JP7397623B2 - Plant condition evaluation device, plant condition evaluation method and program - Google Patents

Description

The present invention relates to a plant condition evaluation device, a plant condition evaluation method, and a program.

There is a known method for evaluating the state, performance, and degree of abnormality of a plant by analyzing the process data to be evaluated based on the process data of a plant model prepared in advance (including what is generally called a simulator). There is.
For example, the summary of Patent Document 1 below states, "[Solution] In a plant operation support system using a plant simulator that performs a simulation that follows the operation of an actual plant using a plant model representing the operation of the actual plant, Using a plant simulator that reproduces the state of an actual plant in real time based on actual measured values of the actual plant, the values of adjustment parameters that have been adjusted in advance based on the actual measured values of the actual plant and estimated values obtained by simulation using the plant simulator are used. It is characterized by determining whether or not it is within a predetermined allowable range and detecting an abnormality in the operating state of the actual plant."

In addition, the summary of Patent Document 2 below states, "[Solution] relates to a plant operation support device that performs process simulation in parallel with the operation of an actual plant and predicts the operation of the actual plant based on the simulation results. Yes, the simulation model is modified as needed based on actual data, the simulation model is modified based on the actual data received from the actual plant, and the modified simulation model is used to perform simulations in parallel with the operation of the actual plant. ” is stated.
In addition, the summary of Patent Document 3 listed below states, "[Solution means] means (1, 2) for extracting feature parameters of the input measurement signal from the input measurement signal, and extracting the feature parameters according to the plant operating state. A database that stores statistical normal range upper and lower limits as control values, and a comparison between the extracted feature parameters and the control values to determine a state in which the extracted parameter is not within the normal range or has a large deviation. means (3, 4) for storing the occurrence of the large deviation state; and determining that the plant is in an abnormal state when the frequency of occurrence of the state is greater than a predetermined value. "A plant monitoring device characterized by having means (6, 7)."
Further, the following non-patent document 1 discloses various design knowledge. Further, an example of an optimization method is shown in Non-Patent Document 2 below.

Japanese Patent Application Publication No. 2009-157550 Japanese Patent Application Publication No. 2005-332360 Japanese Patent Application Publication No. 2001-117626

GE Power Systems, GER-4190,Steam Turbine Thermal Evaluation and Assessment A R. Conn et.al, Introduction to Derivative-Free Optimization, Society for Industrial and Applied Mathematics Philadelphia,2008.

By the way, when a change in any process quantity in a plant is detected, many process quantities and their mutual relationships often exhibit behavior different from normal at the same time. Therefore, it is difficult to appropriately estimate (for example, estimate accurately at an early stage) the cause of abnormal fluctuations in process quantities.
The present invention has been made in view of the above-mentioned circumstances, and it is an object of the present invention to provide a plant condition evaluation device, a plant condition evaluation method, and a program that allow a user to appropriately estimate the cause of abnormal fluctuations in process quantities.

In order to solve the above problems, the plant condition evaluation device of the present invention includes a design knowledge storage unit that stores a plurality of functional formulas that define the relationships between a plurality of measured variables and a plurality of virtual variables in a plant, and The difference values between the plurality of values of the function equations based on the plurality of measured values and the assumed values of the plurality of virtual variables are calculated, and the plurality of difference values including those for at least flow rate and pressure are calculated. The present invention is characterized by comprising a virtual variable optimization calculation unit that calculates a total difference value, which is a sum, and calculates an optimal value of the virtual variable by minimizing the total difference value.

According to the present invention, by comparing the optimal value of each virtual variable with the corresponding measured value, the user can appropriately estimate the cause of the abnormal variation in the process amount.

1 is a block diagram of an online diagnostic system according to a first preferred embodiment; FIG. FIG. 2 is a diagram showing an example of a system configuration of a plant to be analyzed. FIG. 3 is a diagram showing a specific example of measurement variables. FIG. 3 is a diagram showing a specific example of virtual variables. FIG. 7 is a diagram showing another specific example of virtual variables. FIG. 7 is a diagram showing another specific example of virtual variables. FIG. 7 is a diagram showing another specific example of virtual variables. FIG. 7 is a diagram illustrating an example of the correspondence between a function name of a function that derives a virtual variable and a related variable. It is a figure which shows the example of the correspondence of a relational expression and a relational variable. It is a figure showing an example of a display screen. FIG. 7 is a diagram showing another example of a display screen. FIG. 7 is a diagram showing another example of a display screen.

<Overall configuration of embodiment>
FIG. 1 is a block diagram of an online diagnostic system 1 (plant condition evaluation device) according to a preferred embodiment. This online diagnosis system 1 is a system for performing online diagnosis of a plant 2.
The online diagnostic system 1 includes a plant signal input device 100, an arithmetic device 200 (computer), a GUI device 300, and a storage device 400 (computer). Here, the plant signal input device 100 acquires various measured values from the plant 2. The GUI device 300 includes a flat panel display that displays various information, and input devices such as a mouse and a keyboard.

The arithmetic device 200 and the storage device 400 are equipped with general computer hardware such as a CPU (Central Processing Unit), a RAM (Random Access Memory), a ROM (Read Only Memory), and an SSD (Solid State Drive). , the SSD stores an OS (Operating System), application programs, various data, and the like. The OS and application programs are loaded into RAM and executed by the CPU. In FIG. 1, functions realized by application programs and the like are shown as blocks inside the arithmetic device 200 and the storage device 400.

That is, the calculation device 200 includes a virtual variable optimization calculation section 210 (virtual variable optimization calculation means), an optimal calculation result display processing section 220, a measurement data error evaluation processing section 230, and a design knowledge modeling processing section 240. , and a signal preprocessing section 250. The storage device 400 also includes a measurement data storage section 410 and a design knowledge storage section 420 (design knowledge storage means).

FIG. 2 shows an example of the system configuration of the plant 2 to be analyzed.
In the illustrated example, the plant 2 is a power generation plant, and includes a reactor 21, turbines 221, 222, 223, a generator 23, a condenser 24, a condensate pump 25, a feed water heater 261, 262, 263, a water supply pump 27, piping (no reference numerals) connecting these, and a measuring section 28. The measurement unit 28 includes instruments (not shown) arranged in various parts of the plant 2. The items measured by these instruments are called "measurement variables." Further, the specific value of the "measurement variable" is called a "measurement value." The measurement unit 28 supplies these measured values to the plant signal input device 100 of the online diagnosis system 1.

<Data structure>
Next, the data structure in this embodiment will be explained.
FIG. 3 is a diagram showing a specific example of measurement variables measured by the measurement unit 28 (see FIG. 2).
In FIG. 3, names of measurement variables and meanings of measurement variables are listed. Here, the numbers “21”, “221”, “222”, “223”, “24”, or “25” included in the variable names of the measurement variables correspond to the symbols of each component in Figure 2. There is. For example, the measurement variables I_V_P_21_1 and I_V_P_21_2 shown in the first and second lines (excluding the title column; the same applies hereinafter) of FIG. It is a measurement variable.

The measured variables I_V_P_21_1 and I_V_P_21_2 both indicate the "plant pressure" of the reactor 21. This indicates that because the measuring unit 28 (see FIG. 2) redundantly arranges two similarly configured pressure measurement instruments, similar measurement variables I_V_P_21_1 and I_V_P_21_2 can be obtained. As shown in FIG. 3, in this embodiment, the name of the measurement variable is always a character string starting with "I_".

4, FIG. 5, FIG. 6, and FIG. 7 are diagrams showing specific examples of virtual variables.
Here, the virtual variable is a variable calculated by the signal preprocessing unit 250 based on the measurement variable and/or other virtual variables. Further, the values of virtual variables and measured values may be collectively referred to as "process amount." Here, "21", "221", "222", "223", "24", "25", "261", "262", "263", "27" included in the variable name of the virtual variable The numbers correspond to the symbols of each component in FIG. For example, the virtual variable V_P_21_in shown in the second line of FIG. 4 indicates the "plant pressure" of the reactor 21. The virtual variable V_P_21_in can be calculated, for example, by the average value of the above-mentioned measurement variables I_V_P_21_1 and I_V_P_21_2.

FIG. 8 is a diagram showing an example of the correspondence between the function name of a function that derives a virtual variable and a relational variable applied to the function. Here, the term "related variable" is a general term for measurement variables and virtual variables used in the function. Further, a relational variable that is input to the function is sometimes called an input relational variable, and a relational variable that is output is sometimes called an output relational variable. In FIG. 8, the function names start with "Eval_". Then, the character string obtained by removing the "Eval_" part from the function name becomes the variable name of the virtual variable to be determined, that is, the output-related variable.

Further, in FIG. 8, the variable names listed in the "Relationship Variable" column indicate the names of input relational variables for deriving output relational variables. For example, the function Eval_V_P_21_in described in the first line of FIG. 8 is a function that derives the output-related variable, ie, the virtual variable V_P_21_in, by removing the "Eval_" part. The measurement variables I_V_P_21_1 and I_V_P_21_2 described above become input related variables for deriving the virtual variable V_P_21_in. That is, the interrelationship between these relational variables can be expressed by the following equation.
V_P_21_in=Eval_V_P_21_in (I_V_P_21_1, I_V_P_21_2) ...Formula (1)

FIG. 9 is a diagram showing the correspondence between relational expressions and relational variables.
In this embodiment, the function (for example, the function shown in FIG. 8) for deriving a virtual variable is one that applies a "relational expression" to an "input relational variable." For example, the relational expressions TB1, TB2, and TB3 in the first to third lines of FIG. 9 are relational expressions expressing the input/output relationships of the turbines 221, 222, and 223, respectively. Moreover, the relational expression Rtr in the second line from the bottom in FIG. 9 is a relational expression expressing the input/output relationship of the reactor 21. These relational expressions TB1, TB2, and TB3 include "a mathematical expression for deriving the average value of input relational variables."

For example, the relational variables of the relational expression TB1 in the first line of FIG. 9 include the virtual variable V_P_21_in. This indicates that when the virtual variable V_P_21_in is included in the relational variables, the "average value calculation" in relational expression TB1 is applied. Therefore, the function Eval_V_P_21_in shown in FIG. 8 indicates the average value of the related variables I_V_P_21_1 and I_V_P_21_2, as shown in equation (2) below.
Eval_V_P_21_in(I_V_P_21_1, I_V_P_21_2)
= (I_V_P_21_1, I_V_P_21_2)/2...Formula (2)

By the way, when a certain function outputs multiple virtual variables, character strings such as ".1" and ".2" are appended to the function name to distinguish between these virtual variables. For example, assume that the input virtual variables of a certain relational expression R are Vin_1, Vin_2, and Vin_3, and that this relational expression R outputs Vout_1 and Vout_2 as two output virtual variables. Here, if the output virtual variable Vout_1 is the first output of the relational expression R and the output virtual variable Vout_2 is the second output, it may be expressed as the following equations (3) and (4).
Vout_1=R(Vin_1, Vin_2, Vin_3).1...Formula (3)
Vout_2=R(Vin_1, Vin_2, Vin_3).2...Formula (4)

In addition, other examples other than the normal "average value calculation" will also be explained.
In the third line of FIG. 8, the only measured variable, that is, the relational variable, related to the function Eval_V_T_21_out is the relational variable I_V_T_21_out_12. In such a case, the relational variable I_V_T_21_out_12 may be used as it is as the function Eval_V_T_21_out, but it may also be processed, such as by averaging the values of the relational variable on the time axis. For example, if the function Eval_V_T_21_out, the virtual variable V_T_21_out, and the relational variable I_V_T_21_out_12 are functions of time t, and ΔT is a predetermined period, the virtual variable V_T_21_out can be expressed by the following equation (5).
V_T_21_out(t)
=Eval_V_T_21_out(I_V_T_21_out_12).1
=(I_V_T_21_out_12(t-2ΔT) + I_V_T_21_out_12(t-ΔT)
+ I_V_T_21_out_12(t))/3 …Equation (5)

<Details of online diagnosis system 1>
Returning to FIG. 1, the plant signal input device 100 receives various measured values from the plant 2.
(Storage device 400)
Furthermore, the measurement data storage section 410 in the storage device 400 stores measurement values input in the past via the plant signal input device 100. Further, the design knowledge storage unit 420 stores data etc. listed below.
・Temperature versus pressure in saturated steam using steam table function,
・Relationship between turbine input steam enthalpy and output steam enthalpy, temperature, pressure, amount of extracted air, etc.
・Other relationships shown in Non-Patent Document 1, etc.

(Design knowledge modeling processing unit 240)
The design knowledge modeling processing unit 240 stores the plant model PM. The plant model PM defines the relationship between virtual variables. That is, the design knowledge modeling processing unit 240 stores the contents of FIGS. 3 to 9 described above.

(Signal preprocessing unit 250)
The signal preprocessing unit 250 in the arithmetic device 200 correlates the measured values received by the plant signal input device 100 with virtual variables based on the above-mentioned equations (2) to (5) and the like. For example, assume that two pressure gauges are redundantly arranged in order to measure the pressure at a certain location in the plant 2, as described above. The signal preprocessing unit 250 receives two measured values from these two pressure gauges via the plant signal input device 100. Then, the signal preprocessing unit 250 outputs, for example, the average value of these two measured values as the value of the virtual variable.

(Virtual variable optimization calculation unit 210)
The virtual variable optimization calculation unit 210 calculates the optimal value of each virtual variable. In other words, the virtual variable optimization calculation unit 210 optimizes the entire virtual variables so that the relational expression of the virtual variables (see FIG. 9) is best satisfied. The details will be explained below.

Here, values called "hypothetical/measured values" for measured variables and virtual variables will be defined. The "assumption/measured value" of a virtual variable is an arbitrarily assumed value for the virtual variable. Further, the "assumption/measurement value" of a measurement variable (that is, a variable whose name starts with "I_", such as I_V_P_21_1) is a measurement value corresponding to the measurement variable. “Assumption/Measurement value” is written by adding “**” to the name of the virtual variable or measurement variable.

Assume that the input relational variables of a certain relational expression R are Vin_1, Vin_2, and Vin_3, and this relational expression R outputs Vout_1 and Vout_2 as two virtual variables. At this time, "the degree to which the relational expression R is not satisfied" is called the degree of unsatisfaction RH (difference value). The sum of "absolute value of 'output value - function value'" can be used as an expression indicating the degree of unsufficiency RH.
For example, in the above example, the degree of unsufficiency RH can be expressed by the following equation (6).
RH= |Vout_1**− R(Vin_1**, Vin_2**, Vin_3**).1|
+|Vout_2**− R(Vin_1**, Vin_2**, Vin_3**).2| …Formula (6)
Further, instead of this, the sum of "the square value of the absolute value of "output value - function value"" may be used as the degree of unsatisfaction RH.

Here, among the assumed/measured values Vin_1**, Vin_2**, and Vin_3** of input related variables, those that are "hypothetical values of virtual variables" and the assumed values of output related variables Vout_1**, Vout_2** The degree of unsufficiency RH can be minimized by varying RH. As a result, the results of optimizing the assumed/measured values Vin_1**, Vin_2**, Vin_3**, Vout_1**, Vout_2** of the related variables and the output related variables are obtained. The assumptions/measured values optimized in this way are called "optimal/measured values." Optimal/measured values are expressed by adding "*" to the variable name of the virtual variable or measurement variable. For example, the optimal/measured values obtained in the above example are Vin_1*, Vin_2*, Vin_3*, Vout_1*, Vout_2*.

With this notation, the optimal/measured value of each variable is the value that minimizes the total unsatisfied degree RHG (total difference value), which is the sum of the unsatisfied degrees RH of the entire plant 2, for example in the following equation (7). .
RHG=|V_P_21_in** - Eval_V_P_21_in(I_V_P_21_1, I_V_P_21_2).1|
＋|V_F_21_out** - Eval_V_F_21_out(I_V_F_21_1).1|
:
＋|V_F_25_out** - Eval_V_F_25_out(I_V_F_25_out). 1 |
＋|V_F_221_out** -TB1(V_F_21_out**,V_T_21_out**,V_P_221**,V_T_221**,
V_F_221_in**).1|
＋|Eff_221** -TB1(V_F_21_out**,V_T_21_out**,V_P_221**,V_T_221**,
V_F_221_in**).2|
＋|V_F_222_out** -TB2(V_F_22_out**,V_T_22_out**,V_P_222**,V_T_222**).1|
＋|Eff_222** -TB2(V_F_22_out**,V_T_22_out**,V_P_222**,V_T_222**,
V_F_222_in).2|
:
＋|V_Q_F21** -Rtr(V_P_21**, V_T_21_out**).1|
:
+|E_eff** -PLANT(V_E_23**, V_Q_F21**).1|
...Formula (7)

In the above equation (7), the first output of the relational equation TB1, that is, TB1(V_F_21_out**,V_T_21_out**,V_P_221**,V_T_221**,V_F_221_in**).1, is the output from the turbine 221 to the feed water heater 263. is the steam flow rate. Further, the second output of the relational expression TB1, that is, TB1(V_F_21_out**,V_T_21_out**,V_P_221**,V_T_221**,V_F_221_in**).2 is the efficiency of the turbine 221. The relational expressions TB2 and TB3 are also similar to the relational expression TB1. Further, the first output of the relational expression Rtr for the reactor 21, ie, Rtr(V_P_21**, V_T_21_out**).1, corresponds to the thermal output V_Q_21.

Further, in equation (7), the first output of the relational expression PLANT, that is, PLANT(V_E_23**, V_Q_F21**).1 is the efficiency of the plant 2. In this relational expression, electrical output V_E_23 and thermal output V_Q_F21 are input relational variables. Note that in the example shown in equation (7), each term is not particularly standardized, but each term may be standardized.

For example, it is conceivable to normalize the values of each virtual variable using preprocessing so that the average value of each virtual variable becomes "1" based on past measured values. Thereby, it is possible to suppress a situation in which only the error of a specific virtual variable greatly contributes to the overall total degree of unsatisfiedness RHG. By the way, optimization calculations include nonlinear and non-differentiable functions such as steam table functions. In such a case, optimization can be achieved using an optimization method such as that shown in Non-Patent Document 2. That is, the virtual variable optimization calculation unit 210 determines the optimal values V_P_21_in*, V_F_21_out*, V_F_25_out*, etc. of the virtual variables corresponding to the above-mentioned assumed/measured values.

(Optimum calculation result display processing unit 220)
The optimal calculation result display processing unit 220 performs display processing for displaying the optimal value of each virtual variable on the GUI device 300. An example is shown in FIG.
Here, FIG. 10 is a diagram showing an example of the display screen 310 in the GUI device 300.
In FIG. 10, a system diagram image 320 and status display columns 510, 520, 530, and 540 are displayed on the display screen 310. The system diagram image 320 also includes a reactor image 321, turbine images 371, 372, 373, a generator image 323, a condenser image 324, a condensate pump image 325, and a feed water heater image 381, 382. , 383, a water supply pump image 327, and a piping image (without reference numeral) connecting these.

The images included in these system diagram images 320 are the reactor 21, turbines 221, 222, 223, generator 23, condenser 24, condensate pump 25, and feed water heating shown in FIG. 2, respectively. 261, 262, 263, the water supply pump 27, and the piping (not numbered) connecting these.

The status display field 510 displays a graph of the elapsed time regarding the thermal output V_Q_F21 (see the first line of FIG. 4) of the reactor 21 (see FIG. 2) and a numerical value of the current value ("2800" in the illustrated example). it's shown. In addition, the status display column 520 shows a graph of the passage of time of the plant water supply flow rate V_F_21_in (see the third line in FIG. 4) supplied to the reactor 21, and a numerical value of the current value ("1800" in the illustrated example). is displayed.

In addition, the status display field 530 shows a graph of the passage of time of the third feed water heater efficiency Eff_263 (see the third line from the bottom in FIG. 7) in the feed water heater 263 (see FIG. 2), and a numerical value of the current value (as shown in the figure). In the example, "72%") is displayed. In addition, the status display field 540 displays a graph of the time elapsed water pump efficiency Eff_27 (see the second line from the bottom of FIG. 7) in the water pump 27 (see FIG. 2) and a numerical value of the current value (in the illustrated example, "65 %”) is displayed.

The user can operate the GUI device 300 based on design know-how, change the numerical values displayed in the status display columns 510 to 540 as appropriate, and input hypothetical numerical values that are considered to be more correct. The input numerical value is called the "specified numerical value." When the user inputs a specified numerical value for any virtual variable in any of the status display fields 510 to 540, the optimal calculation result display processing unit 220 inputs the specified numerical value to the other virtual variables in the other status display fields. Display the optimal value that is the result of optimizing the variable. An example is shown in FIG.

FIG. 11 is a diagram showing an example of another display screen 312 in the GUI device 300.
FIG. 11 shows a case where the user operates the GUI device 300 to increase the plant water supply flow rate (V_F_21_in in the third line of FIG. 4) from "1800" shown in FIG. 10 to "1850" which is the specified value. This is a display screen 312.
The status display column 520 includes, as the plant water supply flow rate, a pre-change value of "1800", a post-change value of "1850", and an arrow 522 connecting the two, in accordance with the user's operation.

In addition, the third feed water heater efficiency shown in the status display column 530 (see Eff_263 in the third line from the bottom in Figure 7) has decreased from "72%" shown in Figure 10 to "71%". are doing. That is, since the third feed water heater efficiency decreases by "1%", the text "1%" and a downward arrow 532 are shown in the status display column 530. In addition, the water pump efficiency of the water pump 27 shown in the status display field 540 (see Eff_27 in the second line from the bottom in FIG. 7) has changed from "65%" shown in FIG. 10 to "64%". It's changing. That is, since the water supply pump efficiency of the water supply pump 27 decreases by "1%," the text "1%" and a downward arrow 542 are also shown in the status display column 540. Further, the status display field 510 shows a downward arrow 512 and the characters "0.1%". This indicates that the thermal output V_Q_F21 (see the first line of FIG. 4) decreases by "0.1%."

In this way, the user can arbitrarily change the numerical value of a desired virtual variable (for example, to a value that seems more preferable) based on his/her own design know-how. Then, in conjunction with the changed numerical value, the values of other virtual variables are updated and displayed on the GUI device 300 as shown in FIG. 11. The operation in which the user inputs hypothetical numerical values on the GUI device 300 is called "adjustment setting." The contents of the adjustment settings for the virtual variables are stored in the design knowledge modeling processing unit 240 and added to the design knowledge model as constraints.

For example, as shown in Figure 11, when the plant water supply flow rate V_F_21_in is changed from "1800" to "1850", the constraint shown in equation (8) below is added to the design knowledge model corresponding to the adjustment setting. Ru.
V_F_21_in=1850...Formula (8)
Using this equation (8) constraint condition, if each virtual variable is recalculated so as to minimize the total unsatisfied degree RHG shown in equation (7), the heat output will be "0" as described above. The evaluation results show that the efficiency of the third feed water heater and the water pump efficiency of the water pump 27 both decrease by 1%.

Additionally, if there is a statistically large difference between the measurement variable, that is, the measured value of the instrument, and the optimal value of the output-related variable related to the function name shown in Figure 8 as a result of the optimal calculation, the instrument may be malfunctioning. It is suspected that there is. Hereinafter, the value of the measurement variable that generates the optimal value (V_P_21_in*) of the output-related variable (V_P_21_in) related to the function name (eg, Eval_V_P_21_in) shown in FIG. 7 will be referred to as the "evaluation value." If there is a statistically large difference between the measured values (I_V_P_21_1, I_V_P_21_2) and these evaluation values (for example, if the difference is greater than a predetermined value), it is likely that the instrument is malfunctioning. Suspected. Therefore, in such a case, when the user performs a predetermined operation, the measurement data error evaluation processing section 230 displays the evaluation value and the actual measurement value of the measurement variable side by side on the display screen.

FIG. 12 is a diagram showing an example of another display screen 314 in the GUI device 300.
On the display screen 314, a status display field 550 is shown corresponding to a pipe image (no code) connecting the feed water heater image 383 and the reactor image 321. The status display field 550 includes an evaluation value histogram 552 (image) representing the evaluation value and an actual measurement value histogram 554 representing two actual measurement values (IV_F_21_1, IV_F_21_2) regarding the water supply flow rate V_F_21_in to the reactor 21 (see FIG. 2). , 556 (image) are displayed. This allows the user to recognize that "there is a high possibility that a failure has occurred in the instruments that measure the actual measured values IV_F_21_1 and IV_F_21_2."

<Effects of embodiment>
As described above, the plant condition evaluation device (1) includes a design knowledge storage section (420) that stores a plurality of functional formulas that define the relationships between a plurality of measured values and a plurality of virtual variables in the plant (2), and a plurality of Calculate the difference values (RH) between the values of multiple functional expressions based on the measured values and the assumed values of multiple virtual variables, and calculate the total difference value (RHG) that is the sum of the multiple difference values (RH). It is preferable to include a virtual variable optimization calculation unit (210) that calculates the optimal value of the virtual variable by minimizing the total difference value (RHG). This allows the user to compare the optimal value of each virtual variable with the corresponding measured value, and appropriately estimate the cause of the abnormal variation in the process amount.

The plant condition evaluation device (1) also calculates an evaluation value, which is a virtual measurement value corresponding to the optimum value, based on the functional formula and the optimum value, and images (554, 556) and an image (552) corresponding to the evaluation value on the display device (300). As a result, the user can more appropriately estimate the cause of the process quantity fluctuation abnormality by visually viewing the images (554, 556) corresponding to the actual measurement values and the image (552) corresponding to the evaluation value. can.

Further, when a specified numerical value is input to some of the virtual variables, the virtual variable optimization calculation unit (210) optimizes other virtual variables other than the virtual variable to which the specified numerical value is specified. is preferred. Thereby, various virtual variables can be optimized based on the design knowledge possessed by the user.

Further, it is preferable that at least some of the plurality of function expressions (e.g., Eval_V_T_21_out) determine a corresponding virtual variable (e.g., V_T_21_out) based on the passage of time of the corresponding measurement value (e.g., I_V_T_21_out_12). This makes it possible to determine appropriate virtual variables, especially for measured values that vary over time.

<Modified example>
The present invention is not limited to the embodiments described above, and various modifications are possible. The embodiments described above are exemplified to explain the present invention in an easy-to-understand manner, and are not necessarily limited to those having all the configurations described. Further, other configurations may be added to the configuration of the above embodiment, and it is also possible to replace a part of the configuration with another configuration. Furthermore, the control lines and information lines shown in the figures are those considered necessary for explanation, and do not necessarily show all the control lines and information lines necessary on the product. In reality, almost all components may be considered to be interconnected. Possible modifications to the above embodiment include, for example, the following.

(1) Since the hardware of the arithmetic device 200 and the storage device 400 in the above embodiment can be realized by a general computer, programs etc. that execute the various processes described above can be stored in a storage medium or distributed via a transmission path. You may.

(2) Each of the above-mentioned processes has been explained as a software process using a program in the above embodiment, but some or all of them can be implemented using an ASIC (Application Specific Integrated Circuit) or an FPGA (Field It may be replaced with hardware processing using a programmable gate array (Programmable Gate Array) or the like.

1 Online diagnosis system (plant condition evaluation device)
2 Plant 200 Arithmetic device (computer)
210 Virtual variable optimization calculation unit (virtual variable optimization calculation means)
230 Measurement data error evaluation processing unit 300 GUI device (display device)
400 Storage device (computer)
420 Design knowledge storage unit (design knowledge storage means)
552 Evaluation value histogram (image)
554,556 Actual value histogram (image)
RH unsatisfied degree (difference value)
RHG Total unsatisfied degree (total difference value)

Claims (6)

a design knowledge storage unit that stores a plurality of functional formulas that define relationships between a plurality of measured variables and a plurality of virtual variables in the plant;
Calculating the respective difference values between the values of the plurality of functional expressions based on the plurality of measured values, which are specific values of the measurement variables, and the assumed values of the plurality of virtual variables, including those for at least flow rate and pressure. a virtual variable optimization calculation unit that calculates an optimal value of the virtual variable by calculating a total difference value that is the sum of a plurality of the difference values and minimizing the total difference value. Condition evaluation device. An evaluation value that is a virtual measurement value corresponding to the optimum value is calculated based on the function formula and the optimum value, and an image corresponding to the actual measurement value and an image corresponding to the evaluation value are calculated. The plant condition evaluation device according to claim 1, further comprising: a measurement data error evaluation processing unit that displays on a display device. When a specified numerical value is input to some of the virtual variables, the virtual variable optimization calculation unit optimizes the other virtual variables other than the virtual variable to which the specified numerical value is specified. The plant condition evaluation device according to claim 1, characterized in that: The plant state evaluation device according to claim 1, wherein at least some of the plurality of functional expressions determine the corresponding virtual variables based on the passage of time of the corresponding measured values. a process of storing in a design knowledge storage section a plurality of functional formulas that define relationships between a plurality of measured variables and a plurality of virtual variables in the plant;
Calculating the respective difference values between the values of the plurality of functional expressions based on the plurality of measured values, which are specific values of the measurement variables, and the assumed values of the plurality of virtual variables, including those for at least flow rate and pressure. A plant state evaluation method comprising the steps of determining a total difference value that is the sum of a plurality of the difference values, and calculating an optimal value of the virtual variable by minimizing the total difference value. computer,
a design knowledge storage means for storing a plurality of functional formulas defining relationships between a plurality of measured variables and a plurality of virtual variables in a plant;
Calculating the respective difference values between the values of the plurality of functional expressions based on the plurality of measured values, which are specific values of the measurement variables, and the assumed values of the plurality of virtual variables, including those for at least flow rate and pressure. virtual variable optimization calculation means for calculating an optimal value of the virtual variable by calculating a total difference value that is the sum of the plurality of difference values and minimizing the total difference value;
A program to function as

Images