JP2002364311A - Supporting system for dynamic behavior analysis of thermal electric power plant - Google Patents

Abstract

PROBLEM TO BE SOLVED: To shorten an adjusting period of a dynamic behavior model adjusted by using actual machine data, and to accurately support dynamic behavior analysis of a thermal electric power plant. SOLUTION: This supporting system for dynamic behavior analysis of a thermal electric power plant comprises a dynamic behavior model creating means 2000 establishing an attribute 1100 to equipment modules 100 from an attribute of respecting equipment composing a plant, and automatically constructing a dynamic behavior model 310 by combining a dynamic behavior modules 110 by equipment generated from an interface 1200 between the modules; an attribute parameter optimizing means 3000 comparing and assessing actual machine data 300 and a result of the dynamic behavior model, and correcting and adjusting the attribute parameter; and a means 4000 estimating a thermal power plant design value from the corrected and adjusted attribute parameter. The dynamic behavior model creating means partially takes out an adjusting object from the dynamic behavior model of the whole of the plant to create a dynamic behavior part area model, automatically creates an observer estimating an input/output value of the part area model from an actual machine measurement value, and adjusts the parameter of the dynamic behavior model by using the observer.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a thermal power plant dynamic characteristic analysis support system, and more particularly to a plant design which is a part of plant equipment attributes based on equipment-specific attributes and connection information and actual measurement data of the thermal power plant. The present invention relates to a support system for analyzing dynamic characteristics of a thermal power plant for estimating a reverse value.

[0002]

2. Description of the Related Art Conventionally, in the dynamic characteristic analysis of a process such as a thermal power plant or a chemical plant, a physical model based on the law of conservation of mass, energy, and momentum and other characteristic formulas and a certain identification signal are applied to an actual plant. In addition, a statistical model based on the correlation with the state quantity is obtained. The physical model has a feature that a dynamic characteristic model can be constructed without using actual machine data. However, there is a problem that it takes time even for an experienced person to develop a physical equation and construct a model. Further, since the model is simplified, the dynamic characteristics include an error, and an identification method for absorbing the error has not been established. On the other hand, in the statistical model, it is not necessary to develop a physical expression, but since the identification signal is applied to the actual machine and the model is constructed, there is a problem that the analysis cannot be performed until the actual machine can be operated. As described above, if an attempt is made to use a plant from the planning stage, it is necessary to rely on a physical model, but there is a problem that it takes a long time to improve the accuracy after the actual machine starts operation.

[0003]

Problems to be solved in order to construct a high-precision dynamic characteristic model in a short period of time are to generate a dynamic characteristic module unique to the target device and to generate a dynamic characteristic of the entire target by combining the modules. The adjustment is based on the actual model data. Even if the equations that express the dynamic characteristics are known, reusability can be improved if the dynamic characteristic model of the target device is built with the device's attribute parameters and interfaces with other modules incorporated. Becomes lower. It is important to separate the attribute parameters and the interface part of the dynamic characteristics and generalize the model so that it can be specified separately. Even if the dynamic characteristic parameter of the target device is set from the design value, the characteristic equation of the dynamic characteristic is idealized and simplified, so that the accuracy is low as it is. For this reason, parameter adjustment of a dynamic characteristic model based on actual machine data for high accuracy of the dynamic characteristic model is an important issue. The reason is that it is difficult to construct an observer for unobserved state quantities, and it takes a long time to adjust the whole plant model at the same time.
That is, conventionally, the adjustment of the dynamic characteristic model was performed for the entire plant, and thus a long time was required. The reason why the adjustment is required for a long time is that the calculation requires a long time because the static and dynamic characteristics are adjusted while analyzing the entire dynamic characteristic model including the portion where the adjustment is completed. As a means for solving this problem, it is an object to develop a means for decomposing equipment constituting a plant into regions and adjusting a dynamic characteristic model for each region. The reason that adjustment by region cannot be performed is that, in a small region, not all input / output state quantities are measured, and estimation of unobserved state quantities, that is, creation of an observer requires a long time. The characteristics could not be analyzed and adjusted.

[0004] In view of the above circumstances, it is an object of the present invention to shorten the adjustment period of a dynamic characteristic model when adjusting a dynamic characteristic model using actual machine data, and to support dynamic characteristic analysis of a thermal power plant with high accuracy. Is to do.

[0005]

In order to solve the above-mentioned problems, an attribute is set from an attribute of each device constituting a thermal power plant to a module of each device, and an interface section of the module is set based on connection information between the modules. Means for generating a dynamic characteristic module for each device by setting, dynamic characteristic model creating means for automatically constructing a dynamic characteristic model by combining the dynamic characteristic modules for each device, and comparing actual machine data with the result of the dynamic characteristic model Attribute parameter optimizing means for evaluating and correcting / adjusting attribute parameters, and means for estimating a thermal power plant design value from the corrected / adjusted attribute parameters are provided. Here, the dynamic characteristic model creating means extracts a part to be adjusted from the dynamic characteristic model of the whole plant to create a dynamic characteristic partial region model, and estimates input / output values of the dynamic characteristic partial region model from actual machine measured values. An observer is automatically created, and the parameters of the dynamic characteristic model are adjusted using the observer.

[0006]

Embodiments of the present invention will be described below with reference to the drawings. First, FIG. 2 shows a hardware configuration for realizing the present invention. A typical thermal power plant control system includes a central display panel 20 for displaying the status of the whole plant, an operation panel 30 for operation, a control computer 40 for monitoring the state and performance of the plant, and a control device 60 for directly controlling the plant. , Data communication network 5 connecting control computer 40 and control device 60
0, which comprises a data acquisition device 70, monitors the operating state of the thermal power plant 10 to be controlled, and controls so as to output a predetermined output. Since the dynamic characteristic analysis support system of the present invention does not directly control the plant, it is not normally connected to the control device 60. The real machine data is stored in the data recording device 70 or the like,
0 and read it into the off-line computer 90 via a storage medium such as 0. The dynamic characteristic analysis support system of the present invention is implemented by the off-line computer 90. Normally,
It is used offline in this way, but network 5
It is possible to connect directly via 0.

The main problem of the dynamic characteristic analysis support system of the present invention is to support parameter adjustment of a plant control device, and is used at the system planning stage to optimize the equipment configuration and arrangement of the plant. In this dynamic characteristic analysis support system, it is important to construct a highly accurate dynamic characteristic model of a thermal power plant to be controlled. Hereinafter, an embodiment of a thermal power plant dynamic characteristic analysis support system of the present invention will be described with reference to FIG. In FIG. 1, an equipment module class group 100 is a model generalized for each type of equipment constituting a thermal power plant, and is referred to as an equipment module in this embodiment. The basic flow of dynamic characteristic model generation will be described in order. The user selects a class module to be used from the device module class group 100 and specifies its design value and interface information between modules, thereby generating the device module group 110 in which device-specific design data and interfaces are set. I do.

Hereinafter, this equipment module will be described. Here, FIG. 20 shows a schematic water / steam system of a variable pressure once-through type thermal power generation. This operation will be briefly described. The water pressurized by the feedwater pump 500 passes through the feedwater heater 510 and becomes a mixed state of steam and water on the furnace water cooling wall 520. This mixture is separated into steam and water by a brackish water separator 530, and the pressure of the water is increased by a circulation pump 540.
Led to. On the other hand, the steam separated by the steam separator 530 is superheated by the primary superheater 560 and the secondary superheater 570, and is guided to the high-pressure steam turbine 600 via the steam turbine control valve 580. The steam that has worked in the steam turbine 600 is reheated in the reheater 590 and guided to the medium / low pressure steam turbine 610. Work here again and condenser 6
At 30 the water condition is restored. Seawater or air is used as cooling water for condensing water. Work in the steam turbine is converted to rotational energy and converted to electrical output by a generator 620. The above is the outline of the variable-pressure once-through power generation plant. The power generation plant is composed of many devices such as an air supply system, a fuel supply system, and a control device (not shown). In the thermal power plant described above, there are devices that perform the same function in principle, such as the primary superheater 560 and the secondary superheater 570. The equations representing this dynamic characteristic can be considered the same.
What is different is plant design values such as heat transfer area and mass, and input / output values. In the case of simulating dynamic characteristics, in such a case, the model can be efficiently generated by generalizing the model and setting design values and the like later. Since heat exchangers and turbines have different characteristics, naturally different generalized models are required. In the present embodiment, a model generalized in this way is referred to as a class device module group 1
Call it 00. In the example of FIG. 20, a water supply pump,
Heat exchangers, valves, turbines, and generators that generalize furnace water cooling walls, brackish water separators, superheaters, and reheaters are class modules. Examples of the class of the heat exchanger module are shown in FIGS. 4 and 5 (1) and (2).

Hereinafter, the basic structure of the class module will be described. 4 and 5 (1) and (2) show the class modules of the heat exchanger. The module has a cell structure. The first column is a column for describing a block number in the block diagram. The block number is given by four digits.
C means a comment, * is used in combination with END, and indicates that a module is described between * and END. In the second column, the output block variable name of the calculation is described. For comments, this column is used. The third column describes an operation instruction. CONST
Represents a constant, and MULT represents multiplication. The arithmetic elements prepared as standard are shown in FIG. OPT is a special operation element and can be independently set by the user from 1 to 99. Since each coding method is not particularly important, the description is omitted here. 4 and 5, the fourth to sixth columns are input variable names of the operation element, the seventh to tenth columns are constant definition columns, and the twelfth column is used for explanations such as comments. In the model description method as described above, the attributes of the model are described after the attribute designation column, the input of the interface is described after the input interface column, and the output is described after the output interface column. The model element is SCRIPT BEGIN
MODULE and SCRIPT END MODULE
Between the definitions. This SCRIPT B
EGIN MODULE and SCRIPT END MO
A method of using DULE will be described in a model generation method described later. MDL in the variable name is a reserved word, and indicates that the module name is a class. As described above, since a class module group is created, when a device-specific name and design value 200 and interface information are set 1200, a module group 1 for each device is set.
10 can be generated.

Hereinafter, a method of generating the device-specific module group 110 will be described. FIG. 11 shows a detailed flow of the attribute setting unit 1100. The attribute settings are shown in FIG. As shown in FIG. 6, a class module to be used and a device module name to be generated are designated by three characters in the attribute setting table. These three characters are determined for convenience, and the number of characters can be increased. In the example of FIG. 6, the class is a heat exchanger CHE (class heat exch).
An abbreviation for "Anger") and select PP as the module name
1 is specified. This PP1 has no special meaning,
An appropriate name may be described to distinguish the devices. Examples of the class modules of the heat exchanger CHE are as shown in (1) and (2) of FIGS. When CHE is selected, the modules of FIGS. 4 and 5 are read (block 1110). Next, the module name PP1 is read, and the PP specified in FIG.
1 is set (block 1120). Since many names of MDL are used in the class module,
It is convenient to perform batch replacement for PP1. Next is the attribute setting. When the module name is set, the second
Variable name SIPP1, SOP of attribute specified in column
Variable names matching P1, VIPP1, etc. are created. The constant term of the matched variable name is changed to the value specified in FIG. 6 (block 1130).

Next, a method of setting the input / output interface will be described. FIG. 12 shows a detailed flow of the input / output interface setting means 1200. Hereinafter, FIG.
The setting method of the input / output interface will be described with reference to FIG. As shown in FIG. 12, the input interface setting 1210 and the output interface setting 1210
Divided into 20. FIG. 7A shows an input interface designated by the user. The correspondence between the internal variable names used in the module and the input variable names is shown. FIG.
"100 HIMD" is input to the input interface of (1).
LGAIN HGO The first number indicates the block number, the second variable HIMDL indicates the name of the variable used internally, GAIN indicates the coefficient operation, and the next HGO shall be multiplied by 1. A variable name from the outside is written in HGO.
Since the module created this time is PP1, in the first attribute setting, this instruction is "100 HIPP1 GAI
NHGO1. ". That is, MDL
Has been replaced by PP1, so the variable used internally is HIPP1. It can be seen from FIG. 7A that the input variable name is HGOGG1. Therefore, the input variable HGO is replaced with HGOGG1. The results of this setting are shown in FIGS. 9 and 10 (1) and (2). The instruction of block number 100 is "100
HIPP1 GAIN HGOGG1 1. Similarly, in the block number 101, the GI
MDL is replaced by GIPP1 and GS is replaced by GSGG2. The above is the input interface setting method. The output interface is basically set in the same way. FIG. 7B shows the output interface. Unlike the input interface, when the user specifies the device name on the attribute specification sheet (FIG. 6), FIG. 7 (2) is automatically set. When the device name PP1 is also given to the module, the MDL of the class module is all changed to the new module name PP1, and there is no content to be executed again for the output interface. By executing the above processing, a module having the structure shown in FIGS. 9 and 10 (1) and (2) is constructed.

In this embodiment, an example of a heat exchanger is shown. However, in principle, other modules can be produced in the same manner. When actually applying, repeat the processing for several minutes or more of the device to be simulated,
The module group 110 is created.

Next, a method of creating the static / dynamic characteristic model 2000 will be described with reference to FIG. A description will be given from the state where the module group 110 is created by the above-described method. First, the device module 110 to be adjusted is selected. For selection, the name of the device module is specified in the target device module specification table 260. An example of the device module designation table 260 is shown in FIG. Modules are created in the manner described above and are in module group 110.
Therefore, the target module may be specified by specifying the directory containing the module and the module name.
Next, it is necessary to attach an adjustment instruction 2300 to the adjustment model 240 using the adjustment instruction table 270. In order to adjust the model, it is necessary to specify the target variables, the corresponding model variables, and the parameters in the model. The example of FIG. 14 (2) shows an example in which the parameter of the block number (BD #) 5244 is corrected by the integration control. The gain indicates the gain of the integral control, and needs to be set in accordance with the characteristic to be adjusted. In addition, in the integration control for adjustment, an initial value of the integrator is required, and this value indicates that the output is the output of the LMDF01.

Next, a method of creating an observer will be described. It has been described that the adjustment model 240 to be adjusted is created by a combination of device-specific modules. However, in order to adjust the dynamic characteristics, the actual model data
Must be entered. However, not all devices measure the state quantity corresponding to the input of the device. Therefore, it is necessary to estimate an unmeasured input from the measured actual device data 300. This is usually called an observer. There was no generalized technology for creating observers, and it was common for a system developer to create each observer. However, since a reliable method has not been established, development took a long time. In this embodiment, the observer creation is to be automated.

FIG. 16 shows the basic steps for creating an observer. As shown in FIG. 15, the instruction for the dynamic characteristic analysis is made up of input / output and parameters. FIG. 15A shows a basic instruction format of a standard operation element. In FIG. 15 (1), the first column is the block number,
The columns indicate output variables, the third column indicates operation element names (here, called function names), the fourth to sixth columns indicate input variables, and the seventh to tenth indicate constants. In addition, although it has been described that there is a special element also in the above, FIG. 15 (2) shows a format of the special element. FIG. 15 (2)
The first column indicates a block number, the third column indicates a function, an output, an input, or a constant, the fourth to sixth columns indicate an output or input variable, and the seventh to tenth columns indicate a constant. What is important here is that the input and output cells for each block are determined. This enables the automatic design of the observer described below. In the present embodiment, an adjustment target is partially extracted from the dynamic characteristic model of the entire plant, that is, the adjustment model 240 described above, and a dynamic characteristic partial region model is created. The problem is that extracting a part of the whole plant creates variables whose inputs are not determined. This is called an undefined variable in this embodiment. Observer creation creates a model for estimating these undefined variables from actual machine data.

An observer is created in the following procedure. First, an overall dynamic characteristic model is created. The overall dynamic characteristic model is created by the modules included in the module group 110 in FIG. Next, among the variables of the overall model that has been created, an R is added to the beginning of the variable name where the actual device data (measured value) 300 is located (step 2110). Variables preceded by R
This indicates that it can be estimated from the measured values. For example, TS
If 2SH (usually indicating the outlet temperature of the secondary superheater 570) is measured, this TS2SH is converted to RTS2SH.
To change. Next, each block of the overall dynamic characteristic model is checked, and if all inputs are R variables, R is added to the output variables of that block. This means that calculation can be performed using a combination of measured variables. The above process is repeated until there are no more variables with R. The model created by repeating the above is stored as an estimation sheet. Next, the dynamic characteristic model to be adjusted is read, and undefined variables are extracted (step 2120). Since the undefined variable surely exists in the estimation sheet, the corresponding instruction is extracted and added to the target dynamic characteristic model as the observer model 250 (step 2130). An undefined variable is searched for the new dynamic characteristic model thus created, and the presence or absence of the undefined variable is checked (step 2140). If observers can be created for all inputs, there are no undefined variables. In this case, the process ends. However, in many cases, the input variables of the newly added instruction are not defined in one observer creation, and it is necessary to repeat Steps 2120 and 2130 several times. Also,
If the estimated sheet, which is an auxiliary sheet for creating an observer, is not sufficient, a complete observer may not be created. In this case, the same undefined variable as before exists. This check is performed (step 2150), and if the same undefined variable appears, the process ends because an observer cannot be created. In this case, it is necessary to enhance the estimation sheet so that the undefined variables can be estimated. Returning to FIG. 13, by incorporating the observer model 250 formed based on the observer model creation instruction group 280 in the observer model creation 2200 into the adjustment model 240 by the observer installation 2400, the dynamic characteristic model 3
10,320 can be created.

Next, a characteristic adjustment method will be described with reference to FIGS. 3, 17, and 18. FIG. FIG. 3 is a detailed step of the flow from the module group 110 to the dynamic characteristic model 320 shown in FIG. 1, and shows a method of adjusting static characteristics and dynamic characteristics separately as characteristic adjustment. First, the adjustment of the static characteristics is performed using the module group 110. First, a static characteristic adjustment model 2100 is created from the module group 110. This creating method is shown in FIG. Using this, the static characteristics 130
While analyzing 0, the attribute parameter 3100 is optimized so that its characteristic matches the target value or the actual device data 300. FIG. 17 shows an example of the optimization method. The input condition is determined using the actual device data 300 or the target value 300 as the input of the dynamic characteristic model 310. Here, since the target is usually to match the actual machine data, it has the same meaning as the target value. It is also possible to give a numerical value separately as a target value instead of the actual machine data. The output for this input condition, here the exit steam temperature is shown as an example, but a variable to be adjusted is specified. A deviation is set so that the outlet steam temperature becomes the actually measured outlet steam temperature (step 3120), and proportional integral control is performed (step 31).
30) Then, the dynamic characteristics are calculated so that the deviation converges to 0, and the attribute parameters are corrected (step 311).
0). The parameters to be modified are plant efficiency, heat transfer coefficient, pressure drop coefficient. When the static characteristics can be adjusted by the above processing, the static characteristic adjusted module group 1 in FIG.
Store as 20. Next, the adjustment of the dynamic characteristics is performed using the module group 120 for which the static characteristics have been adjusted. The creation of the dynamic characteristic adjustment model 2100 is as shown in FIG. Dynamic characteristic 13 using dynamic characteristic adjustment model 330
00 is analyzed and its attribute parameter 3200 is optimized. The adjustment of the dynamic characteristic is different from the adjustment of the static characteristic, in order to match the transient state of the target model with the actual machine. FIG.
An example of the optimization method will be described with reference to FIG. First, dynamic characteristics are analyzed using the dynamic characteristic model 330 and the actual machine data 300 (step 3210). Next, an evaluation index J indicating the degree of approximation of the transient state is calculated (step 3220). Here we are choosing time integral of the square of the difference between the actual θ output value of the _R and model θ as an evaluation function. This evaluation function can be changed according to the purpose at that time, such as the maximum value of the deviation, the maximum value or the minimum value in the evaluation. Next, it is checked whether the evaluation value is within a target range or whether the evaluation value has reached an optimum value (step 3230). If it is not the optimum value outside the target range, the parameter is corrected (step 3240). The steepest descent method is most commonly used for this parameter calculation. In this method, the parameter dependence (sensitivity for each parameter) of the evaluation function is checked, and the parameter is corrected in the direction of highest sensitivity T → T ′. After modifying the parameters, the model parameters are set to T '(step 3250). If the optimum value has been reached, it is output as a dynamic characteristic model 320 with dynamic characteristics adjusted.

Finally, a method of estimating plant design values [FIG.
Of the plant design value estimating means 4000] will be described. FIG. 19 shows a specific design value estimation method. In the design value estimation, first, the heat transfer area A and the mass M of the heat transfer tube, which are the plant design values, are assumed (step 4010). It is, of course, better to be as accurate as possible, but it need not be entirely accurate. Next, a dynamic characteristic adjustment model is created, static characteristic adjustment is performed, the static characteristic parameter A · α is determined (step 3100), dynamic characteristic adjustment is performed, and the dynamic characteristic parameter M · TM is determined (step 3100). 3200). These two adjustment functions have already been described. Next, it is checked whether or not the optimum value has been reached using the evaluation function implemented in the dynamic characteristic adjustment (step 402).
0). If it is not the optimum value, the heat transfer area A and the mass M are changed by the same method as described in the dynamic characteristic adjustment. If the optimum value has been reached, the heat transfer area A and the mass M at that time are the optimum estimated values.

[0019]

As described above, according to the present invention,
Since the dynamic characteristic model can be created by combining the device modules, the dynamic characteristic model can be easily created. In addition, since an observer for estimating a state quantity that is not observed from actual machine data is automatically created, a highly accurate dynamic characteristic model can be created. In addition, by automatically creating an observer for estimating unobserved state quantities and adjusting the range of the adjustment target, adjustment including the part that has already been adjusted is unnecessary, and the dynamic characteristic model Adjustment period can be shortened. Further, since the design value can be accurately estimated from the actual device data, a high-precision dynamic characteristic simulator can be produced when the actual device data is provided, even if an accurate design value is not provided. Using this high-precision dynamic characteristic simulator, it is possible to improve the dynamic characteristics of the plant and to construct a dynamic characteristic model-based control system even for a plant to which no design value is given.

[Brief description of the drawings]

FIG. 1 shows an embodiment of a thermal power plant dynamic characteristic analysis support system according to the present invention.

FIG. 2 is a diagram showing a hardware configuration for realizing the present invention.

FIG. 3 is a diagram showing detailed steps related to optimization of attribute parameters according to the present invention.

FIG. 4 shows an example of a class module for each device according to the present invention (1).

FIG. 5 shows an example (2) of a class module for each device according to the present invention.

FIG. 6 shows an example of an attribute of a device module according to the present invention.

FIG. 7 shows an example of input and output of the device module of the present invention.

FIG. 8 shows an example of a dynamic characteristic analysis command according to the present invention.

FIG. 9 shows an example (1) of a dynamic characteristic model according to the present invention.

FIG. 10 shows an example (2) of a dynamic characteristic model according to the present invention.

FIG. 11 shows an example of a device module attribute setting unit according to the present invention.

FIG. 12 shows an example of an interface setting means of the present invention.

FIG. 13 shows an example of a dynamic characteristic model creation means by module selection according to the present invention.

FIG. 14 is a diagram illustrating a specification format of a module selection and a specification format of an adjustment circuit according to the present invention.

FIG. 15 is a standard instruction format for dynamic characteristic analysis according to the present invention.

FIG. 16 is a diagram showing a procedure for automatically creating an observer according to the present invention.

FIG. 17 is a configuration diagram of static characteristic adjustment according to the present invention.

FIG. 18 is a basic flowchart of dynamic characteristic adjustment according to the present invention.

FIG. 19 is a basic flowchart of design value estimation according to the present invention.

FIG. 20 is a diagram of a water / steam system of a thermal power plant

[Explanation of symbols]

100: equipment module class group, 110: equipment module group, 120: module group, 200: module name and design value, 300: actual machine data, 310: dynamic characteristic model, 320: dynamic characteristic model, 330: dynamic characteristic model, 1
100: attribute setting means, 1200: interface section setting means, 1300: dynamic characteristic analysis, 2000: static / dynamic characteristic model creation, 2100: static characteristic adjustment model creation, dynamic characteristic adjustment model creation, 2200: observer model creation, 3000: characteristic comparison attribute parameter optimizing means, 3
100: Static characteristic comparison attribute parameter optimizing means, 320
0: Dynamic characteristic comparison attribute parameter optimizing means, 4000 ...
Plant design value estimation means

──────────────────────────────────────────────────続 き Continued on the front page (51) Int.Cl. ⁷ Identification symbol FI Theme coat ゛ (Reference) G06F 17/60 110 G06F 17/60 110 (72) Inventor Yo Osawa 7-2, Omikamachi, Hitachi City, Ibaraki Prefecture No. 1 Inside Hitachi, Ltd. Power and Electricity Development Laboratory (72) Inventor Yukinori Katagiri 7-2, Omika-cho, Hitachi City, Ibaraki Prefecture Inside Hitachi, Ltd. Electric Power and Electricity Development Laboratory (72) Inventor Toru Kimura (2-1) Inventor Tadao Uenaka 2-4-1 Hamamatsucho, Minato-ku, Tokyo Inside Babcock Hitachi, Ltd. (2-1) 5-2-1 Omika-cho, Hitachi City, Ibaraki Prefecture F term (reference) 3G071 AA04 AA08 AB01 BA31 EA02 EA05 EA06 FA03 FA05 FA06 JA02 JA03 5H004 GA30 GB04 JB21 KC28 LA15 MA40

Claims (4)

[Claims] 1. A dynamic characteristic module for each device is generated by setting an attribute to a module for each device from an attribute for each device constituting a thermal power plant, and setting an interface section of the module based on connection information between the modules. Means, a dynamic characteristic model creating means for automatically constructing a dynamic characteristic model by combining the dynamic characteristic modules for each device, and an attribute for comparing and evaluating actual machine data and a result of the dynamic characteristic model, and correcting and adjusting an attribute parameter A thermal power plant dynamic characteristic analysis support system, comprising: parameter optimizing means; and means for estimating a thermal power plant design value from the corrected and adjusted attribute parameters. 2. The dynamic characteristic model creating means according to claim 1, wherein the dynamic characteristic model creating means creates a dynamic characteristic partial area model by partially extracting an adjustment target from the dynamic characteristic model of the whole plant. A thermal power plant dynamic characteristic analysis support system, which automatically creates an observer for estimating input / output values from actual machine measured values and adjusts parameters of the dynamic characteristic model using the observer. 3. The method according to claim 2, wherein in the creation of the observer, a model that can be estimated from actual machine measurement values among variables of the dynamic characteristic model of the entire plant is created as an estimation sheet, and the model is obtained from the dynamic characteristic model to be adjusted. Extract definition variables,
The undefined variable is searched from the estimation sheet, added as an observer model to the dynamic characteristic model to be adjusted, the undefined variable is searched for the newly created dynamic characteristic model, and the presence or absence of the undefined variable is determined. A thermal power plant dynamic characteristic analysis support system characterized by performing checks. 4. The thermal power plant dynamic characteristic according to claim 1, further comprising static characteristic attribute parameter optimizing means for comparing and evaluating actual machine data and a result of the dynamic characteristic model, and correcting and adjusting an attribute parameter. Analysis support system.