of the solution and / or a setting change process of the parameter are provided.
23. The program according to 1 or 22. 24. A computer-readable recording medium on which the program according to claim 20 is recorded.