Information processing device and simulation method

Abstract

The present invention provides an information processing device and simulation method for obtaining each element that influences the target evaluation indicator, improvement measures, and evaluation results before and after improvement. [Solution] An information processing device for performing a simulation of predicted values, comprising: a model generation unit 6 that generates a prediction model of an evaluation index for one or more devices based on device data obtained from a sensor 3 and a predetermined model formula; a simulation unit 7 that predicts the evaluation index using the prediction model; and an output unit 8 that outputs the prediction model and / or the predicted evaluation index. The simulation unit 7 calculates at least one improvement measure for the evaluation index by interpreting physical phenomena from the predetermined model formula, calculates a first prediction value which is the predicted value of the evaluation index when the calculated improvement measure is applied, and a second prediction value which is the predicted value of the evaluation index when the calculated improvement measure is not applied, and the output unit 8 outputs the improvement measure, the first prediction value and the second prediction value.

Description

Technical Field

[0001] The present invention relates to an information processing apparatus and a simulation method.

Background Art

[0002] In recent years, in electronic devices, mechanical devices, mechanical systems, etc., it has often been carried out to construct a simulation model and utilize it for design, control, operation, maintenance, etc. For example, when predicting and controlling the room temperature and power consumption in a house, an office building, a factory, etc., a model is constructed with inputs such as the structure and dimensions of the building, materials, types, numbers, and arrangements of lighting and electrical products, the performance and operation methods of air conditioners, the outside air temperature, and solar radiation conditions, and simulation is performed. As a result of the simulation, air conditioners and electrical equipment are appropriately controlled for the purpose of reducing electricity bills and improving comfort. At that time, for example, by performing three-dimensional unsteady thermal fluid analysis as a simulation, it becomes possible to predict the temporal changes in room temperature and power consumption.

[0003] However, generally, three-dimensional unsteady thermal fluid analysis has a large computational load, so a more simplified low-dimensional model is often used. The low-dimensional model may be configured deductively from a physical model that reproduces the behavior of each component, or may be configured from a model constructed inductively using sensor data installed in a device or system, or may be configured by mixing those models. Generally, when using a low-dimensional model, although the accuracy and resolution are lower compared to the case of using a high-dimensional model, it is often possible to perform a wider or longer-term simulation. For example, in the case of predicting the room temperature and power consumption described above, while a high-dimensional model could simulate for several hours in one room, a low-dimensional model may be able to predict several days or months later for the entire building or area.

[0004] Patent Document 1 describes an information processing device, an information processing method, and a program that can more easily identify the influence of multiple variables on the output of a model obtained by regression analysis using statistics and machine learning.

[0005] Furthermore, Non-Patent Document 1 proposes a sparse identification method for nonlinear dynamic systems as a modeling technique using data. [Prior art documents] [Patent Documents]

[0006] [Patent Document 1] Japanese Patent Publication No. 2023-131558 [Non-patent literature]

[0007] [Non-Patent Document 1] SL Brunton, JL Proctor, JN Kutz, Discovering governing equations from data by sparse identification of nonlinear dynamical systems, PNAS, Vol.113 (2016), pp.3932-3937. [Overview of the Initiative] [Problems that the invention aims to solve]

[0008] Patent Document 1 calculates the degree of influence and the selected frequency for each explanatory variable in the constructed regression model over a period selected as a variable affecting the output data, and displays the calculated degree of influence and the selected frequency in relation to each other. Users can extract and identify appropriate factors without omission, referring to the characteristics of each category. In other words, the influence of multiple variables on the output can be identified more easily.

[0009] However, conventional information processing devices, including Patent Document 1, have the problem that the physical governing factors and improvement measures for each element influencing the target evaluation index (for example, each term when the evaluation index is expressed as a polynomial), as well as the evaluation results before and after improvement, are unclear. For example, if the electricity cost of a base consisting of multiple areas is used as the target evaluation index, and the electricity cost due to air conditioning in some areas is one of the physical governing factors for the base's electricity cost, it has not always been possible to accurately present evaluations and improvement measures such as how much improvement in the electricity cost due to air conditioning in those areas would improve the overall electricity cost of the base.

[0010] This invention was made to solve the above-mentioned problems, and aims to provide a technology for obtaining the elements and improvement measures that affect the target evaluation index, as well as the evaluation results before and after the improvement. [Means for solving the problem]

[0011] To achieve the above objective, the present invention provides an information processing device for performing simulations of predicted values, comprising: a model generation unit that generates a prediction model of one or more evaluation indicators for a device based on device data obtained from a sensor and a predetermined model formula; a simulation unit that predicts the evaluation indicator using the prediction model; and an output unit that outputs the prediction model and / or the predicted evaluation indicator. The simulation unit calculates at least one improvement measure for the evaluation indicator by interpreting physical phenomena from the predetermined model formula, calculates a first prediction value which is the predicted value of the evaluation indicator when the calculated improvement measure is applied, and a second prediction value which is the predicted value of the evaluation indicator when the calculated improvement measure is not applied, and the output unit outputs the improvement measure, the first prediction value, and the second prediction value. [Effects of the Invention]

[0012] According to the present invention, for a target evaluation index, each element that affects it, improvement measures, and evaluation results before and after improvement can be obtained.

Brief Description of the Drawings

[0013] [Figure 1A] It is a diagram showing the configuration and data flow of an information processing apparatus according to a first embodiment of the present invention. [Figure 1B] It is a diagram showing an example of a computer schematic diagram. [Figure 2] It is a flowchart showing the flow from data acquisition to result output of an information processing apparatus according to a first embodiment of the present invention. [Figure 3] It is a diagram showing an input unit of an information processing apparatus according to a first embodiment of the present invention. [Figure 4] It is a diagram showing a display example of an output unit of an information processing apparatus according to a first embodiment of the present invention. [Figure 5] It is a diagram showing another display example of an output unit of an information processing apparatus according to a first embodiment of the present invention. [Figure 6] It is a diagram showing another display example of an output unit of an information processing apparatus according to a first embodiment of the present invention. [Figure 7] It is a diagram showing another display example of an output unit of an information processing apparatus according to a first embodiment of the present invention. [Figure 8] It is a diagram showing the configuration and data flow of an information processing apparatus according to a second embodiment of the present invention. [Figure 9] It is a diagram showing the configuration and data flow of an information processing apparatus according to a third embodiment of the present invention. [Figure 10] It is a flowchart showing the flow from data acquisition to result output of an information processing apparatus according to a third embodiment of the present invention.

Modes for Carrying Out the Invention

[0014] Hereinafter, embodiments of the present invention will be described with reference to the drawings. The examples are illustrative for explaining the present invention, and for the sake of clarity of explanation, appropriate omissions and simplifications have been made. The present invention can also be implemented in various other forms. Unless otherwise limited, each component may be singular or plural. In the drawings, the positions, sizes, shapes, ranges, etc. of each component shown may not represent the actual positions, sizes, shapes, ranges, etc. in order to facilitate the understanding of the invention. Therefore, the present invention is not necessarily limited to the positions, sizes, shapes, ranges, etc. disclosed in the drawings.

[0015] When there are a plurality of components having the same or similar functions, they may be described with the same reference numeral and different subscripts. Also, when it is not necessary to distinguish these plurality of components, the subscripts may be omitted in the description.

[0016] In the examples, the processes performed by executing a program may be described. Here, the computer executes a program by a processor (e.g., CPU, GPU), and performs the processes defined by the program while using storage resources (e.g., memory) and interface devices (e.g., communication ports), etc. Therefore, the subject of the process performed by executing the program may be the processor. Similarly, the subject of the process performed by executing the program may be a controller, device, system, computer, or node having a processor. The subject of the process performed by executing the program only needs to be an arithmetic unit and may include a dedicated circuit for performing a specific process. Here, the dedicated circuit is, for example, an FPGA (Field Programmable Gate Array), an ASIC (Application Specific Integrated Circuit), a CPLD (Complex Programmable Logic Device), etc.

[0017] The program may be installed on the computer from the program source. The program source may be, for example, a program distribution server or a storage medium readable by the computer. If the program source is a program distribution server, the program distribution server includes a processor and storage resources for storing the program to be distributed, and the processor of the program distribution server may distribute the program to other computers. In addition, in the embodiment, two or more programs may be implemented as one program, or one program may be implemented as two or more programs.

[0018] The following examples will be described with reference to the drawings. [Examples]

[0019] An information processing device according to the first embodiment of the present invention will be described below with reference to Figures 1A to 10.

[0020] Figure 1A shows the configuration and data flow of the information processing device 100 according to the first embodiment. In this embodiment, indoor temperature control using air conditioners etc. is described as an example, but the application of the present invention is not limited to this, and can be applied to, for example, the design of electronic and medical equipment, power demand control using storage batteries, prediction of power generation amount from solar power and storage / discharge management, energy management in factories, adjustment capacity management using distributed energy sources, response to demand response, participation in adjustment capacity markets and power markets, etc. It can also be applied to cyber-physical systems where the use of 1D simulation is envisioned, operational optimization, real-time control, and other applications.

[0021] The information processing device 100 according to this embodiment includes a data acquisition unit 2 that acquires data from sensors 3 installed on a target system 1 to be simulated, a storage unit 4 that stores the data, an input unit 5 that inputs data to the storage unit 4, a model generation unit 6 that generates a prediction model using the data acquired from the sensors 3 and the data input from the input unit 5, a simulation unit 7 that performs a simulation using the generated prediction model, and an output unit 8 that outputs the simulation results. The data from the sensors 3 acquired by the data acquisition unit 2 includes, for example, the intake temperature of the air conditioner, the room temperature, the outside air temperature, the set temperature, the compressor rotation speed, and the fan rotation speed. The storage unit 4 can also acquire data from an external system 10. The input unit 5 takes, for example, an evaluation index, a sensor number, a variable name, and a variable type as input. The external system 10 includes, for example, the date, time, prediction results for outside temperature and humidity, forecast results for outside temperature and humidity obtained via the internet, and prediction results from an external program that interpolates data obtained from the data acquisition unit 2.

[0022] The information processing device 100 shown in Figure 1A can be realized by a general-purpose computer 1600, which includes, for example, a CPU 1601, memory 1602, an external storage device 1603 such as an HDD (Hard Disk Drive), a reader 1607 for reading and writing information to a portable storage medium 1608 such as a CD (Compact Disk) or USB memory, an input device 1606 for receiving various types of information such as a keyboard and mouse, an output device 1605 such as a display for outputting various types of information that are input and used for processing, a communication device 1604 such as a NIC (Network Interface Card) for connecting to a communication network, and an internal communication line (called a system bus) 1609 such as a system bus that connects these.

[0023] Furthermore, various data stored in the information processing device 100 or used for processing can be realized by the CPU 1601 reading and using data from memory 1602 or external storage device 1603. In addition, each part of the information processing device 100 (for example, data acquisition unit 2, model generation unit 6, simulation unit 7) can be realized by the CPU 1601 loading a predetermined program stored in external storage device 1603 into memory 1602 and executing it.

[0024] The aforementioned programs and data may be stored (downloaded) from the storage medium 1608 via the reading device 1607, or from the network via the communication device 1604, into the external storage device 1603, and then loaded onto the memory 1602 and executed by the CPU 1601. Alternatively, they may be loaded directly onto the memory 1602 via the reading device 1607 from the storage medium 1608, or from the network via the communication device 1604, and then executed by the CPU 1601.

[0025] In the following example, the information processing device 100 is described as being composed of a single computer, but all or part of these functions may be distributed across one or more computers, such as a cloud, and similar functions may be realized by them communicating with each other via a network.

[0026] In Figure 1B, the information processing device 100 is shown as an example in which it has an input device 1606, which is hardware that functions as an input unit 5. However, it is not necessary for the information processing device 100 to have this input device, and it may be located outside of the information processing device 100. Similarly, the output device 1605, which is hardware that functions as an output unit 8, may also be located outside of the information processing device 100.

[0027] Figure 2 is a flowchart showing the flow from data acquisition to result output of the information processing device according to the first embodiment.

[0028] In step S10, setting information is obtained from the input unit 5, which functions as an input device 1606. Here, Figure 3 is a diagram showing an example of information that the input unit 5 of the information processing device according to the first embodiment accepts as input. In the input unit 5 of this embodiment, the electricity rate of a certain location (location A) is set as an example of an evaluation indicator to be improved. Here, a location may be, for example, a city block, a factory, a commercial building, a hospital, a school, etc., or even a single floor or room.

[0029] Furthermore, the evaluation indicators in this embodiment are not limited to electricity charges, but may also include physical indicators such as temperature, pressure, and energy; economic indicators such as electricity charges, operating costs, and revenue; and environmental indicators such as CO2 emissions, greenhouse gas emissions, and renewable energy usage. Additionally, there is no need to use only one evaluation indicator; there may be multiple. If there are multiple evaluation indicators, optimization techniques or the like can be used to adjust them to obtain favorable results for the user of this information processing device.

[0030] Furthermore, the data accepted by the input unit 5 includes the sensor number, variable name, and variable type, as shown in Figure 3. The sensor number links each variable to the sensor installed in the target system. The variable name is a name used to identify the variable, and the variable type is also entered. The variable type is used when linking to physical phenomena, which will be described later. In Figure 3, for example, it is shown that processing is performed using data obtained from sensors 1 to M in order to determine the evaluation index "electricity cost at site A". Also, the data obtained from the sensor identified by sensor number "sensor 1" is defined as a variable represented by the variable name "set temperature of area A1", and this variable is defined as a variable related to "temperature".

[0031] In step S11, the model generation unit 6 determines the evaluation indicators for this information processing device based on the information input in step S10. For example, "electricity charges for site A" shown in Figure 3 is determined as the evaluation indicator.

[0032] In step S12, the model generation unit 6 sets the necessary target variable and explanatory variables from the variables input in step S10 in order to predict and evaluate the evaluation indicators. The target variable may be the same as the evaluation indicators or may be different. The target variable is a variable that this information processing device predicts in order to calculate the evaluation indicators (i.e., the electricity charges for site A in this embodiment).

[0033] Furthermore, for example, in this embodiment, the electricity cost of base A is used as an evaluation metric. However, if "electricity costs due to air conditioning in area A1" are one of the physical governing factors, this may be set as a new evaluation metric or as one of the objective variables. Alternatively, the time derivative or time integral of "electricity costs due to air conditioning in area A1" may be set as the objective variable. On the other hand, there may be other factors that influence the evaluation function besides those set as objective variables. For example, if there are electronic devices that cannot be shut down, the electricity costs resulting from them will affect the evaluation function, but since they cannot be reduced, there is no need to predict them in this information processing device, nor is there a need to set them as objective variables. Therefore, although the electricity costs of such electronic devices are not objective variables, they are calculated as part of the evaluation function. The evaluation function is a function for calculating the difference between the predicted value output by the prediction model and the actual correct value.

[0034] The dependent variable can be, for example, power consumption, comfort level, or renewable energy usage, and is not necessarily limited to one; multiple dependent variables can be set. If multiple dependent variables are set, multiple model equations are usually constructed from step S30 onward. The user may set the target values ​​for one or more dependent variables, or these may be stored in the memory unit 4 beforehand. If there are multiple dependent variables, they may be adjusted using known optimization methods. The independent variables can be, for example, the intake temperature of the air conditioner, the room temperature, the outside temperature, the set temperature, the compressor speed, and the fan speed.

[0035] In step S20, the model generation unit 6 acquires data from at least one of the data acquisition unit 2 and the external system 10.

[0036] In step S25, the model generation unit 6 determines whether or not to perform modeling. This is because the time when data acquisition is performed in step S20 and the time when modeling needs to be performed do not necessarily coincide. For example, if you want to always build a predictive model from the latest data and perform a simulation immediately when necessary, you need to perform modeling each time data is acquired. On the other hand, if you do not need to always build the latest predictive model, it may be sufficient to build the predictive model at intervals longer than the data acquisition interval. Therefore, if you do not want to perform modeling (step S25; No), you return to step S20 and acquire data. If you want to perform modeling (step S25; Yes), you proceed to step S30.

[0037] In step S30, the model generation unit 6 determines candidate terms for each term in the model equation. The model equation can be expressed, for example, as follows:

[0038]

number

[0039] Here, F is the model equation representing the dependent variable, and can be set as, for example, "electricity cost due to air conditioning in area A1" or "time derivative of electricity cost due to air conditioning in area A1". In this embodiment, the case where the dependent variable is "electricity cost due to air conditioning in area A1" will be described below. C is the model coefficient (C0 is the constant term), f is the function that constitutes each term of the prediction model, x is the explanatory variable, n is the number of explanatory variables, m is the number of terms excluding the constant term, and i is an integer. The function f may be a function form that has been stored in the memory unit 4 in advance, or it may be a function form that has been determined from the data input from the data acquisition unit 2, the input unit 5, and the external system 10. For example, it may be determined by the type of dependent variable and explanatory variable. Also, f may be expressed as a polynomial of the explanatory variable x. In this way, the user may select any form of model equation for each candidate term in the model equation.

[0040] In step S40, the model generation unit 6 calculates the values ​​of the model coefficients C of the model equation F described above. The model coefficients C may be calculated using data input from the data acquisition unit 2, the input unit 5, and the external system 10, using known equations, known statistical methods, probabilistic methods, or machine learning. Specifically, regression analysis or sparse identification methods as described in Non-Patent Literature 1 may be used.

[0041] In step 50, the model generation unit 6 determines at least one model equation using the calculated model coefficients. Here, if the calculated model coefficient C is small and its influence is deemed small, C can be considered as zero.

[0042] In step 60, the simulation unit 7 determines whether or not to perform the simulation. If prediction or evaluation is necessary by performing the simulation as described later (step S60; Yes), the process proceeds to step S70. If the simulation is not necessary and data acquisition takes priority, or if it is time for data acquisition (step S60; No), the process returns to step S20. Alternatively, if, for example, neither the simulation nor the data acquisition time is required, the process can wait in step S60.

[0043] In step S70, the simulation unit 7 executes a simulation using the constructed prediction model. The simulation may include the above prediction model as a part of it. For example, the simulation as a whole may predict the power consumption of a site or building, with the air conditioning-related portion represented by the above model equation, and other electrical equipment and demand-side energy resources represented by other mathematical models, statistical models, machine learning models, etc. Furthermore, this simulation may be a prediction model that can be evolved over time. That is, it may predict the future values ​​of evaluation indicators or target variables.

[0044] In step S80, the simulation unit 7 predicts the values ​​of each term in the model equation, i.e., C, which represents the explanatory variable in equation (1). i *f i (x1,···,x n The simulation unit 7 calculates the value of ). The simulation unit 7 also calculates the influence of each term. The influence of each term may be, for example, the value of each term, the absolute value of each term, or a ratio with the value F of the model equation as the denominator.

[0045] In other words, the degree of influence represents the magnitude of the value or proportion of each term relative to the dependent variable. The larger the value or proportion of each term, the greater its influence on the dependent variable, and the more dominant it is considered to be. For example, if the magnitude of the value or proportion of each term exceeds a certain level relative to the dependent variable, it can be determined that the phenomenon represented by that term or the explanatory variable included in that term is a dominant factor of the dependent variable, and the simulation unit 7 calculates the degree of influence for each term. As described above, the model generation unit 6 can either treat the explanatory variable that it has determined to be a dominant factor of the dependent variable as a new evaluation index, or set it as one of the new dependent variables. In that case, the simulation unit 7 can perform a simulation on the newly set dependent variable.

[0046] Next, in step S90, the simulation unit 7 performs a physical interpretation of these terms. The physical interpretation of each term in the model equation, that is, its connection to physical phenomena, may be set in advance in the memory unit 4, or it may be determined by considering the functional form of the model equation determined in step S50. For example, if a term in the model equation is a function of the difference between the outdoor temperature To and the indoor temperature Ti, i.e., (To-Ti), then that term can be considered to represent the effect of the air conditioning load. Also, for example, if the target system 1 is a general thermal system, and the heat flux q is expressed as a term of some spatial temperature gradient, then the physical phenomenon related to that term can be estimated to be heat conduction. Similarly, if it is a temperature difference, it can be linked to heat transfer, and if it is proportional to the fourth power of temperature T, it can be linked to a radiation phenomenon. In this way, it is possible to link the functional form of each term to physical phenomena.

[0047] In step S100, the simulation unit 7 considers improvement measures to improve the evaluation indicators and decides on the improvement measures to actually implement, based on the predicted values ​​of the model equation obtained in step S80 and the physical interpretation obtained in step S90.

[0048] When considering improvement measures, the simulation unit 7 may change the explanatory variables involved in the terms whose impact is relatively large as evaluated in step S80, calculate the impact of such changes on the evaluation index, and then search for conditions under which a desirable evaluation index can be obtained based on this. Alternatively, if there are multiple explanatory variables, an optimization method may be used for the search. For example, if the evaluation index is the electricity cost at site A and the objective variable is the electricity cost due to air conditioning in area A1, the simulation unit 7 may consider the impact of each term and its physical interpretation to find explanatory variables that can reduce the objective variable, as well as their values ​​and ranges (consideration of improvement measures).

[0049] For example, the simulation unit 7 considers improvement measures such as changing the set temperature of the air conditioning (heating) in area A1, which is one of the explanatory variables, within a certain time range, and lowering the set temperature by 0.5°C or more, or turning off the air conditioning for 30 minutes or more. More specifically, the simulation unit 7 calculates the objective variable when the set temperature, which is one of the explanatory variables, is lowered, and calculates the electricity bill, which is the evaluation index. Conversely, it also determines the set temperature that should be set from the value of the electricity bill that it wants to reduce. In this case, if there are multiple explanatory variables, optimization methods may be used. From among these improvement measures, an actual improvement measure is determined, such as "lowering the set temperature of the air conditioning in area A1 by 1.0°C," so that the comfort level, calculated from room temperature, humidity, etc., does not deteriorate beyond a predetermined acceptable range.

[0050] Next, in step S110, the simulation unit 7 sorts the results, such as the magnitude of the influence of each term, and in step S120, outputs the simulation results to the output unit 8 as described later. In the sorting process described above, for example, the simulation results are rearranged in descending order of the magnitude of the influence of each term.

[0051] In step S130, the simulation unit 7 determines whether or not to reacquire the setting information. For example, if the evaluation indicator changes or the setting information changes (step S130; Yes), the simulation unit 7 returns to step S10 and reacquires the setting information from the input unit 5. An example of when a change in the evaluation function may be necessary is when it is determined that economic indicators such as electricity costs have been sufficiently met at a certain time, in which case the priority of environmental indicators such as CO2 emissions increases, and the evaluation indicators may be changed. An example of when the setting information changes is when a sensor malfunctions and is shut down or replaced, or when a sensor is added.

[0052] In step S140, the simulation unit 7 determines whether to continue acquiring data. If it decides to acquire data (step S140; Yes), it returns to step S20. On the other hand, if further acquisition is not necessary (step S140; No), this flow terminates.

[0053] Figure 4 shows an example of information including simulation results output by the output unit of the information processing device according to the first embodiment. The information output by the output unit 8 includes an improvement measure display unit 801 that displays improvement measures for evaluation indicators, improvement effect display units 802-1 and 802-2 that display improvement effects, and an improvement measure candidate display unit 803 that displays candidate improvement measures. In this example, the output unit 8, which functions as an output device 1605, displays each piece of information on the screen.

[0054] The improvement measure display unit 801 is an area that displays the improvement measures determined in step S100. In this example, in step S100, the simulation unit 7 indicates that the improvement measure is "a decrease of ○°C in the set temperature of area A1," which is the phenomenon, its value and range, or the explanatory variable constituting the said term, that has the greatest influence on each term constituting the model equation for calculating the evaluation index "electricity charges for site A."

[0055] The improvement effect display unit 802-1 displays the time change of the predicted values ​​of the evaluation indicator before and after the application of the improvement measures. The evaluation indicator before the application of the improvement measures (electricity charges for site A) includes one or more target variables F calculated using the predicted values ​​in step S80 of Figure 2. More specifically, in this embodiment, it includes electricity charges for air conditioning in area A1, electricity charges for lighting, electricity charges for other electronic equipment, etc. In this example, the simulation unit 7 displays the predicted values ​​of the model equation obtained in step S80 and the predicted values ​​of the model equation representing the determined improvement measures displayed in the improvement measure display unit 801 obtained in step S100 in a comparable manner.

[0056] The improvement effect display unit 802-2 shows the improvement results of the evaluation indicator at a specific time. For example, the difference 401 showing the improvement effect at a certain time Tn of the model equation displayed in the improvement effect display unit 802-1 is displayed as a reduction rate of "X%".

[0057] Furthermore, the improvement measure candidate display unit 803 displays at least one improvement measure candidate, and the improvement measure displayed in the improvement measure display unit 801, the improvement effect display unit 802-1, and 802-2 can be selected. The improvement measures shown here are those considered in step S100, and the effect of the selected improvement measure is displayed in the improvement effect display unit 802-1. In the example shown in Figure 4, the simulation unit 7 calculates the objective variable F (electricity cost due to air conditioning in area A1) when the explanatory variable x, the set temperature, is lowered by 1 degree, and the evaluation index (electricity cost at base A) including this F is displayed as after the improvement measure has been applied. Also, if multiple improvement measures are selected in the improvement measure candidate display unit 803, their effects are reflected and displayed in the improvement measure display unit 801, the improvement effect display unit 802-1, and 802-2.

[0058] Furthermore, Figure 5 shows a different display example from Figure 4 regarding the information, including simulation results, output by the output unit of the information processing device according to the first embodiment. The information output by the output unit 8 shown in Figure 5 includes the improvement effect display unit 802 and the improvement measure candidate display unit 803, as well as the influence degree display unit 804-1 which displays the influence degree of each term in the model equation, and the influence degree display unit 804-2 which displays the influence degree by physical factor. In Figure 5, the improvement measure candidate display unit 803 displays a difference 401 instead of a graph.

[0059] The influence display section 804-1 illustrates the magnitude of the influence of each term in the model equation. The magnitude of the influence of each term in the model equation may be an instantaneous value at a certain time, or an average value over a certain period. Furthermore, as mentioned above, the influence may be the value or absolute value of each term calculated from the model equation, or it may be displayed as a percentage with the value F of the model equation as the denominator. In addition, the values ​​may be displayed in descending order, or in any order.

[0060] Furthermore, the impact display unit 804-2 displays the degree of impact for each physical factor. This displays the physical phenomena shown in each term of the model equation used by the simulation unit 7 in step S90 shown in Figure 2, categorized by physical factor.

[0061] When the simulation unit 7 displays information on the impact display unit 804-2, it may display each item individually or display a sum of multiple items. When summing by physical factors, the items used and the level of detail in the classification can be suitably adjusted in consideration of the purpose of use of this information processing device.

[0062] Figure 6 shows a different display example from Figures 4 and 5 regarding the information, including simulation results, output by the output unit of the information processing device according to the first embodiment. The information output by the output unit 8 includes the improvement effect display unit 802-1, the improvement measure candidate display unit 803, and the impact level display unit 804-1.

[0063] The impact display unit 804-1 displays the impact calculated from the magnitude of each term in the model equation at time t in the improvement effect display unit 802-1. The improvement effect display unit 802-1 accepts user input and allows the user to select a time by moving a dotted line indicating time t on the screen. As shown in Figure 6, the improvement effect display unit 802-1, the impact display unit 804-1, and the improvement measure candidate display unit 803 are all displayed on a single screen, allowing the user to grasp the effects of the improvement measures and the physical factors influencing each term in the model equation at a glance.

[0064] Figure 7 shows an example of displaying information, including simulation results, output by the output unit of the information processing device according to the first embodiment, which differs from the display examples shown in Figures 4 to 6. The information output by the output unit 8 includes an improvement effect display unit 802-1, an improvement measure candidate display unit 803, and an impact degree display unit 804-1.

[0065] The improvement effect display unit 802-1 displays the degree of influence of each term in the model equation on the evaluation indicator. In this example, the degree of influence of one to three items from each term in the model equation is displayed.

[0066] Furthermore, the impact display unit 804-1 shows the time change from time t to time t+T, allowing users to see how the impact of each item changes over time. This makes it possible to visualize how the improvement measures selected by the improvement measure candidate display unit 803 affect each item, and how that further affects the evaluation indicators.

[0067] As described above, according to this embodiment, in simulations performed under various conditions by an information processing device that constructs a model using data obtained from sensors, etc., the degree of influence of each element, the physical governing factors, improvement measures, and quantitative evaluation results before and after improvement can be output based on physical evidence. In other words, for the target evaluation indicator, the physical governing factors and improvement measures for each element that influences it, as well as evaluation results before and after improvement, can be obtained.

[0068] Figure 8 shows the configuration and data flow of an information processing device 200 according to a second embodiment of the present invention. The difference between the information processing device according to the second embodiment and the information processing device 100 according to the first embodiment is the control unit 9. Therefore, here, the same reference numerals are used for parts common to the information processing device 100 according to the first embodiment, and redundant explanations are omitted. The control unit 9 receives the results of the simulation unit 7 and applies the above-described improvement measures to suitably control the target system 1 in order to set the evaluation index to a target value. For example, when power consumption is used as the evaluation index, in order to reduce the evaluation index, the control unit 9 changes the set temperature of the target system, i.e., the air conditioner, to a predicted value after applying the above-described improvement measures. The functions of the control unit 9 can be realized, for example, by the CPU 1601 shown in Figure 1B.

[0069] Figure 9 shows the configuration and data flow of the information processing device 300 according to the third embodiment. The difference between the information processing device according to the third embodiment and the information processing device 200 according to the second embodiment is the coefficient calculation unit 11. Therefore, here, the same reference numerals are used for parts common to the information processing device 200 according to the second embodiment, and redundant explanations are omitted. The function of the coefficient calculation unit 11 can be realized by, for example, the CPU 1601 shown in Figure 1B, similar to the control unit 9. In this embodiment, the model coefficients are updated in the coefficient calculation unit 11 without performing prediction model generation in the model generation unit 6.

[0070] Figure 10 is a flowchart showing the flow from data acquisition to result output of the information processing device according to this embodiment.

[0071] In step S25, if the model generation unit 6 determines that modeling should not be performed (step S25; No), the coefficient calculation unit 11 determines in step S26 whether or not to update the model coefficients. If the coefficient calculation unit 11 determines that the model coefficients should be updated (step S26; Yes), the update is performed in step S27. In this case, the model coefficients are updated without updating the model equation. The model coefficients may be updated using the same method as model generation, or they may be updated using data assimilation. This allows, for example, the shape of the function shown in equation (1) to remain unchanged, while the difference between the predicted values ​​of each term obtained from the model equation of the prediction model and the actual data obtained from sensors, etc., is analyzed by data assimilation. Only the model coefficients are modified so that the predicted values ​​approach the actual data, thereby improving the prediction accuracy for the acquired data. In other words, the coefficient calculation unit 11 modifies the model coefficients so that the predicted values ​​included in the simulation results by the simulation unit 7 approach the actual data obtained from sensors. This makes it possible to improve the accuracy of the simulation without changing the model equation of the prediction model.

[0072] As described above for each embodiment, in an information processing device that performs simulation of predicted values ​​(for example, information processing devices 100, 200, 300), a model generation unit (for example, model generation unit 6) generates a prediction model for one or more evaluation indicators of the above device based on device data obtained from a sensor (for example, data output from sensor 3) and a predetermined model equation (for example, the model equation shown in (Equation 1)), a simulation unit (for example, simulation unit 7) predicts the above evaluation indicator using the above prediction model, and outputs the above prediction model and / or the above predicted evaluation indicator. The simulation unit has an output unit (for example, output unit 8) and calculates at least one improvement measure for the evaluation index by interpreting physical phenomena from the predetermined model formula, and calculates a first predicted value which is the predicted value of the evaluation index when the calculated improvement measure is applied (for example, the graph after applying the improvement measure shown in Figure 4), and a second predicted value which is the predicted value of the evaluation index when the calculated improvement measure is not applied (for example, the graph before applying the improvement measure shown in Figure 4). The output unit outputs the improvement measure, the first predicted value, and the second predicted value (for example, the information of the improvement measure display unit 801 and the improvement effect display unit 802 shown in Figure 4). As a result, it is possible to obtain each element that affects the target evaluation index (for example, the electricity bill at base A) (the set temperature of area A1 at base A), the improvement measure (for example, the information shown in the improvement measure display unit 801 shown in Figure 4), and the evaluation results before and after the improvement (for example, the information shown in the improvement effect display unit 802-1 shown in Figure 4).

[0073] Furthermore, as explained using step S100 in Figure 2 and Figure 4, the output unit outputs at least one of the improvement measures calculated by the simulation unit, and outputs the first predicted value when the selected improvement measure is applied. This allows the user to obtain simulation results for various improvement measures.

[0074] Furthermore, as explained using Figures 5 and 6, the simulation unit identifies at least one governing factor (for example, "set temperature of area A1") which is a physical factor affecting the evaluation index in the predetermined model equation that constitutes the prediction model. The simulation unit then calculates the degree of influence of the governing factor on the evaluation index in the prediction model (for example, the information displayed in the influence display unit 804-2 shown in Figure 5, the degree of influence A%) when the governing factor is the set temperature. The output unit then outputs the calculated governing factor and its degree of influence. This allows the user to grasp the degree of influence of the governing factor on the evaluation index at a glance.

[0075] Furthermore, as explained using Figures 5 and 6, the above-mentioned predetermined model equation is expressed as a mathematical formula consisting of at least one term (for example, the polynomial shown in (Equation 1)), and the simulation unit identifies the governing factors and calculates the degree of influence of the governing factors in each term of the mathematical formula that represents the above-mentioned model equation that constitutes the prediction model. This makes it possible to calculate the degree of influence for each governing factor and to easily adjust the degree of influence.

[0076] Furthermore, as explained using Figures 5 and 6, the above prediction model may use an equation model, a statistical model, or a stochastic model. By using any of these models, simulation and evaluation results corresponding to each model can be obtained.

[0077] Furthermore, as explained using Figures 5 and 6, the model generation unit generates the prediction model using either sparse identification or regression analysis. Therefore, it can generate a prediction model corresponding to either of these methods.

[0078] Furthermore, as explained using step S10 in Figure 2, and Figures 5 and 6, the above evaluation indicators may be physical indicators, economic indicators, or environmental indicators. These indicators can be defined as evaluation indicators, and simulations can be performed.

[0079] Furthermore, as explained using Example 3, Figures 9 and 10, the system includes a model coefficient calculation unit (for example, a coefficient calculation unit 11) that updates the prediction model by modifying the model coefficients through data assimilation so that the first predicted value obtained by the simulation unit approaches the actual data obtained from the sensor. This allows for improving the prediction accuracy for acquired data by modifying the model coefficients so that the difference between the predicted value of each term obtained from the model equation of the prediction model through data assimilation and the actual data obtained from the sensor, etc., is eliminated, without changing the shape of the function represented by the polynomial shown in Equation 1.

[0080] Furthermore, as explained using Example 2, the control unit, Figure 8, etc., the system has a control unit (for example, control unit 9) that controls the equipment according to the improvement measures calculated by the simulation unit. This makes it possible to suitably control the target system 1 by applying the above-mentioned improvement measures to bring the evaluation index to the target value based on the simulation results.

[0081] Furthermore, as explained using Figure 2, etc., when the model generation unit generates the prediction model by sparse identification, the simulation unit calculates the first prediction value and / or the second prediction value for multiple functional forms composed of combinations of explanatory variables appearing in each term of the model equation constituting the prediction model. This makes it possible to calculate these prediction values ​​by calculating the model coefficients by sparse identification.

[0082] The present invention is not limited to the embodiments described above, and in the implementation stage, the components can be modified and implemented without departing from the gist of the invention, or the multiple components disclosed in the embodiments can be appropriately combined. [Explanation of Symbols]

[0083] 1. Target System 2. Data Acquisition Unit 3. Sensor 4...Storage section 5. Input section 6. Model Generation Unit 7. Simulation Department 8. Output section 9. Control Unit 10. External Systems 11. Coefficient Calculation Unit 100, 200, 300... Information processing device

Claims

1. An information processing device that performs simulations of predicted values, A model generation unit generates a predictive model for one or more evaluation indicators of the equipment based on equipment data obtained from sensors and a predetermined model formula. A simulation unit that predicts the evaluation index using the prediction model, It has an output unit that outputs the prediction model and / or the predicted evaluation index, The simulation unit calculates at least one improvement measure for the evaluation index by interpreting the physical phenomenon from the predetermined model equation, calculates a first predicted value which is the predicted value of the evaluation index when the calculated improvement measure is applied, and a second predicted value which is the predicted value of the evaluation index when the calculated improvement measure is not applied. The output unit outputs the improvement measures, the first predicted value, and the second predicted value. An information processing device characterized by the following:

2. An information processing apparatus according to claim 1, The output unit outputs at least one candidate improvement measure calculated by the simulation unit, and outputs the first predicted value when the selected improvement measure is applied. An information processing device characterized by the following:

3. An information processing apparatus according to claim 1, The simulation unit identifies at least one dominant factor, which is a physical factor, for the evaluation index in the predetermined model equation that constitutes the prediction model. The simulation unit calculates the degree of influence of the governing factors on the evaluation indicators in the prediction model, The output unit outputs the calculated governing factors and their respective influence levels. An information processing device characterized by the following:

4. An information processing apparatus according to claim 3, The aforementioned predetermined model equation is expressed as a mathematical formula consisting of at least one term, The simulation unit performs the identification of the governing factors and the calculation of the degree of influence of the governing factors in each term of the mathematical formula that represents the model formula that constitutes the prediction model. An information processing device characterized by the following:

5. An information processing apparatus according to claim 1, The aforementioned prediction model is one of the following: an equation model, a statistical model, or a stochastic model. An information processing device characterized by the following:

6. An information processing apparatus according to claim 1, The model generation unit generates the prediction model by either sparse identification or regression analysis. An information processing device characterized by the following:

7. An information processing apparatus according to claim 1, The aforementioned evaluation indicator is one of the following: a physical indicator, an economic indicator, or an environmental indicator. An information processing device characterized by the following:

8. An information processing apparatus according to claim 1, The system includes a model coefficient calculation unit that updates the prediction model by modifying the model coefficients through data assimilation so that the first predicted value obtained by the simulation unit approaches the actual data obtained from the sensor. An information processing device characterized by the following:

9. An information processing apparatus according to claim 1, The system has a control unit that controls the equipment according to the improvement measures calculated by the simulation unit, An information processing device characterized by the following:

10. An information processing apparatus according to claim 6, When the model generation unit generates the prediction model by sparse identification, The simulation unit calculates the first predicted value and / or the second predicted value for a plurality of functional forms composed of combinations of explanatory variables appearing in each term of the model equation that constitutes the prediction model. An information processing device characterized by the following:

11. A simulation method performed on an information processing device that simulates predicted values, The model generation unit generates a predictive model for one or more evaluation indicators of the equipment based on the equipment data obtained from the sensor and a predetermined model equation. The simulation unit predicts the evaluation indicator using the prediction model, The output unit outputs the prediction model and / or the predicted evaluation index. The simulation unit calculates at least one improvement measure for the evaluation index by interpreting the physical phenomenon from the predetermined model equation, calculates a first predicted value which is the predicted value of the evaluation index when the calculated improvement measure is applied, and a second predicted value which is the predicted value of the evaluation index when the calculated improvement measure is not applied. The output unit outputs the improvement measures, the first predicted value, and the second predicted value. A simulation method characterized by the following:

Images