JP7709931B2 - Economic evaluation simulation device and economic evaluation simulation method - Google Patents

Description

This disclosure relates to a simulation device and a simulation method for evaluating economic efficiency.

Research and development is being conducted on methods to evaluate the economic viability, business viability, productivity, etc. of robots or machines in advance.

For example, Patent Document 1 discloses a production performance evaluation device for a machine system that analyzes production performance in a machine system that automatically produces products, which defines the state transition probability between the part states before and after each task and the work time for each task for a specific part state among multiple tasks of a machine system that performs multiple tasks in sequence, extracts all transition paths from an initial state to a target state through multiple part states based on the state transition probability, calculates a path cycle time for each transition path from the state transition probability and the work time, and calculates an expected takt time, which is the time it takes for one part in the machine system to reach the target state, from the path cycle time and the extracted transition paths.

Patent Document 2 also discloses a business feasibility evaluation device that clearly separates the determination of the income/expense structure from the input of prerequisites, and aims to visualize changes in management indicators when the prerequisites are interactively changed. The device accepts input from a user of a calculation process for calculating values of evaluation items for evaluating a business and prerequisites that are input values for the calculation process, creates values of the evaluation items in a time series based on the prerequisites and the calculation process, displays the values of the evaluation items created in a time series, accepts re-input of prerequisites by the user, recreates values of the evaluation items in a time series based on the re-input prerequisites and calculation process, and displays the values of the evaluation items created again in a time series.

Patent Document 3 discloses an equipment evaluation system that estimates a probability density distribution of a parameter representing the performance of a target equipment based on measurement data of the target equipment measured up to each of a plurality of times, and stores the usage pattern in association with the probability density distribution of the target equipment, identifies a usage pattern of the target equipment similar to the usage pattern of a first equipment different from the target equipment, and evaluates future performance deterioration of the first equipment using a probability density distribution group corresponding to the identified usage pattern, for the purpose of evaluating performance deterioration of the equipment over a long period in anticipation of the entire life cycle. Patent Document 3 also discloses that Bayesian estimation can be used to estimate parameters, and that methods for determining posterior probability density distributions include Markov chain Monte Carlo methods (MCMC: Markov chain Monte Carlo methods including the Gibbs method and the Metropolis method, and a particle method, which is a type of sequential Monte Carlo method, etc.

Patent No. 6516663 JP 2005-32023 A JP 2016-177682 A

For example, when using simulations to predict the productivity of work performed by machines or robots in the field, such as in construction or civil engineering work, it is not possible to accurately determine in advance the physical properties and boundary conditions that serve as inputs for the simulation. Also, because the tasks and environments are complex, it is difficult for the simulation model itself to accurately mimic the behavior of the robot. For example, in civil engineering work, productivity varies greatly depending on the condition of the soil at the site (hardness, moisture content, etc.), the topography, weather conditions, etc., making it difficult to make accurate predictions in advance using simulations.

In conditions like these where predictions through prior simulations are difficult, it is difficult to directly apply the productivity evaluation method at a production site where the operating environment is relatively predictable, as in Patent Document 1.

In addition, in the business feasibility evaluation device described in Patent Document 2, in order to evaluate the profitability of a business, a probability distribution is assigned to items related to business feasibility, and the business feasibility evaluation is also performed probabilistically. However, even in this case, the target is mainly plants and other environments with relatively constant conditions.

The equipment evaluation system described in Patent Document 3 also relates to the technical field of remotely providing various services for monitoring, controlling, and diagnosing facilities such as apartment complexes, buildings, and plants, so the environmental conditions are not so unstable that they require prior prediction. For this reason, it is believed that the system focuses on the ex post probability density distribution.

The primary purpose of this disclosure is to improve the accuracy of simulations for evaluating the economic viability of robotics businesses and other projects that deal with unknown environments where physical properties, boundary conditions, etc. cannot be determined accurately in advance.

The economic evaluation simulation device disclosed herein includes a work simulation execution unit that performs a simulation of the work content in a business, an economic parameter input unit that inputs economic parameters, an economic simulation execution unit that performs a simulation of economics using the economic parameters, an actual operation data acquisition unit that acquires actual operation data related to the work content and the economic parameters, and a data assimilation unit that uses the results obtained by the economic simulation execution unit and the actual operation data before the actual operation data acquisition unit acquires the actual operation data, and modifies the results by data assimilation processing.

This disclosure makes it possible to improve the accuracy of economic evaluation simulations for robotics businesses and other projects that deal with unknown environments where physical properties, boundary conditions, etc. cannot be determined accurately in advance.

FIG. 1 is a configuration diagram showing a simulation device for evaluating economic efficiency of a robot business according to a first embodiment. 2 is a diagram showing a screen of a specific example of processing in the robot simulation execution unit 111 of FIG. 1. FIG. 2 is a diagram showing a screen of a specific example of processing in the economic parameter input unit 112 of FIG. 1. 2 is a graph showing a specific example of an evaluation result by the economic evaluation result display unit 114 of FIG. 1 . 2 is a graph for explaining an ensemble Kalman filter, which is an example of processing in the data assimilation unit 122 in FIG. 1 . 1 is a graph showing the process of correcting simulation results through data assimilation. FIG. 1 is a configuration diagram showing a simulation system for evaluating the economic feasibility of a robot business. FIG. 11 is a process flow diagram of probability distribution evaluation by simulation according to the second embodiment. FIG. 1 is a diagram for explaining a Monte Carlo method. FIG. 13 is a diagram for explaining probability density distribution estimation. FIG. 13 is a configuration diagram showing a simulation device for evaluating economic efficiency of a robot business according to a third embodiment.

This disclosure relates to a simulation device and a simulation method for performing an economic evaluation. In particular, this disclosure relates to a simulation device and a simulation method for performing an economic evaluation of a robotics business. Note that in the following explanation, a robotics business is used as an example, but the contents of this disclosure are not limited thereto. As described below, it is extremely difficult to evaluate the economic viability of a robotics business that uses work robots, etc. in an unknown environment where physical property values, boundary conditions, etc. cannot be accurately determined in advance, and this has not been the subject of conventional simulation technology. The technology of this disclosure can also be used for businesses with such unknown conditions.

The contents of this disclosure will be explained below using examples.

Figure 1 is a diagram showing the configuration of a simulation device for economic evaluation of a robot business.

In this diagram, the robot business economic evaluation simulation device 100 includes a calculation processing unit 101 and an input/output terminal 130. The calculation processing unit 101 includes a pre-processing unit 110 and an operation processing unit 120. The pre-processing unit 110 has a robot simulation execution unit 111, an economic parameter input unit 112, an economic simulation execution unit 113, and an economic evaluation result display unit 114. The operation processing unit 120 has an actual operation data acquisition unit 121, a data assimilation unit 122, and an economic evaluation result display unit 123. The input/output terminal 130 has an input unit 131 and an output unit 132.

The actual operation data acquisition unit 121 acquires actual operation data during business operations. Here, actual operation data refers to data that includes actual measured values of state variables.

Figure 2 shows a screen showing a specific example of processing in the robot simulation execution unit 111 in Figure 1.

In Figure 2, as an example, the screen of the output section 132 (display section) of the input/output terminal 130 is shown, which is aimed at the task of removing rubble from a construction site or the like.

The robot simulation execution unit 111 (Figure 1) performs a simulation of the work of the robot 202. The robot 202 (working robot) performs the work of removing rubble 201 scattered at the work site. The simulation is performed by calculating physical conditions such as the robot's center of gravity and weight, friction at the joints, friction between the ground and the wheels, and the weight of the rubble. Items evaluated by the simulation include those related to work productivity such as work time and whether the work can be performed, energy consumption which is the operating cost of the robot, items related to maintenance such as damage and deterioration of robot parts, and the amount of materials required for the work. These items may be displayed on the screen as display items 203.

Note that while FIG. 2 shows an example of simulating the work of the robot 202, it can also be generally applied to simulating the content of work in a business with unknown conditions. In this case, the robot simulation execution unit 111 functions as a work simulation execution unit.

Figure 3 shows a screen showing a specific example of processing in the economic parameter input section 112 of Figure 1.

In the economic parameter input section 112 (Figure 1), parameters (economic parameters) required to execute an economic simulation are input. In this case, the input may be done manually by the user or the device may perform automatic optimization.

Figure 3 shows a screen for inputting prerequisites 301, items related to income 302, and items related to expenses 303, all displayed side by side.

For the prerequisites 301, parameters such as currency units, project years, operating rate, number of robots, depreciation years, interest rate, sales price, electricity cost (unit price) and so on are input. If variation in the parameters is assumed, for example, a normal distribution is assumed and the mean and standard deviation are input. Also, if yearly variation is assumed, for example, a value for each year is input.

For the income-related item 302, parameters such as sales, loan amount, and subsidy are entered. Here, for example, sales are calculated using the following formula using the productivity results calculated by the robot simulation execution unit 111.

(Sales) = (Sales price) x (Production volume per unit) x (Number of units) x (Utilization rate)
When it is assumed that the unit price or production volume will change in the future, the future values for the simulation are input. When it is assumed that there will be variation in each item, for example, a normal distribution is assumed and the average and standard deviation values are input. The processing contents when the value of the standard deviation of sales is input from the results of a robot simulation will be described in the third embodiment.

Next, on the input screen for expenditure item 303, parameter values for the initial capital investment amount, power consumption, maintenance costs, labor costs, depreciation costs, corporate tax, insurance premiums, etc. are entered. Of these, power consumption, maintenance costs, etc. are calculated using values calculated by the robot simulation execution unit 111. For example, the electricity cost is calculated using the following formula.

(Electricity bill) = (Electricity charge (unit price)) x (Electricity consumption)
If you assume that these will change, enter the values for each year in the future for which you will run the simulation. If you assume that each item will vary, for example, assume a normal distribution and enter the average and standard deviation values.

The economics simulation execution unit 113 (Figure 1) evaluates the economics of the robot business based on the above input parameters. The calculation method is, for example, to simulate a forecast of future cash flow (income, expenditure) trends using the method described in Patent Document 2. Specifically, the items shown in Figure 3 are added up for each period (for example, each fiscal year) to perform the calculation.

As an indicator of the economic viability of a business, for example, the IRR (internal rate of return) is calculated from the cash flow using the following formula (1). In other words, the IRR is calculated by solving the equation in which "r" in the following formula (1) is an unknown.

In the formula, C _n is the cash flow (profit before amortization) in the nth period, C ₀ is the initial investment amount, and r (0<r) is the discount rate.

The discount rate r that satisfies the above formula (1) is the IRR. The IRR is an index that indicates the value of an investment, and the higher the value, the higher the investment value. Note that there are various economic evaluation indices other than the IRR, such as net present value and payback period, but in this disclosure, the IRR will be used as an example.

Figure 4 is a graph showing a specific example of the evaluation results from the economic evaluation result display unit 114 in Figure 1. The horizontal axis shows the number of years, and the vertical axis shows the amount of money.

In this figure, future income 402 and expenditure 403 calculated by the economic simulation execution unit 113 are displayed. The income error bar 404 and the expenditure error bar 405 indicate the respective variations, and show, for example, the standard deviation value. This allows the user to confirm the balance of future cash flows and their accuracy. Note that while income and expenditure are displayed here, it is also possible to display, for example, only the sales portion of income, or the production volume that is the result of a robot simulation. In addition, the IRR (406) may be displayed.

Figure 5 is a graph for explaining the ensemble Kalman filter method, which is an example of processing in the data assimilation unit 122 in Figure 1. Note that in this embodiment, the ensemble Kalman filter method is explained, but as a data assimilation method, a Kalman filter method other than the ensemble Kalman filter method, a particle filter method, an adjoint method, a variation method, etc. may also be used. The Kalman filter method, and especially the ensemble Kalman filter method, has the advantage of requiring less calculation and shorter calculation time than the particle filter method, etc.

Here, data assimilation refers to a method of improving the accuracy of a simulation by correcting (assimilating) a simulation model using actual measured values. In this disclosure, after the robot business begins operation, the results of a preliminary economic simulation are corrected using actual operation data to improve the accuracy of future forecast values.

In Figure 5, the state variables 501 on the horizontal axis are variables of the simulation that performs data assimilation, such as the production volume of work performed by the robot, the amount of energy consumption, or the amount of income, expenditure, etc. The vertical axis represents the probability density.

The upper part of this figure shows the probability density distribution 502 (solid line) of the simulated predicted values and the probability density distribution 503 (dashed line) of the actual measured values. Here, for example, a normal distribution is assumed, and the standard deviation in this case is also assumed by the user. A method for estimating the variance of the simulated predicted values by simulation will be explained in Example 2. Here, for example, a normal distribution is assumed. For the probability distribution of the actual measured values, if it is known that the state variable is error-free, such as monetary amount, the standard deviation of the normal distribution can be set to an extremely small value.

The probability density distribution 504 (dotted line) of the simulation after data assimilation processing shown in the lower part of this figure was calculated using the probability density distribution 502 of the simulation's prior predicted values and the probability density distribution 503 of the actual measured values, according to Bayes' theorem expressed in the following formula (2).

In the formula, p(x|y) represents the probability density distribution 504 of the simulation after data assimilation processing, p(y|x) represents the probability density distribution 503 of the measured values, and p(x) represents the probability density distribution 502 of the prior simulation.

As described above, by using Bayesian estimation, in which the prior simulation results are the prior probability, the measurements are the likelihood, and the simulation after data assimilation processing is the posterior probability, the prior simulation results can be corrected based on the measurements.

Figure 6 is a graph showing the process of correcting simulation results through data assimilation. The horizontal axis represents time and the vertical axis represents state variables.

In this figure, the changes in state variables over time are predicted based on the results of a prior simulation. After that, two data assimilation processes are performed using the actual measured values of the state variables, and the prediction of the changes in state variables over time is revised. Note that all values represented by the curves and points in the graphs in the figure contain variation (probability density distribution), but for simplicity's sake this is omitted here, and representative values such as the average and median obtained from each probability density distribution are shown.

Curve 601 is the change over time in the state variables based on the results of a prior simulation, with the portion before the first data assimilation 602 shown as a solid line and the portion after the first data assimilation 602 shown as a dashed line.

When an actual measurement value 604 is obtained at the time of the first data assimilation 602, the predicted value 603 by prior simulation shown by curve 601 is corrected to a value 605 after data assimilation by the first data assimilation 602. Here, the reason why the value 605 after data assimilation differs from the actual measurement value 604 is because Bayesian estimation was performed under the assumption that there is a variance (probability density distribution) as shown in Figure 5.

Furthermore, by re-running the simulation using the post-data assimilation values 605, the simulation results from the first data assimilation 602 onwards are also corrected from curve 601 to curve 606. The portion before the second data assimilation 607 is shown by a solid line, and the portion after the second data assimilation 607 is shown by a dashed line.

Next, using the actual measured values 608 obtained at the time of the second data assimilation 607, the predicted values 609 by the simulation are corrected to values 610 after data assimilation by the second data assimilation 607. Furthermore, by re-executing the simulation using the corrected values, the simulation results after the second data assimilation 607 are corrected from curve 606 to curve 611.

By repeating the above process whenever actual measurement data is obtained, the simulation can be corrected to get closer to the actual measurement value. As mentioned above, if it is considered that there is no error in the actual measurement data, the value after data assimilation will match the actual measurement value by setting the variance of the actual measurement data to a very small value.

The economic evaluation result display unit 123 displays the simulation results after the data assimilation process. The display format is the same as the display method shown in FIG. 4. However, in this case, since operation has already started, the actual measured values and values obtained by simulation up to the present time, as well as the values after the data assimilation process, may be displayed separately for income, expenditure, and IRR, so that the actual measured values and values obtained by simulation up to the present time as well as the values after the data assimilation process can be seen. Furthermore, if the state variables assimilated with data are final economic evaluation parameters such as income and expenditure, the results of data assimilation can be displayed directly, but if they are parameters such as production volume and energy consumption, the final economic evaluation parameters can be obtained by re-executing the economic simulation execution unit 113 using the values after the data assimilation process.

Figure 7 is a configuration diagram showing a simulation system for evaluating the economic feasibility of a robotics business.

As shown in this figure, the economic evaluation simulation system for the robotics business is configured by connecting a computer 700 corresponding to the economic evaluation simulation device 100 for the robotics business (Figure 1) to an input/output device 716.

The computer 700 has a CPU (701), a RAM (702), a ROM (703), a HDD (704), a communication I/F (705), an input/output I/F (706), and a media I/F (707). Here, CPU is an abbreviation for Central Processing Unit, RAM is an abbreviation for Random Access Memory, ROM is an abbreviation for Read Only Memory, HDD is an abbreviation for Hard Disk Drive, and I/F is an abbreviation for Interface.

The communication I/F (705) is connected to an external communication device 715. The input/output I/F (706) is connected to an input/output device 716. The media I/F (707) reads and writes data from a recording medium 717. Furthermore, the CPU (701) controls each processing unit by executing a program (also called an "application" or its abbreviation "app") loaded into the RAM (702). This program can be provided via a communication line or distributed by recording it on a recording medium 717 such as a CD-ROM.

In the second embodiment, a method for estimating the variance of a pre-predicted value obtained by simulation is described. Note that a description of the contents common to the first embodiment is omitted. When estimating the variance of a pre-predicted value obtained by simulation, the parameters of the analysis conditions for the robot simulation execution unit 111 (FIG. 1) are varied, and multiple simulations are performed using random numbers generated by the Monte Carlo method.

Figure 8 is a flow diagram showing the process of evaluating probability distributions through simulation.

In this diagram, first, the probability distribution of the input parameters, which are the analysis conditions of the simulation, is input (step S801). For example, the expected degree of variation in parameters such as the weight and shape of the object grasped by the robot is input. Next, probability distribution estimation (step S802) is performed. Here, random numbers are generated using the Monte Carlo method (step S803) for the probability distribution input in step S801. Next, a robot simulation is performed the number of times corresponding to the generated random numbers (step S804), and the obtained analysis results are used to estimate the probability distribution (step S805). Finally, the estimated probability distribution is output (step S806). The output probability distribution is used as the probability density distribution 502 (Figure 5) of the prior predicted value by the simulation during data assimilation.

Figure 9 is a diagram to explain the Monte Carlo method.

In this figure, the left side shows a graph of a curve 901 indicating the probability density for the input parameter. Meanwhile, the right side shows a table of pairs of input parameters 902 and output parameters 903 obtained by simulation. Here, the weight of the grasped object is shown as an example of an input parameter.

Specifically, random numbers are generated a specified number of times (e.g., 1000 times) according to the probability density input in the probability distribution input parameter input (step S801). In other words, random numbers are generated so that the generation density increases as the value on the vertical axis of curve 901 increases.

Next, a robot simulation is executed the number of times corresponding to the generated random numbers, and pairs of input parameters 902 and output parameters 903 are recorded. The output parameters 903 are items to be evaluated in the robot simulation, such as the working time and energy consumption. In this embodiment, the input parameters 902 and the output parameters 903 are each described as one case, but one or both may be multiple. In this embodiment, the simulation is executed for all the specified number of times in the Monte Carlo method, but if the simulation execution time is long, it is also possible to execute the simulation a small number of times in advance, store the results in a database, and estimate the value of the output parameter 903 from the value of the input parameter 902 using machine learning or the like.

Next, we explain probability density distribution estimation.

Figure 10 is a diagram for explaining probability density distribution estimation.

In this figure, graph 1001 on the left side shows the output parameters 903 (Fig. 9) of the robot simulation in the form of a histogram. On the right side, a smooth probability density distribution curve 1002 is shown that is estimated by probability density distribution estimation from the histogram of graph 1001. For example, a method such as KDE (kernel density estimation) is used here. The probability density distribution curve 1002 obtained here is output in step S806 (Fig. 8) and used as the probability density distribution 502 (Fig. 5) of the advance predicted value by the simulation during data assimilation.

Figure 11 is a configuration diagram showing a simulation device for evaluating the economic feasibility of a robotics business when replanning the operation of the robotics business is carried out using simulation results whose accuracy has been improved by data assimilation processing. Note that a description of the contents common to Example 1 will be omitted.

In this diagram, an operation replanning unit 1101 has been added to the operation time processing unit 120 in FIG. 1. The operation replanning unit 1101 replans operations based on future economic evaluation based on the results of data assimilation processing obtained by the data assimilation unit 122. For example, if the simulation results in high maintenance costs for the robots, adjustments to the plan such as replacing the robots or reviewing the number and placement of robots may be considered.

In this case, replanning may be done manually by the user by looking at what is displayed on the economic evaluation result display unit 123, or the system may automatically perform optimization.

When the user does this manually, for example, he or she can change the values of the operating conditions (preconditions) on the economic parameter input screen in Figure 3, rerun the economic simulation, and then check the IRR values before and after the change.

When the system performs this automatically, the items entered in the economic parameter input unit 112 are used as design variables, and the difference between income and expenditure evaluated in the economic simulation or the IRR is used as the objective variable, and optimization is performed to maximize or minimize the objective variable. Optimization methods used in this case include the gradient method and GA (genetic algorithm).

The following summarizes preferred embodiments of this disclosure.

It is preferable that the economic evaluation simulation device further has an economic evaluation result display unit that displays at least one of the results obtained by the economic simulation execution unit and the corrected results, which are data corrected by the data assimilation process.

It is desirable for the work simulation execution unit to evaluate at least one of the following: work productivity, energy consumption, maintenance-related items, and the amount of materials required for the work.

It is desirable that the economic evaluation simulation device further includes an operation re-planning unit, the data assimilation unit performs a future economic evaluation, and the operation re-planning unit re-plans operations based on the future economic evaluation.

The data assimilation unit preferably uses the Kalman filter method, particle filter method, adjoint method or variational method.

It is desirable that the economic parameters include at least one of the following: sales, borrowing amount, subsidy, initial investment amount, labor cost, insurance premium, electricity cost, corporate tax, maintenance cost, depreciation cost, and repair and management cost.

It is desirable that the work simulation execution unit processes the input data for the simulation performed before the start of actual operation by introducing variability using the Monte Carlo method, and that the data assimilation unit uses input data with variability.

Note that the present disclosure is not limited to the above-described examples, but includes various modified examples. For example, the above-described examples are described in detail to clearly explain the present disclosure, and are not necessarily limited to those having all of the configurations described.

It is also possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment.

In addition, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations. Furthermore, some or all of the above configurations, functions, processing units, processing means, etc. may be realized in hardware, for example by designing them as integrated circuits.

In addition, each of the above configurations, functions, etc. may be realized in software by a processor interpreting and executing a program that realizes each function.

Information such as programs, tables, and files that realize each function can be stored in a memory, a hard disk, a recording device such as an SSD (Solid State Drive), or a recording medium such as an IC card (IC here is an abbreviation for Integrated Circuit), an SD card, or a DVD (Digital Versatile Disc). The cloud can also be utilized.

In addition, the control lines and information lines shown are those that are considered necessary for the explanation, and do not necessarily show all control lines and information lines in the product. In reality, it can be assumed that almost all components are interconnected.

Furthermore, the communication means connecting each device is not limited to wireless LAN, but may be changed to wired LAN or other communication means.

100: Robot business economic evaluation simulation device, 101: Calculation processing unit, 110: Pre-processing unit, 111: Robot simulation execution unit, 112: Economic parameter input unit, 113: Economic simulation execution unit, 114, 123: Economic evaluation result display unit, 120: Operation processing unit, 121: Actual operation data acquisition unit, 122: Data assimilation unit, 130: Input/output terminal, 131: Input unit, 132: Output unit.

Claims (7)

a work simulation execution unit that performs a simulation of the contents of work in a business having unknown conditions ;
an economic parameter input unit for inputting economic parameters of the business ;
an economics simulation execution unit that performs a simulation of the economics of the business using a first result of the evaluation by the work simulation execution unit and the economics parameters, and evaluates the economics of the business;
an actual operation data acquisition unit that acquires actual operation data relating to the work content and the economic efficiency parameters;
a data assimilation unit that corrects the second result by a data assimilation process using a second result obtained by the economic simulation execution unit and the actual operation data before the actual operation data acquisition unit acquires the actual operation data,
the work simulation execution unit calculates physical conditions to evaluate at least one of work productivity, energy consumption, items related to maintenance, and amounts of materials required for the work;
The economic parameters include at least one of prerequisites, items related to income, and items related to expenditures . 2. The economic evaluation simulation device according to claim 1, further comprising an economic evaluation result display unit that displays at least one of the second result obtained by the economic simulation execution unit and a correction result which is data corrected by the data assimilation process. An operation replanning unit is further provided,
The data assimilation unit performs a future economic evaluation,
2. The economic evaluation simulation device according to claim 1, wherein the operation re-planning unit re-plans the operation of the business based on the future economic evaluation. The economic evaluation simulation device according to claim 1, wherein the data assimilation unit uses a Kalman filter method, a particle filter method, an adjoint method, or a variational method. The economic evaluation simulation device according to claim 1, wherein the economic parameters include at least one of sales, borrowing amount, subsidy, initial investment amount, labor cost, insurance premium, electricity cost, corporate tax, maintenance cost, depreciation cost, and repair management cost. The operation simulation execution unit performs a process of giving variability to input data of a simulation performed before starting actual operation using a Monte Carlo method,
The economic evaluation simulation device according to claim 1 , wherein the data assimilation unit uses the input data having the variability. A method executed by an economic evaluation simulation device having an operation simulation execution unit, an economic parameter input unit, an economic simulation execution unit, an actual operation data acquisition unit, and a data assimilation unit, comprising:
The work simulation execution unit performs a simulation of the contents of work in a business having unknown conditions ,
The economic parameter input unit inputs economic parameters of the business ,
the economic efficiency simulation execution unit performs a simulation of the economic efficiency of the business using a first result of the evaluation by the work simulation execution unit and the economic efficiency parameters, and evaluates the economic efficiency of the business;
the actual operation data acquisition unit acquires actual operation data relating to the content of the work and the economic efficiency parameters;
the data assimilation unit corrects the second result by a data assimilation process using a second result obtained by the economic simulation execution unit before the actual operation data acquisition unit acquires the actual operation data and the actual operation data ;
the work simulation execution unit calculates physical conditions to evaluate at least one of work productivity, energy consumption, items related to maintenance, and amounts of materials required for the work;
The economic evaluation simulation method, wherein the economic parameters include at least one of prerequisites, items related to income, and items related to expenditures .