JP2014002290A - Simulator and simulation execution method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To achieve a flexible response to user operation, suppressing burden in adjustment when making an adjustment to reduce a difference between an actual piece of data and a simulation result without reducing reproductivity of response time.SOLUTION: A simulator 10 for analyzing a phenomenon by using a computer includes: simulation execution means 18 for executing simulation; and correction mode execution means 15 for correcting a model expression in a source program on the basis of actual data so that a difference between the actual data acquired by a sensor and a simulation result corresponding to the actual data which is acquired through simulation execution by the simulation execution means 18 is within a set range.

Description

  The present invention relates to a simulator and a simulation execution method.

  Simulators for nuclear power plants and chemical plants, etc., simulate the equipment that makes up the plant system with model formulas, etc., which do not model all physical phenomena. In general, it is targeted at various parameters. For example, specifications are determined in consideration of calculation time, calculation accuracy, etc., rated values are extracted from design / structure data, and model formulas are constructed based on these values. For this reason, the behavior of the simulator may be different from that of the actual machine in a portion that is not modeled, and the operator who operates the actual machine may feel uncomfortable.

  About the difference which the behavior of a simulator generate | occur | produces with an actual machine, the model formula which simulated the flow volume is demonstrated as a specific example (FIG. 15).

FIG. 15 is an explanatory diagram (graph) showing the pump flow rate with respect to the pump rotational speed in comparison with the measured data 1 and the simulation result 2. In the graph shown in FIG. 15, the horizontal axis represents the pump rotation speed (%), and the vertical axis represents the pump flow rate (m ³ / hr). Reference numeral 3 indicates a difference between the actual measurement data 1 and the simulation result 2.

The pump flow rate F is expressed by the following equation (1) using the pump flow rate (rated flow rate) Fc at the rated time and the pump rotation speed (ratio to the rating:%) Pv.
F = Fc * Pv (1)

  The above equation (1) models the so-called proportional relationship, and is a model equation for calculating the pump flow rate F by multiplying the rated flow rate Fc by the pump rotational speed Pv. In such a model formula, when the pump rotational speed is 100%, the rated flow rate Fc is obtained, so that the difference from the actual machine is almost zero. However, for example, the difference 3 between the actual measurement data 1 and the simulation result 2 may become large at an intermediate rotation speed such as 40% or 60%. Factors that cause the large difference 3 are considered to be factors that are not reflected in the model formula, such as device characteristics and piping shapes.

  Of course, if all the elements such as the characteristics of the equipment and the shape of the piping are reflected in the model formula, the simulation accuracy can be improved and the difference 3 can be reduced, but the modeling load, confirmation of the elements to be considered, etc. Actually, reflecting all the factors affecting the pump flow rate F in the model formula is too much labor and time. Moreover, the increase in the calculation formula also leads to an increase in calculation time, which affects not only the double speed calculation but also the real time calculation. Furthermore, the increase in the calculation formula also increases the factors causing bugs.

  In view of the above-described circumstances, for example, a difference between the actual measurement data 1 and the simulation result 2 is obtained by setting an adjustment coefficient that can be changed in advance in the model formula or replacing the collected actual measurement data 1. There is a simulator provided with adjusting means for reducing 3. As an example of a simulator provided with the adjustment means for reducing such a difference 3, there exists what is described in Unexamined-Japanese-Patent No. 2011-123187 (patent document 1), for example.

JP 2011-123187 A

  In the conventional simulator including the adjusting means as described above, for example, when adjustment is performed so as to reduce the difference between the actual measurement data and the simulation result by adjustment using the adjustment coefficient, the simulation result and the actual machine are used to determine the coefficient value of the adjustment coefficient. Humans compare data, determine numerical values by desk study, and confirm the results by simulation again. Even in this work, in order to check the result after adjustment, it may be necessary to wait for an operation and a long-time response in order to bring the simulator into a target state. This occurs remarkably in the chemical plant simulator because it simulates a chemical reaction.

  One of the main uses of the simulator is driving training. In the operation training, in addition to starting / stopping the plant and normal operation at the rated time, training in the event of an abnormality or an accident is also performed. In particular, training using a simulator is indispensable because training at the time of an abnormality or accident cannot use an actual machine. In this case, there is a problem that the reproducibility of response time (real time property) may be lowered because it is necessary to design a model formula or the like in detail if the simulation is to be performed with high accuracy.

  On the other hand, in the case of the adjusting means that uses the measured data as it is, the difference between the measured data and the simulation result may be reduced without reducing the reproducibility of the response time under specific conditions.

  However, the response may be inappropriate under conditions different from the scene where the actual measurement data is obtained (for example, when an operation different from the scene where the actual measurement data is obtained). In other words, even if it is possible to output data and simulation results according to the scene from which the actual measurement data was obtained, if the operation is different from the scene from which the actual measurement data was obtained (deviates from the assumed scenario), the simulation results are inconsistent. May occur.

  Therefore, in the conventional simulator provided with the adjustment means which uses measured data as it is, there exists a subject that a flexible response cannot be obtained with respect to a user's operation. For example, in a simulator that uses driving training as one of the main applications, such as driving training devices, it is desired to be able to perform training assuming various scenes that do not limit the training scenario. It is important to meet the demand to get.

  The present invention has been made in consideration of the above-described circumstances, and suppresses the burden of adjustment at the time of adjustment to reduce the difference between the actual measurement data and the simulation result, while reducing the reproducibility of response time. An object of the present invention is to provide a simulator and a simulation execution method that obtain a flexible response to the operation of the above.

  A simulator according to an embodiment of the present invention is a simulator that analyzes a phenomenon using a computer in order to solve the above-described problem, and includes a simulation execution unit that executes a simulation, actual measurement data acquired from a sensor, and the simulation Correction mode execution means for correcting the model formula in the source program based on the actual measurement data so that the difference from the simulation result corresponding to the actual measurement data obtained by executing the simulation by the execution means falls within the set range. It is characterized by comprising.

  In order to solve the above-described problem, a simulation execution method according to an embodiment of the present invention is a simulation applied to a simulator including a means for executing a simulation and a means for adjusting a simulation parameter for executing the simulation. The method for adjusting the simulation parameter is a method for setting the difference between the actual measurement data acquired from the sensor and the simulation result corresponding to the actual measurement data obtained by executing the simulation by the means for executing the simulation. The step of determining whether or not the measured parameter is within the range and the means for adjusting the simulation parameter correspond to the actual measurement data and the actual measurement data obtained by executing the simulation in the determination step. When it is determined that the difference from the simulation result is not within the set range, the difference is generated by referring to the sensor value / variable correspondence database in which the sensor value held in the storage unit is associated with the variable. A step of extracting a variable associated with a sensor value of an existing sensor to identify the variable in which the difference occurs, and a calculation for adjusting a simulation parameter to simulate a phenomenon held in a storage unit A source program that executes calculation of the model formula including the variable identified in the step of identifying the variable in which the difference occurs is extracted from a source program database that stores a source program that executes the difference, and the difference is extracted from the source program. The model expression that includes the variable identified in the step of identifying the variable in which Characterized by comprising the steps of: leaving, the.

  According to the present invention, it is possible to obtain a flexible response to a user's operation without reducing the reproducibility of the response time while suppressing the burden of adjustment during adjustment to reduce the difference between the actual measurement data and the simulation result. Can do.

The block diagram which showed roughly the structure of the simulator which concerns on embodiment of this invention. The block diagram which shows roughly the flow of the data input / output to the simulator which concerns on embodiment of this invention. The processing flowchart which shows the process step of the simulation mode determination procedure which the simulator which concerns on embodiment of this invention performs. The block diagram which showed roughly the structure of the correction mode execution means with which the simulator which concerns on embodiment of this invention comprises. The processing flowchart which shows the process step of the 1st correction mode execution procedure which the simulator which concerns on embodiment of this invention performs. The processing flowchart which shows the process step of the malfunction determination process in the 1st correction mode execution procedure which the simulator which concerns on embodiment of this invention performs. The processing flowchart which shows the process step of the malfunction correction process in the 1st correction mode execution procedure which the simulator which concerns on embodiment of this invention performs. Explanatory drawing which shows the example in the case where the measurement data used for correction | amendment can be approximated as a primary approximation formula in the correction mode of the simulator which concerns on embodiment of this invention. Explanatory drawing which shows the example in the case where the actual measurement data used for correction | amendment can be approximated as an approximate expression of a polynomial in the correction mode of the simulator which concerns on embodiment of this invention. The processing flowchart which shows the process step of the 2nd correction mode execution procedure which the simulator which concerns on embodiment of this invention performs. The block diagram which showed roughly the structure of the training mode execution means with which the simulator which concerns on embodiment of this invention comprises. The processing flowchart which shows the processing step of the training mode execution procedure which the simulator which concerns on embodiment of this invention performs. The block diagram which showed roughly the structure of the analysis mode execution means with which the simulator which concerns on embodiment of this invention comprises. The processing flowchart which shows the processing step of the analysis mode execution procedure which the simulator which concerns on embodiment of this invention performs. Explanatory drawing (graph) which showed the pump flow volume with respect to pump rotation speed by comparing actual measurement data with a simulation result.

  Hereinafter, a simulator and a simulation execution method according to embodiments of the present invention will be described with reference to the accompanying drawings. A simulator according to an embodiment of the present invention described below is an example of a simulator that simulates a plant phenomenon.

  FIG. 1 is a block diagram schematically showing a configuration of a simulator 10 which is an example of a simulator according to an embodiment of the present invention.

  The simulator 10 is a device that simulates a phenomenon that occurs in, for example, a device that constitutes a plant system by calculating a model formula or the like that simulates a phenomenon using an arithmetic processing device such as a computer. The simulator 10 includes, for example, an input unit 11, an output unit 12, a communication unit 13, a mode selection unit 14, a correction mode execution unit 15, a training mode execution unit 16, an analysis mode execution unit 17, and a simulation execution that are each unit of the simulator 10. By causing a computer to execute a simulator program (hereinafter, abbreviated as PG) 24 that functions as the means 18, display processing means 19, storage means 21, and control means 22, the computer and software resources, which are hardware resources, are used. Realized in collaboration with a program.

  Further, in the simulator 10, as an example of various data accessed by each means of the simulator 10, a simulator PG 24, an actual measurement data database (hereinafter abbreviated as DB) 25, a sensor value / variable correspondence DB 26, a source PG 27, The initial value DB 28 and the analysis result information 29 are held in the storage unit 21.

  The input unit 11 is realized by, for example, an input device connected to a computer via an interface or an input device such as a keyboard and a mouse provided in the computer itself. The input unit 11 gives the input information to the control unit 22.

  The output unit 12 is realized by, for example, a display unit connected to a computer via an interface or a display unit such as a display provided in the computer itself, a printing unit such as a printer connected to the computer via an interface, or the like. When receiving the display request, the output unit 12 displays the content corresponding to the display request on the screen. Further, when receiving the print request, the output unit 12 prints out the content corresponding to the print request.

  The communication unit 13 has a function of transmitting / receiving data to / from an external device. The communication unit 13 transmits the data received from the control unit 22 to the external device, while giving the data transmitted from the external device to the control unit 22.

  The mode selection unit 14 has a function of recognizing which mode is selected from among the correction mode, the training mode, and the analysis mode set in the simulator 10 and switching the mode so as to execute the simulation according to the selected mode. . In the simulator 10, for example, in response to a request for a mode that the user wants to execute from the input unit 11, the mode selection unit 14 selects one mode to be executed from the correction mode, the training mode, and the analysis mode.

  The correction mode execution means 15 executes a correction mode in which the source PG 27 is automatically corrected so that the difference between the measured data and the simulation result falls within a set range (desired range), or manual correction is urged. For example, a correction mode execution procedure shown in FIGS. 5 and 10 described later is executed.

  The training mode execution means 16 is a means for adjusting various parameters (simulation parameters) when executing the simulation so as to be in a mode (training mode) suitable for driving training of the operator when executing the simulation. For example, a training mode execution procedure shown in FIG. In the training mode, execution speed has priority over calculation accuracy. That is, in the training mode, even if the calculation accuracy is lowered, the simulation is executed with an increased reproduction speed reproduction accuracy. In the simulator 10, the actual measurement data and the calculation result in the analysis mode are used in order to supplement the reduced calculation accuracy.

  The analysis mode execution means 17 is a means for adjusting various parameters (simulation parameters) when executing the simulation so as to be in a mode (analysis mode) suitable for analyzing the phenomenon when the simulation is executed. An analysis mode execution procedure shown in FIG. 14 described later is executed. In the analysis mode, calculation accuracy has priority over execution speed. That is, in the analysis mode, the calculation accuracy is improved and the event is simulated in detail.

  The simulation execution means 18 is a means for executing a simulation by executing a source PG 27 for calculating a model formula or the like for simulating a phenomenon. When the simulation execution unit 18 executes the simulation, the simulation is executed in any one of the correction mode, the training mode, and the analysis mode selected by the mode selection unit 14.

  The display processing unit 19 has a function of generating display information for displaying information on a display unit such as a display. For example, when the display processing unit 19 receives information such as a simulation result performed by the simulation execution unit 18 or a defective model formula or a model formula element determined by the correction mode execution unit 15, the display processing unit 19 displays the content of the received information. Display information is generated, and the generated display information is given to the control means 22. The display method may be initially set or may be set from the input means 11 each time.

  The storage unit 21 includes a storage area where data can be read (read) and written (write), and has a function of holding data in the storage area. Each means of the simulator 10 can access the storage means 21, and information held in the storage means 21 is referred to as necessary.

  The control means 22 is a means for controlling processing of the entire apparatus of the simulator 10, and includes input means 11, output means 12, communication means 13, mode selection means 14, correction mode execution means 15, training mode execution means 16, analysis mode. Data is exchanged with the execution means 17, the simulation execution means 18, the display processing means 19, and the storage means 21, and has a function of controlling them.

  When the control means 22 receives information from the input means 11, the output means 12, communication means 13, mode selection means 14, correction mode execution means 15, training mode execution means according to the type of information received by the input means 11. 16, the analysis mode execution means 17, the simulation execution means 18, the display processing means 19 and the storage means 21 give a request based on the received information.

  When the control means 22 receives information from the communication means 13, the output means 12, the mode selection means 14, the correction mode execution means 15, the training mode execution means 16, and the analysis mode execution according to the type of information received from the communication means 13. The received information is given to any one of the means 17, the simulation executing means 18, the display processing means 19, and the storage means 21.

  When receiving the display of the mode selected by the mode selection unit 14 or the simulation result executed by the simulation execution unit 18, the control unit 22 receives information corresponding to the received display request, and displays the received information as a display process. The display information which gives to the means 19 and displays the said information is produced | generated.

  In addition to the information on the mode selected by the mode selection unit 14, the control unit 22 is an automatic correction mode for automatically correcting the model expression of the source PG 27 when the correction mode is selected. It is recognized whether it is a manual correction mode for correcting the model formula of PG27. The control means 22 controls the correction mode execution means 15 depending on whether the automatic correction mode is selected or the manual correction mode is selected, that is, whether the automatic correction mode is on (ON) or off (OFF). .

  Upon receiving the display information generated by the display processing means 19, the control means 22 gives the received display information to the output means 12 as a display means together with a display request. The output means 12 displays the display content based on the given display information.

  The simulator PG 24 is a program for causing a computer to function as the simulator 10. In other words, the simulator PG 24 is a program that causes a computer to execute processing procedures such as a correction mode execution procedure to be described later.

  The actual measurement data DB 25 is actual measurement data acquired by each device of each system constituting the plant. The actual measurement data is stored in association with information that uniquely identifies the equipment such as the system number and system name assigned to each system, and the equipment number and equipment name assigned to each equipment. Therefore, if information that uniquely identifies the equipment constituting the plant, such as the system number and equipment number, is given as a search key, actual measurement data corresponding to the given search key can be searched and extracted.

  Further, the actual measurement data DB 25 includes information on the date and time when the actual measurement data is stored as arbitrary information. This is from the viewpoint of improving the convenience of the user. For example, when searching the actual measurement data required by the user, information on the date and time when the actual measurement data is stored is provided as reference information.

  The sensor value / variable correspondence DB 26 has information in which a sensor value in a plant to be simulated is associated with a variable used in a model formula obtained by modeling a sensor that acquires the sensor value. Therefore, if the sensor value is used as a search key, information on the variable associated with the sensor value can be searched and extracted.

The source PG 27 is a program that executes a calculation that simulates a phenomenon. In the case of a simulator that simulates a plant, it is often modularized in units of parts such as valves and pumps that constitute the plant. In the simulator that simulates a plant, for example, a source PG 27 that simulates components such as valves is provided. is there. The source PG27 represented by the expression (2) is an example in the case of obtaining the flow rate of the flow path constituted by a single valve, F is the flow rate of the flow path, Fc is the rated flow rate of the valve, and Zv is the valve flow rate. Opening degree.
F = Fc * Zv (2)

  The initial value DB 28 includes information associating a physical quantity representing the state of the plant such as a flow rate and pressure actually measured in the plant with an initial value of the physical quantity. Therefore, the initial value of the plant state quantity can be specified from the state quantity of the plant.

  The analysis result information 29 includes information that uniquely identifies an analysis simulation to be executed, such as the execution date and time or execution conditions of the analysis simulation, and information that associates the analysis result of the analysis simulation. Therefore, if the execution date and time of the analysis simulation or the execution condition is used as a search key, the analysis simulation result associated with the search key can be searched and extracted.

  The simulator 10 configured in this way identifies the cause of the failure by paying attention to a sensor whose difference between the actual measurement data and the simulation result exceeds a preset range. The cause of the failure is identified by focusing on sensors whose difference between the measured data and the simulation results exceeds the preset range. In addition, the validity of the response was confirmed by the human system with respect to the behaviors to be simulated such as abnormal events and accident events, etc., but the fault found in this confirmation work is the value simulated as a sensor In view of the fact that it can often be determined from the above.

  The simulator 10 described above includes, for example, an input unit 11, an output unit 12, a communication unit 13, a mode selection unit 14, a correction mode execution unit 15, a training mode execution unit 16, an analysis mode execution unit 17, and a simulation execution unit 18. The display processing unit 19, the storage unit 21, and the control unit 22 have been described. However, it is not always necessary to include all the units 11 to 19, 21 and 22 described above.

  For example, if communication with the outside is not performed, the simulator 10 that does not include the communication unit 13 can be configured. In addition, if driving training can be eliminated as an application of the simulator 10, the simulator 10 that does not include the training mode execution means 16 can be configured, and more detailed phenomenon analysis (in the analysis mode) can be used as the application of the simulator 10. If the simulation execution) can be eliminated, the simulator 10 that does not include the analysis mode execution means 17 can be configured. Furthermore, in the simulator 10 that does not include the training mode execution means 16 and the analysis mode execution means 17, the simulation execution mode is one. In this case, the mode selection means 14 can be omitted.

  FIG. 2 is a block diagram schematically showing the flow of data inputted to and outputted from the simulator 10.

  The simulator 10 has an input (INPUT) given to the actual measurement data output unit 30 corresponding to a sensor or a device of the actual plant, and an output (OUTPUT) from the actual measurement data output unit 30 obtained based on this input. Given. The output of the actual measurement data output unit 30 is stored in the actual measurement data DB 25 in association with a desired search key.

  In the simulator 10, the same input as the input to the actual measurement data output unit 30 is given to the simulation execution means 18 that simulates the phenomenon occurring in the plant. The simulation execution means 18 outputs the result calculated based on the given input to the comparator 31. On the other hand, in addition to the calculation result of the simulation execution means 18, the output of the actual measurement data output unit 30 is given to the comparator 31.

  The comparator 31 compares the output of the actual measurement data output unit 30 with the calculation result of the simulation execution means 18 and gives the comparison result to the correction mode execution means 15. Based on the output (comparison result) of the comparator 31, the correction mode execution means 15 determines whether or not the difference between the actual measurement data and the simulation result is within the set range.

  FIG. 3 is a process flow diagram showing processing steps (steps S1 to S6) of the simulation mode determination procedure executed by the simulator 10.

  In the simulator 10, when a mode selection command indicating which simulation mode is selected is given, the mode selection unit 14 determines whether the command is a command for selecting a correction mode, a training mode, or an analysis mode. The mode related to the determination result is determined as the simulation execution mode. That is, the simulation is executed in the mode determined by the mode selection unit 14.

  In the simulation mode determination procedure (steps S1 to S6), first, when a mode selection command indicating which simulation mode is selected is given (in the case of YES in step S1), the mode selection means 14 then gives It is determined whether the received mode selection command is a correction mode selection command (step S2).

  If the given mode selection command is a correction mode selection command (YES in step S2), the mode selection means 14 determines that the given mode selection command is a correction mode selection command, and the determination result. Is provided to the control means 22. The control means 22 generates a simulation execution command and gives it to the simulation execution means 18, and also generates a correction mode execution command obtained as a determination result of the mode selection command and gives it to the correction mode execution means 15. 10, the simulation is executed in the correction mode (step S3). When it is determined that the simulation execution mode is the correction mode, the simulation mode determination procedure ends (END).

  On the other hand, when the given mode selection command is not a correction mode selection command (NO in step S2), it is subsequently determined whether or not the given mode selection command is a training mode selection command ( Step S4). If the given mode selection command is a training mode selection command (YES in step S4), the mode selection means 14 determines that the given mode selection command is a training mode selection command, and the determination result. Is provided to the control means 22.

  The control means 22 generates a simulation execution command and gives it to the simulation execution means 18, and also generates a training mode execution command obtained as a determination result of the mode selection command and gives it to the training mode execution means 16. 10, the simulation is executed in the training mode (step S5). When it is determined that the simulation execution mode is the training mode, the simulation mode determination procedure ends (END).

  If the given mode selection command is neither a correction mode selection command nor a training mode selection command (NO in step S2 → NO in step S4), the mode selection means 14 analyzes the given mode selection command. The mode selection command is determined, and the determination result is given to the control means 22.

  The control means 22 generates a simulation execution command and gives it to the simulation execution means 18, and also generates an analysis mode execution command obtained as a determination result of the mode selection command and gives it to the analysis mode execution means 17. 10, the simulation is executed in the analysis mode (step S6). When it is determined that the simulation execution mode is the analysis mode, the simulation mode determination procedure ends (END).

  Note that the processing flow shown in FIG. 3 is an example, and the order in which the modes are determined is not limited to the order shown in FIG. In the simulation mode determination procedure executed by the simulator 10, the procedure is arbitrary as long as it is determined whether or not any two of the three modes that can be selected are selected.

  For example, in the simulation mode determination procedure, whether or not the analysis mode not shown in FIG. 3 is selected instead of the processing step (step S2 or step S4) for determining whether or not the correction mode or the training mode is selected. There may be provided a processing step for determining whether or not. Further, it may be determined whether or not the training mode is selected before determining whether or not the correction mode is selected.

  FIG. 4 is a block diagram schematically showing the configuration of the correction mode execution means 15 provided in the simulator 10.

  The correction mode execution means 15 includes a difference sensor value determination unit 33, a difference occurrence variable identification unit 34, a source PG search unit 35, a model formula extraction unit 36, a defective element extraction unit 37, and an approximate expression calculation unit 38. The source PG correction unit 39 and the operation control unit 40 are provided.

  The difference sensor value determination unit 33 has a function of determining whether or not the difference between the measured data acquired for each device from the comparator 31 (FIG. 2) and the simulation result is within a set range (desired range). The determination result is given to the control means 22. When the difference sensor value determination unit 33 determines that the difference between the actually measured data and the simulation result has occurred beyond the desired range, the control means 22 determines that a defect has occurred and automatically corrects the simulator 10. Alternatively, the user is informed that a problem has occurred and is prompted to correct it.

  The difference occurrence variable identification unit 34 generates the difference when the difference sensor value determination unit 33 determines that the difference between the actual measurement data and the simulation result exceeds the set desired range (allowable value). Has the function of identifying variables. The difference occurrence variable identification unit 34 refers to the sensor value / variable correspondence DB 26 and extracts a variable associated with the sensor value in which the difference has occurred, thereby identifying the variable in which the difference has occurred. Information on the variables identified by the difference occurrence variable identification unit 34 is given to the control means 22.

  The source PG search unit 35 has a function of searching for a source PG 27 that executes calculation of a model formula including a variable given as a search key using a variable as a search key. The source PG search unit 35 receives the variable information identified by the difference occurrence variable identification unit 34 from the control unit 22 and searches for the source PG 27 that executes the calculation of the model expression including the received variable. Information on the result of the search by the source PG search unit 35 is given to the control means 22.

  The model formula extraction unit 36 has a function of extracting a model formula including a variable identified by the difference occurrence variable identification unit 34 from the source PG 27 obtained as a search result of the source PG search unit 35. Information on the model formula extracted by the model formula extraction unit 36 is given to the control unit 22 and can be displayed on a display unit such as a display constituting the output unit 12, and the user can display the model formula extraction unit via the display unit. The contents of the model formula extracted by 36 can be confirmed.

  The defect element extraction unit 37 has a function of extracting the components of the model expression extracted by the model expression extraction unit 36 and whether or not there is actual measurement data corresponding to the components of the model expression extracted by the model expression extraction unit 36. And a function for confirming.

  The defect element extraction unit 37 extracts constituent elements from the model formula extracted by the model formula extraction unit 36, and checks whether or not there is actually measured data corresponding to the constituent elements of the extracted model formula. Information about the result of confirmation as to whether or not there is a constituent element of the model formula extracted by the defective element extraction unit 37 and actual measurement data corresponding to the constituent element is given to the control means 22.

  Here, whether or not the actual measurement data corresponding to the component exists is determined by, for example, information such as a system number and a device number corresponding to the part simulated by the source PG 27 obtained as a search result of the source PG search unit 35. It can be extracted from the actual measurement data DB 25 as a search key. Further, once the components of the model formula extracted by the operator are confirmed, information such as the corresponding system number and device number can be manually given.

The approximate expression calculation unit 38 uses a primary expression (y = ax + b) or a quadratic or higher order for the measured data corresponding to the constituent elements of the model expression extracted by the model expression extraction unit 36 that is determined to have the defect element extraction unit 37. polynomial ^{(e.g., y = ax 3 + bx 2} + cx + d) and function of determining whether or not can be approximated by, and a function of calculating an approximate expression if it can approximate the measured data by a linear equation or a second or higher polynomial Have.

  The function of calculating the approximate expression possessed by the approximate expression calculation unit 38 can be realized by utilizing a known program such as a program for calculating a polynomial approximate expression using the method of least squares. Information about the determination result of whether or not the approximate expression calculation unit 38 can obtain the approximate expression and information on the result of calculating the approximate expression are given to the control means 22.

  The source PG correction unit 39 has a function of correcting the source PG 27. The correction of the source PG 27 is performed according to a correction command (in the case of the manual correction mode) based on a user input operation or a correction command to an approximate expression indicating actual measurement data given from the control means 22 (in the case of the automatic correction mode). .

  The operation control unit 40 has a function of extracting information (operation information) related to the user's operation from the actual measurement data. The operation information extracted by the operation control unit 40 is given to the control means 22. The operation information given to the control means 22 is given to the simulation execution means 18 as operation information when a second correction mode execution procedure (FIG. 10) described later is executed.

  Note that the correction mode execution means 15 shown in FIG. 4 is an example, and is not necessarily limited to the difference sensor value determination unit 33, the difference occurrence variable identification unit 34, the source PG search unit 35, the model formula extraction unit 36, and the failure element described above. It is not necessary to include all of the extraction unit 37, the approximate expression calculation unit 38, the source PG correction unit 39, and the operation control unit 40. For example, if the second correction mode execution procedure (FIG. 10) described later is not executed as the correction mode execution procedure, the correction mode execution means 15 that does not include the operation control unit 40 can be configured.

  FIG. 5 is a process flow diagram showing the processing steps (step S11 and step S12) of the first correction mode execution procedure executed by the simulator 10, and FIG. 6 is the processing steps (steps S11a to S11c) of the defect determination step (step S11). ), And FIG. 7 is a process flow diagram showing processing steps (steps S12a to S12j) of the defect correction step (step S12).

  The first correction mode execution procedure is an example of a simulation execution method according to the embodiment of the present invention. For example, as shown in FIG. 5, a failure determination step of determining whether or not a failure occurs in the simulator 10. (Step S11) and a defect correction step (step S12) for correcting the defect determined in the defect determination step.

  In the first correction mode execution procedure, when execution of the processing step is started (START), first, a failure determination step (step S11) is executed, and then a failure correction step (step S12) is executed. (END). The defect determination step and the defect determination step are executed by the correction mode execution means 15.

  In the defect determination step (step S11) of the first correction mode execution procedure, for example, as shown in FIG. 6, first, when the processing step of the defect determination step is started (ENTER), step S11a is executed. .

  In step S11a, the difference sensor value determination unit 33 determines whether or not the difference between the actual measurement data acquired for each device and the simulation result is within a set range. As a result of the determination, if the difference exceeds the set allowable value, it is determined that there is a sensor whose difference exceeds the set allowable value (YES in step S11a), and then step S11b is executed.

  In step S11b, the variable in which the difference is generated is identified from the sensor determined that the difference exceeding the range (allowable value) set by the difference occurrence variable identification unit 34 has occurred. If the variable in which the difference has occurred is identified, step S11c is subsequently executed.

  In step S11c, first, the source PG search unit 35 searches from the source PG 27 that executes calculation of the model formula including the variable identified in step S11b. Then, in response to the search result of the source PG search unit 35, the model formula extraction unit 36 extracts the model formula including the variable identified by the difference occurrence variable identification unit 34 from the source PG 27 obtained as the search result. When step S11c is completed, the defect determination step (step S11) of the first correction mode execution procedure is completed (RETURN).

  On the other hand, if the difference sensor value determination unit 33 determines whether or not the difference between the actual measurement data acquired for each device and the simulation result is within the set range, the difference does not exceed the set allowable value. Then, it is determined that there is no sensor whose difference exceeds the set allowable value (NO in step S11a), the process returns to step S11a, and the processing steps after step S11a are executed.

  In the first correction mode execution procedure, when the defect determination step (step S11) is completed, a defect correction step (step S12) is subsequently executed. More specifically, in the defect correcting step (step S12), for example, as shown in FIG. 7, when a processing step is started (ENTER), step S12a is executed.

  In step S12a, the defect element extraction unit 37 extracts a component from the model expression extracted in the defect determination step (step S11), and determines whether or not actual measurement data corresponding to the component of the extracted model expression exists. Check. As a result of the confirmation, if there is actual measurement data corresponding to the extracted component of the model formula (YES in step S12a), then step S12b is executed.

  In step S12b, the approximate expression calculation unit 38 determines whether or not the approximate expression can be calculated for the actual measurement data corresponding to the component of the model expression determined to exist in step S12a.

  As a result of determining whether or not the approximate expression calculation unit 38 can calculate the approximate expression, if the approximate expression can be calculated (YES in step S12b), if the automatic correction mode is selected ( In the case of YES in step S12c), the source PG correction unit 39 subsequently corrects (corrects) the model expression so that the expression calculated in step S12b becomes an approximate expression (step S12d). When step S12d is completed, the defect correction process is completed (RETURN).

  On the other hand, in step S12a, when actual measurement data corresponding to the extracted model formula component does not exist (NO in step S12a), step S12e is executed. In step S12e, the model formula extracted in the defect determination step (step S11) is displayed on a display unit such as a display, and the contents of the model formula extracted by the model formula extraction unit 36 are presented to the user.

  When the presentation of the model formula is completed, the processing step of the defect correction process proceeds from step S12e to step S12f, and accepts an input operation for correcting the source PG 27 (step S12f). When the correction work for the source PG 27 is completed (YES in step S12g), the source PG correction unit 39 reflects the correction contents of the source PG 27 (step S12h). When step S12h is completed, the defect correction process is completed (RETURN).

  Further, in step S12b, as a result of determining whether or not the approximate expression calculation unit 38 can calculate the approximate expression, if the approximate expression cannot be calculated (NO in step S12b), step S12i is executed. The In step S12i, the defect element extraction unit 37 displays the component extracted from the model formula in step S12a on a display unit such as a display, and presents the contents of the component of the model formula extracted from the model formula to the user. When step S12i is completed, the processing step of the defect correction process proceeds from step S12i to step S12f, and thereafter, the processing steps after step S12f are executed.

  Furthermore, if the manual correction mode is selected in step S12c (NO in step S12c), step S12j is executed. In step S12j, the approximate expression calculated by the approximate expression calculation unit 38 is displayed on display means such as a display, and the contents of the calculated approximate expression are presented to the user. When step S12j is completed, the processing step of the defect correction process proceeds from step S12j to step S12f, and thereafter, the processing steps after step S12f are executed.

  The first correction mode execution procedure described above includes the defect determination process (step S11) and the defect correction process (step S12), but necessarily includes both processes (step S11 and step S12). There is no need. For example, if it is assumed that the user corrects the defect separately, the first correction mode execution procedure may be stopped only for specifying the defect. That is, the simulator 10 is applied, and the correction mode execution procedure that does not include the defect correction step (step S12) can be executed as an example of the simulation execution method according to the embodiment of the present invention.

  FIGS. 8 and 9 are examples in the case where the actual measurement data used for correction can be approximated in the correction mode of the simulator 10, and FIG. 8 is an explanatory diagram showing an example in which the actual measurement data can be approximated as a first-order approximation formula. FIG. 9 is an explanatory diagram (graph) showing an example in which the measured data can be approximated as an approximate expression of a polynomial.

  For example, in the example shown in FIG. 8, the actual measurement data 1 (solid line) is a straight line that satisfies the relationship y = 0.3x, and the simulation result 2 (broken line) is the straight line that satisfies the relationship y = 0.2x. is there. In this case, the source PG correction unit 39 changes the proportional coefficient “a” of y = ax from 0.2 to 0.3 (adjusts), thereby correcting an error generated between the actual measurement data 1 and the simulation result 2. adjust.

On the other hand, in the example shown in FIG. 9, the actual measurement data 1 (solid line) is a curve that satisfies the relationship y = 8E-06x ³ −0.0024x ² + 0.3549x + 0.3497, and the simulation result 2 (dashed line) is y. = A straight line that satisfies the relationship of 0.2x.

In this case, when the source PG 27 is created as a second-order or higher-order polynomial to be corrected, for example, y = ax ³ + bx ² + cx + d, the source PG correction unit 39 adjusts each adjustment coefficient a, Measured by changing b, c, d from a = 0 to 8E-06, b = 0 to -0.0024, c = 0.2 to 0.3549, and d = 0 to 0.3497. The error generated between the data 1 and the simulation result 2 is adjusted.

  In the source PG 27, for example, a model expression to be corrected is created (modeled) with a lower order than the approximate expression of the actual measurement data 2 such as a linear expression, and the adjustment coefficient of the model expression to be corrected is changed. If it is not possible to deal with this problem alone, the source PG 27 itself needs to be corrected. In this case, the approximate expression calculation unit 38 determines that the approximate expression cannot be obtained, and accepts the source PG 27 by the user (manual input) in the defect correction process.

  FIG. 10 is a process flow diagram showing process steps (steps S21, S22, S11, S12) of the second correction mode execution procedure executed by the simulator 10.

  The second correction mode execution procedure, which is an example of the simulation execution method according to the embodiment of the present invention, differs from the first correction mode execution procedure in that it further includes step S21 and step S22. There is no substantial difference in other respects. Therefore, in the description of the second correction mode execution procedure, the difference from the first correction mode execution procedure will be mainly described, and the processing steps having the same processing contents are denoted by the same step numbers and description thereof is omitted.

  When the second correction mode execution procedure receives an execution start request for the second correction mode execution procedure, its processing step is started (START).

  When the processing step is started, step S21 is first executed. In step S21, the operation control unit 40 extracts operation information included in the actual measurement data stored in the actual measurement data DB 25. The extracted operation information is given to the simulation execution means 18. When step S21 is completed, step S22 is subsequently executed.

  In step S22, the simulation execution means 18 receives the operation information extracted in step S21, acquires the plant information included in the actual measurement data read from the actual measurement data DB 25, and executes the simulation. When executing the simulation, it is necessary to set an initial value.

  For the initial value setting, the simulation execution means 18 refers to the initial value DB 28, specifies the initial value corresponding to the plant information (time-series data) included in the actual measurement data read from the actual measurement data DB 25, and specifies the specified initial value. This is done by setting The initial value is identified by first extracting the value with the earliest time from the plant information (time-series data) time-series data as an initial value, and the matching rate between the extracted initial value and each initial value in the initial value DB 28. This is done by calculating and selecting the initial value with the highest matching rate.

  As shown in FIG. 2, the result of the simulation executed by the simulation execution unit 18 is given to the comparator 31 and compared with the execution result of the simulation execution unit 18. The result of comparison by the comparator 31 is given to the correction mode execution means 15. When the execution of the simulation is started, the process proceeds to step S11. After step S11, the same processing steps as those in the first correction mode execution procedure are executed. When step S12 is completed, the second correction mode is executed. The procedure ends (END).

  Thus, in the second correction mode execution procedure, the operation information and the plant information included in the actual measurement data are extracted with reference to the actual measurement data DB 25, and the initial value of the physical quantity corresponding to the extracted plant information is set as the initial value DB 28. From the obtained operation information and plant information and the initial value of the physical quantity corresponding to the extracted plant information, the result obtained by executing the simulation and the measured data are used to determine the presence or absence of a defect. Therefore, for example, the simulation result is valid up to a certain point, but it is effective when a malfunction occurs after operating for a long time, such as when a malfunction occurs in the sensor value.

  FIG. 11 is a block diagram schematically showing the configuration of the training mode execution means 16 provided in the simulator 10.

  The training mode execution means 16 includes an n-times speed execution determination unit 41, a system determination unit 43, a model accuracy change availability determination unit 45, a time step change availability determination unit 46, a model accuracy change unit 47, and a time step change unit. 48.

  The n-times speed execution determination unit 41 has a function of determining whether or not it is necessary to execute a simulation at a speed higher than n (n> 1) times speed, that is, constant speed (1 time speed). Whether or not to execute the program at the n-times speed is determined based on, for example, information set in advance or given from the input means 11 afterwards.

  Here, as an example of the case where it is effective to execute the program at n times speed, the waiting time after the actual operation is long, and if the simulation (reproduction) is performed at the same time as the actual, the training time cannot be effectively used. This is the case. The n-times speed of the simulation can be realized by increasing the number of executions per unit time or changing the time step.

  The n-times speed execution determination unit 41 notifies the control unit 22 of the result of determining whether it is necessary to execute the simulation at n times speed. When executing a program at n-times speed, for example, selecting whether to increase the number of executions per unit time or change the time step is to set whether or not to execute the program at n-times speed Follow the settings made separately.

  The system determination unit 43 has a function of determining whether or not the system that executes the simulation at the n-times speed is a system to be trained. Whether or not the system is a training target system can be determined from an initial value at the time of simulation execution. Information on the result of determining whether or not the system determination unit 43 is a system to be trained is given to the control means 22.

  The model accuracy change possibility determination unit 45 has a function of determining whether or not the accuracy of a model that simulates a phenomenon can be changed when a simulation is executed. Whether or not the model accuracy can be changed is determined, for example, by determining whether or not the model accuracy can be dealt with by reducing the number of calculation nodes or the number of area divisions. When the model accuracy change possibility determination unit 45 executes the simulation, information on the determination result as to whether or not the accuracy of the model simulating the phenomenon can be changed is given to the control means 22.

  The time step change possibility determination unit 46 has a function of determining whether or not the time step can be changed when the simulation is executed. Whether or not the time step can be changed is determined by, for example, the type of parameter. Information on the result of determining whether the time step can be changed when the time step change enable / disable determining unit 46 executes the simulation is given to the control means 22.

  The model accuracy changing unit 47 has a function of changing the accuracy of the model when the model accuracy changing possibility determining unit 45 determines that the accuracy of the model simulating the phenomenon can be changed when executing the simulation. The model accuracy changing unit 47 changes the model accuracy in accordance with the changed model accuracy received from the input unit 11 or preset contents, and provides the control unit 22 with information on the changed model accuracy.

  The time step change unit 48 has a function of changing the time step when the time step change possibility determination unit 46 determines that the time step can be changed when the simulation is executed. The time step changing unit 48 changes the time step according to the changed time step received from the input unit 11 or the preset content, and provides the changed model accuracy information to the control unit 22.

  FIG. 12 is a process flow diagram showing processing steps (steps S31 to S40) of the training mode execution procedure executed by the simulator 10.

  The training mode execution procedure is an example of the simulation execution method according to the embodiment of the present invention. When a training mode execution procedure execution start request is received, the processing step is started (START). When the processing step is started, step S31 is first executed. In step S31, the n-times speed execution determination unit 41 determines whether to execute at n-times speed. As a result of determining whether or not to execute at n-times speed, if it is executed at n-times speed (YES in step S31), step S32 is subsequently executed.

  In step S32, the system determining unit 43 determines whether or not the system that executes the simulation at the n-times speed is a system to be trained, and the system that executes the simulation at the n-times speed is the system to be the training target. (In the case of YES at step S32), step S33 is subsequently executed.

  In step S33, it is determined whether or not the model accuracy change possibility determination unit 45 can change the accuracy of the model of the system that executes the simulation at the n-times speed, that is, the system to be trained.

  When the accuracy of the model of the system to be trained can be changed (YES in step S33), the model accuracy changing unit 47 changes the accuracy of the model of the system to be trained, that is, the accuracy of the simulator. (Reduced) (Step S34), the simulation execution means 18 applies the actual measurement data or the detailed calculation result such as the simulation execution result in the analysis mode to be described later to complement the lowered accuracy of the simulator (Step S34). S35).

  When the simulator accuracy change and the simulator accuracy supplement by the actually measured data are completed, the simulation execution means 18 starts the simulation at n times speed. Thereafter, when the training end request is received or the execution of the simulation is completed, the training mode execution procedure ends (END).

  On the other hand, as a result of determining whether or not the n-times speed execution determination unit 41 executes at n-times speed in step S31, if not executed at n-times speed (in the case of NO in step S31), the processing step of the training mode execution procedure In step S36, the simulation execution means 18 maintains the accuracy of the system model and starts executing the simulation as usual, that is, at the same magnification (1 × speed). Thereafter, when the training end request is received or the execution of the simulation is completed, the training mode execution procedure ends (END).

  In step S32, when the system determination unit 43 determines that the system that executes the simulation at the n-times speed is not the system to be trained (NO in step S32), the processing steps of the training mode execution procedure are as follows: Proceed to step S37. In step S <b> 37, it is determined whether the time step can be changed when the time step change enable / disable determination unit 46 executes the simulation.

  If the time step change enable / disable determining unit 46 determines that the time step can be changed when executing the simulation (YES in step S37), the time step changing unit 48 sets the time step of the system model for executing the simulation. Change (step S38). Then, the simulation execution means 18 starts an n-times speed simulation with the time step changed. Upon receiving the training end request or completing the execution of the simulation, the training mode execution procedure ends (END).

  Furthermore, when the model accuracy change possibility determination unit 45 determines in step S33 that the accuracy of the model of the system to be trained cannot be changed (NO in step S33), the processing steps of the training mode execution procedure are: Then, the process proceeds to step S36, and the processing steps after step S36 are executed.

  Furthermore, in step S37, when the time step change possibility determination unit 46 determines that the time step cannot be changed when the simulation is executed (NO in step S37), the simulation execution unit 18 stops the system model. At the same time (step S39), the accuracy of the lowered simulator is complemented by applying actual calculation data or detailed calculation results such as simulation execution results in an analysis mode to be described later (step S40).

  When the simulator accuracy change and the simulator accuracy supplement by the actually measured data are completed, the simulation execution means 18 starts the simulation at n times speed. Thereafter, when the training end request is received or the execution of the simulation is completed, the training mode execution procedure ends (END).

  As described above, according to the training mode execution means 16 and the training mode execution procedure, by reducing the calculation accuracy within a range that does not affect the training, while using the measured data or the like to supplement the reduced calculation accuracy, n times speed Even at the time of execution, higher training accuracy can be realized without increasing the calculation load. That is, conventionally, it is possible to achieve both the training accuracy and the calculation load that are in a trade-off relationship.

  FIG. 13 is a block diagram schematically showing the configuration of the analysis mode execution means 17 provided in the simulator 10.

  The analysis mode execution means 17 includes a system determination unit 43, a model accuracy change enable / disable determination unit 45, a time step change enable / disable determination unit 46, a model accuracy change unit 47, a time step change unit 48, and a data collection unit 51. Is provided. Here, the system determination unit 43, the model accuracy change availability determination unit 45, the time step change availability determination unit 46, the model accuracy change unit 47, and the time step change unit 48 include the system determination unit 43, the model provided in the training mode execution means 16. This is substantially the same as the accuracy change enable / disable determining unit 45, the time step change enable / disable determining unit 46, the model accuracy changing unit 47, and the time step changing unit 48.

  The data collection unit 51 has a function of collecting simulation execution results in the analysis mode as analysis result information 29. The data collection unit 51 can save the simulation execution result in the analysis mode as the analysis result information 29, for example, in the storage unit 21.

  In the simulator 10, the data collection unit 51 collects simulation results corresponding to actually measured data that is difficult to collect, so that a more accurate simulation execution result can be used instead of the actually measured data that is difficult to collect. That is, in the simulator 10, the data collection unit 51 has higher accuracy than the collected data corresponding to the actually measured data that cannot be basically collected, such as actually measured data in an abnormal state (abnormal state) such as an accident. The simulation execution result can be used to improve the accuracy of the simulator 10.

  The above-described analysis mode execution means 17 is an example including the data collection unit 51. However, even if the analysis mode execution means 17 does not include the data collection unit 51, the simulation can be executed in the analysis mode. it can. That is, the presence or absence of the data collection unit 51 does not affect the execution of the simulation in the analysis mode.

  FIG. 14 is a process flow diagram showing the processing steps (steps S51 to S57) of the analysis mode execution procedure executed by the simulator 10.

  The analysis mode execution procedure is an example of the simulation execution method according to the embodiment of the present invention. When an analysis mode execution procedure execution start request is received, the processing step is started (START). When the processing step is started, step S51 is first executed. In step S51, when the system determination unit 43 determines whether or not the system for executing the simulation in the analysis mode is the system to be analyzed, and determines that the system is to execute the simulation in the analysis mode (YES in step S51). Step S52 is subsequently executed.

  In step S52, the model accuracy change availability determination unit 45 determines whether the model accuracy can be changed for a model of a system that executes simulation in the analysis mode. When it is determined that the model accuracy can be changed for the model of the system that executes the simulation in the analysis mode (YES in step S52), the model of the model of the system that the model accuracy changing unit 47 executes the simulation in the analysis mode The accuracy is changed (improved) (step S53).

  Subsequently, when the model accuracy changing unit 47 improves the model accuracy of the model of the system that executes the simulation in the analysis mode, the simulation executing unit 18 starts the simulation, and the data collecting unit 51 collects the simulation execution result ( Step S54). Thereafter, the analysis mode execution procedure ends when the analysis end request is received or the execution of the simulation is completed (END).

  On the other hand, when the system determination unit 43 determines in step S51 that the system for executing the simulation in the analysis mode is not the system to be analyzed (NO in step S51), the processing step of the analysis mode execution procedure is performed in step S51. Proceed to S55. In step S55, it is determined whether the time step can be changed when the time step change enable / disable determining unit 46 executes the simulation.

  When the time step change enable / disable determining unit 46 determines that the time step can be changed when executing the simulation (YES in step S55), the time step changing unit 48 sets the time step of the system model for executing the simulation. The time step is changed to a finer time step (step S56). Then, the simulation is started with the simulation execution means 18 changing the time step, and the data collection unit 51 collects the simulation execution results (step S54). Thereafter, the analysis mode execution procedure ends when the analysis end request is received or the execution of the simulation is completed (END).

  Further, in step S52, when the model accuracy change possibility determination unit 45 determines that the model accuracy cannot be changed for the model of the system that executes the simulation in the analysis mode (NO in step S52), the analysis mode execution procedure In step S57, the simulation execution means 18 maintains the accuracy of the system model (step S57) and starts executing the simulation as usual, that is, at the same magnification (1 × speed). 51 collects simulation execution results (step S54). Thereafter, the analysis mode execution procedure ends when the analysis end request is received or the execution of the simulation is completed (END).

  Furthermore, if it is determined in step S55 that the time step changeability determination unit 46 cannot change the time step when executing the simulation (NO in step S55), the processing step of the analysis mode execution procedure is performed in step S55. Proceeding to S57, the processing steps after step S57 are executed.

  Thus, according to the analysis mode execution means 17 and the analysis mode execution procedure, the calculation accuracy is prioritized over the execution speed, the time step is made as fine as possible (short) as much as possible, and the region can be divided. Since the simulation can be executed with the maximum precision to improve the calculation accuracy, the obtained simulation execution result can be used instead of the actual measurement data that is difficult to collect, such as actual measurement data at the time of an abnormality such as an accident. Can be used. In other words, insufficiently collected actual measurement data can be supplemented with simulation execution results, for example, even if there is no actual measurement data corresponding to the elements constituting the model formula extracted in the defect correction process of the correction mode execution procedure. If there is a simulation execution result collected in the mode, the defect correction process can be executed by using the simulation execution result.

  As described above, according to the simulator 10 and the correction mode execution procedure, the training execution mode execution procedure, and the analysis mode execution procedure as the simulation execution method, when the difference between the actual measurement data and the simulation result becomes larger than the preset range, the actual measurement data Etc. can be used to correct the error.

  The verification of simulators targeting plants has traditionally confirmed the validity of responses for the simulation target operations such as abnormal events and accident events in addition to plant startup, shutdown and steady operation. Since the simulator 10 performs a work corresponding to the confirmation work, it is possible to suppress the burden of adjustment during adjustment (error correction) that reduces the difference between the actual measurement data and the simulation result. Moreover, since the measured data is not reproduced as it is but used for adjusting the model formula, a flexible response to the user's operation can be obtained without reducing the reproducibility of the response time.

  Also, in the training execution mode execution procedure, the calculation accuracy is reduced within a range that does not affect the training, while the measured data is used to supplement the reduced calculation accuracy, thereby increasing the calculation load even during n-times speed execution. Therefore, it is possible to achieve higher training accuracy than before. That is, it is possible to achieve both training accuracy and calculation load.

  Furthermore, in the analysis mode execution procedure, it is possible to prioritize the calculation accuracy over the execution speed and execute the simulation with improved calculation accuracy. The result can be supplemented. For example, even if there is no actual measurement data corresponding to the elements constituting the model expression extracted in the defect correction step of the correction mode execution procedure, if there is a simulation execution result collected in the analysis mode, the simulation execution result It is possible to execute the defect correction process by substituting.

  It should be noted that the present invention is not limited to the above-described embodiment as it is, and can be implemented in various forms other than the above-described examples in the implementation stage, and various modifications can be made without departing from the spirit of the invention. Can be omitted, added, replaced, or changed. These embodiments and modifications thereof are included in the scope and gist of the invention, and are included in the invention described in the claims and the equivalents thereof.

  In addition, the method described in the above-described embodiment is applied as a program that can be executed by a computer to a storage medium such as a magnetic disk, an optical disk, or a semiconductor memory and applied to various apparatuses, or transmitted by a communication medium. It can also be applied to various devices. A computer that implements the apparatus reads the program recorded in the storage medium, and executes the above-described processing by controlling the operation by the program.

  DESCRIPTION OF SYMBOLS 1 ... Actual measurement data, 2 ... Simulation result, 3 ... Difference between actual measurement data and simulation result, 10 ... Simulator, 11 ... Input means, 12 ... Output means, 13 ... Communication means, 14 ... Mode selection means, 15 ... Correction mode Execution means, 16 ... training mode execution means, 17 ... analysis mode execution means, 18 ... simulation execution means, 19 ... display processing means, 21 ... storage means, 22 ... control means, 24 ... simulator PG, 25 ... actual data DB, 26 ... Sensor value / variable correspondence DB, 27 ... Source PG, 28 ... Initial value DB, 29 ... Analysis result information, 30 ... Actual measurement data output unit, 31 ... Comparator, 33 ... Difference sensor value determination unit, 34 ... Difference occurrence Variable identification unit, 35... Source PG search unit, 36... Model formula extraction unit, 37... Defective element extraction unit, 38. Correction unit, 40 ... operation control unit, 41 ... n-times speed execution determination unit, 43 ... system determination unit, 45 ... model accuracy change availability determination unit, 46 ... time step change availability determination unit, 47 ... model accuracy change unit, 48 ... Time step changing unit, 51... Data collecting unit.

Claims (13)

It is a simulator that analyzes phenomena using a computer,
A simulation execution means for executing the simulation;
In the source program based on the actual measurement data, the difference between the actual measurement data acquired from the sensor and the simulation result corresponding to the actual measurement data obtained by the simulation execution means executing the simulation is within a set range. And a correction mode executing means for correcting the model formula. The correction mode execution means, based on the output of a comparator that compares the actual measurement data and a simulation result corresponding to the actual measurement data obtained by the simulation execution means executing the simulation, the actual measurement data and the simulation execution A difference sensor value determination unit that determines whether a difference from a simulation result corresponding to the actual measurement data obtained by executing the simulation is within the set range;
When the difference sensor value determination unit determines that the difference between the actual measurement data and the simulation result corresponding to the actual measurement data obtained by executing the simulation by the simulation execution unit is not within the set range, Refers to the sensor value / variable correspondence database that associates the sensor value and variable held in the storage means, extracts the variable associated with the sensor value of the sensor in which the difference occurs, and generates the difference A difference occurrence variable identification unit for identifying a variable
Source program search for extracting a source program for calculating a model formula including a variable identified by the difference occurrence variable identification unit from a source program database for storing a source program for performing a calculation for simulating a phenomenon held in a storage means And
A model formula extraction unit that extracts the model formula including the variable identified by the difference occurrence variable identification unit from the source program extracted by the source program search unit;
Extracting a component from the model formula extracted by the model formula extraction unit, and checking whether there is actually measured data corresponding to the extracted component of the model formula;
An approximate expression calculation unit that calculates the approximate expression when the approximate expression of the actual measurement data corresponding to the component of the model expression determined that the defective element extraction unit is present;
The simulator according to claim 1, further comprising: a source program correction unit that corrects the source program stored in the source program database. The simulator according to claim 2, wherein the correction mode execution unit further includes an operation control unit that extracts information related to a user operation included in the actual measurement data. The correction mode execution means displays the model formula extracted by the model formula extraction unit when there is no actual measurement data corresponding to the component of the model formula extracted from the model formula extracted by the model formula extraction unit In the case where it is not possible to calculate an approximate expression of the actual measurement data corresponding to the constituent element of the model formula that is configured to be displayed on the means and determined that the defective element extraction unit exists, the defective element extraction unit 4. The simulator according to claim 2, wherein the simulator is configured to display a component of the model formula determined to be present on a display unit. The correction mode execution means automatically adjusts the coefficient of the model formula of the source program stored in the source program database by the source program correction unit so as to match the approximate formula calculated by the approximate formula calculation unit. 5. The simulator according to claim 2, further comprising: a correction mode; and a manual correction mode that accepts a user input operation and corrects the source program stored in the source program database. 6. . Training mode execution means for adjusting simulation parameters by giving priority to increasing the reproduction accuracy of execution speed over calculation accuracy, and means for adjusting simulation parameters when the simulation execution means executes the simulation. The simulator according to any one of claims 1 to 5, comprising the simulator. The training mode execution means includes an n-times speed execution determination unit that determines whether or not to execute the simulation at an n-times speed when n> 1.
A system determination unit that determines whether or not the system for executing the simulation at the n-times speed is a system to be trained;
A model accuracy change possibility determination unit that determines whether it is possible to change the accuracy of the model of the system to be trained;
A model accuracy changing unit that changes the accuracy of the model of the system to be trained when the model accuracy changeability determining unit determines that the accuracy of the model of the system to be trained can be changed; ,
A time step changeability determination unit that determines whether or not it is possible to change a time step of a model of a system that is not a target of the training when the simulation is performed;
When the time step change enable / disable determining unit determines that it is possible to change the time step of the model of the system that is not subject to training when executing the simulation, the time step change that changes the time step The simulator according to claim 6, further comprising: a unit. When the model accuracy changing unit changes the accuracy of the model of the system to be trained, the simulation execution unit reads the actual data of the system to be trained from the measured data held in the storage unit. The simulator according to claim 7, wherein the simulator is configured to be applied. When the simulation execution unit determines that the time step changeability determination unit cannot change the time step of the model of the system that is not the target of the training, the simulation is performed for the model of the system that is not the target of the training. 9. The system according to claim 7 or 8, characterized in that the measurement data of the model of the system not subject to training is read out and applied from the actual measurement data held in the storage means. Simulator. A means for adjusting a simulation parameter when the simulation execution means executes the simulation, and further includes an analysis mode execution means for adjusting the simulation parameter in preference to increasing the calculation accuracy over the execution speed. The simulator according to claim 1, wherein: The analysis mode execution means, a system determination unit for determining whether or not the system for executing the simulation is a system to be analyzed,
A model accuracy change possibility determination unit that determines whether it is possible to change the accuracy of the model of the system to be analyzed;
A model accuracy changing unit that changes the accuracy of the model of the system to be analyzed when the model accuracy changeability determining unit determines that the accuracy of the model of the system to be analyzed can be changed; ,
A time step changeability determination unit that determines whether it is possible to change a time step of a model of a system that is not an object of the analysis when the simulation is executed;
When the time step change enable / disable determining unit determines that it is possible to change the time step of the model of the system that is not the analysis target when executing the simulation, the time step change that changes the time step The simulator according to claim 10, further comprising: a unit. 12. The analysis mode execution means further includes a data collection unit that collects the result of the simulation executed by the simulation parameter that prioritizes increasing the calculation accuracy over the execution speed. The simulator described. A simulation execution method applied to a simulator comprising means for executing simulation and means for adjusting simulation parameters for executing the simulation,
The means for adjusting the simulation parameter is within a set range of a difference between the actual measurement data acquired from the sensor and the simulation result corresponding to the actual measurement data obtained by the simulation execution unit executing the simulation. Determining whether or not
The means for adjusting the simulation parameter determines that the difference between the actual measurement data and the simulation result corresponding to the actual measurement data obtained by executing the simulation is not within a set range in the determination step. In this case, referring to the sensor value / variable correspondence database in which the sensor value held in the storage means is associated with the variable, the variable associated with the sensor value of the sensor in which the difference occurs is extracted and the Identifying the variable where the difference occurs;
The variable identified in the step of identifying the variable in which the difference occurs from a source program database storing a source program that stores a source program for executing a simulation that simulates a phenomenon held in the storage means, Extracting a model program that includes the identified variable in the step of identifying a variable in which the difference occurs from the source program, and extracting a source program that executes calculation of the model formula including the method. A characteristic simulation execution method.

Images