JP2003049214A - Analyzing system for operation analysis in steelmaking refining furnace - Google Patents

Abstract

PROBLEM TO BE SOLVED: To perform a operation analysis in a steelmaking refining furnace at high accuracy. SOLUTION: This analysis system is provided with a process control unit 7 which has a control model 15 for simulating the movement of the steelmaking refining furnace 6 and collects the actual information in time series of the steelmaking refining furnace and controls to drive the steelmaking refining furnace by using this actual information and the control model, an analyzing server 8 which is connected with this process control unit through a transmission path 5 and stores the actual information by receiving the actual information transmitted from the process control unit and performs the operating analysis of the steelmaking refining furnace by using this stored actual information, and operating terminals 9 which are connected with the analyzing server through the transmission path and performs a analyzing designation to the analyzing server.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an operation analysis system for a steel refining furnace which uses a computer to analyze the operation of the steel refining furnace in a steelmaking plant.

[0002]

2. Description of the Related Art In order to efficiently operate a steelmaking and refining furnace incorporated in a large-scale steelmaking plant, as shown in FIG.
The operation control of the steelmaking and refining furnace 1 is performed by the process control device 2 including a computer. In this process control device 2, a control model 3 simulating the operation of the steel refining furnace to be controlled is provided.
Is remembered.

The process controller 2 outputs the exhaust gas flow rate, the furnace body temperature, the actual hot metal payout amount, the hot metal component actual result, the molten steel temperature, etc., which are output in time series from the steel refining furnace 1, such as process amounts (control). Quantity). Then, the operation of the steel refining furnace 1 is controlled by using the process amount (control amount) and the control model 3 thus taken in. More specifically, the control model 3
By using, so that each process amount (control amount) output from the steelmaking and refining furnace 1 matches a preset target amount, the amount of input scrap, the amount of input auxiliary material to the steelmaking and refining furnace 1,
The manipulated variables such as ... Are calculated, and the steel refining furnace 1 is operated based on the calculated manipulated variables.

In such a method of operating the steelmaking and refining furnace 1, in order to operate the steelmaking and refining furnace 1 with higher accuracy and higher efficiency, the steelmaking of the control model 3 stored in the process control device 2 is performed. It is necessary to always maintain the simulation accuracy for the refining furnace 1 in the best state. More specifically, a large number of parameters (coefficients) incorporated in the control model 3 are updated to optimum values.

In the identification calculation for adjusting the parameters of the control model 3 to the optimum values, for each process amount item of the output process amount (control amount), each operation amount affecting the corresponding item is obtained and analyzed. Obtain each parameter statistically or statistically. In addition, the configuration of the control model 3 itself may have to be changed. Therefore, in order to identify each parameter of the control model 3 with high accuracy, an enormous amount of data (information) including a process amount and an operation amount and an enormous amount of arithmetic processing are required.

Conventionally, the updating process of the control model 3 set in the process control device 2 has been performed by the host computer 4 which manages the entire steelmaking plant. In addition to the updating process of the control model 3 described above, the host computer 4 also develops a control model that always performs optimal operation, or investigates the cause of abnormality when an operation abnormality occurs.

[0007]

However, as shown in FIG.
As shown in (1), the operation analysis method of the steelmaking and refining furnace for performing the operation analysis of the steelmaking and refining furnace 1 including the update of the control model 3 using the host computer 4 still has the following problems to be improved.

As described above, the host computer 4 collects, retains and centrally manages various information in all steelmaking facilities from hot metal pretreatment in a steelmaking plant to a continuous casting machine. Therefore, the information about the steelmaking and refining furnace 1 held in the host computer 4 is mainly the production management information for controlling the operation of the steelmaking and refining furnace 1 such as the target amount for each controlled object.

As described above, in order to identify the control model 3, there is a large amount of actual results including the operation amount applied from the process control device 2 to the steelmaking and refining furnace 1 and the process amount output from the steelmaking and refining furnace 1. Information is needed. Therefore,
For example, there is a lack of time-series information such as gas flow rate and furnace temperature, and there is information loss such as temperature waveform data obtained by the temperature measurement sampler is not stored. It was difficult to reproduce (simulation) on the desk concerning (driving).

In order to supplement these missing information, it is considered that a recorder is built in the process control device 2, the missing information is printed out on the recording paper by this recorder, and the data is managed by this recording paper. . However, with this method, it is not possible to share this information with other analysts, and it takes a lot of effort to collect information in advance when analyzing past results, and as a result, the operation (operation of the steel smelting furnace 1) There is a concern that the reflection will be delayed and inconvenience may occur in stabilizing the operation and maintaining the high accuracy of the control model 3.

Since it is difficult to reproduce the operating condition on a desk,
If the cause analysis at the time of occurrence of an operation abnormality is insufficient, or if the accuracy calculation and the parameter identification of the control model are performed, the analysis is performed using the actual information including inadequate data, and the cause of the operation abnormality cannot be identified. , The accuracy of the control model could not be improved and maintained.

Although it is possible for the process control device 2 to retain the actual performance information including the temperature waveform data for a long period of time, the original function of the process control device 2 is to make steel using a control model. Since the operation control of the refining furnace 1 is performed, if the operation analysis function including the identification of the control model 3 that requires a large amount of data processing is executed using the process control device 2, the operation control of the steelmaking refining furnace 1 has priority. For this reason, the operation analysis function processing requires a considerable amount of time, and the process analysis device 2 cannot realistically execute the operation analysis work.

As described above, since a large-capacity storage device and a processing device are required to hold the time-series performance information and analyze the operation, the storage device and the processing device are used as the host computer 4. However, providing the process controller 2 with the process controller 2 causes problems in terms of maintainability, expandability, and efficiency.

The present invention has been made in view of such circumstances, and a large amount of time-series performance information generated in the steel refining furnace to be controlled is stored and held in a dedicated analysis server to ensure early operation stability. It is an object of the present invention to provide an operation analysis system of a steel refining furnace that can improve the analysis efficiency by developing and maintaining the control model while maintaining the controllability and high precision control model. .

[0015]

In order to solve the above problems, in the operation analysis system of the steelmaking and refining furnace of the present invention, a control model simulating the operation of the steelmaking and refining furnace is provided. A process control device that collects series performance information and uses this performance information and control model to control the operation of the steel smelting furnace, and is connected to this process control device via a transmission line and transmitted from the process control device. The analysis server that receives and stores the recorded performance information and analyzes the operation of the steel smelting furnace using the stored performance information, and is connected to this analysis server via the transmission line and analyzes the analysis server. An operation terminal for giving instructions is provided.

In the operation analysis system of the steel refining furnace configured as described above, the performance information collected from the steel refining furnace by the process controller is used for controlling the steel refining furnace using the control model. Is transmitted to the analysis server, stored and retained in the analysis server, and used for operation analysis of the steel refining furnace including analysis of the control model.

As described above, by providing the analysis server for storing and storing a large amount of performance information and exclusively performing the operation analysis, the operation analysis of the steelmaking and refining furnace including the analysis of the control model can be performed more efficiently and with higher accuracy. Can be implemented in.

Another aspect of the present invention is that the analysis server in the operation analysis system for a steelmaking and refining furnace according to the above-mentioned invention writes the result information database for storing the result information and the result information received from the process control device to the result information database. A performance information writing means to be inserted, a performance information search means for searching the performance information database for performance information in the range of the analysis instruction designated from the operation terminal, and a parameter of the control model is identified based on the searched performance information. And parameter identification means.

In the analysis server in the steel-refining furnace operation analysis system configured as described above, the performance information received from the process control device is once written in the performance information database, and then the analysis instruction specified by the operation terminal is performed. The performance information in the range is searched to identify the parameters of the control model.

Another aspect of the present invention is that the analysis server in the operation analysis system for steel refining furnace of the above-mentioned invention writes the result information database storing the result information and the result information received from the process control device to the result information database. A performance information writing unit to be inserted, a performance information search unit for searching the performance information database for performance information in the range of the analysis instruction designated from the operation terminal, and a reliability degree calculation unit for calculating the reliability degree of the retrieved performance information. And an accuracy calculation means for calculating the accuracy of the record information from the calculated reliability,
And a parameter identification means for identifying the parameter of the control model based on the calculated reliability frequency and the actual result information.

In the operation analysis system of the steel refining furnace configured as described above, since the reliability and accuracy of the performance information retrieved from the performance information database are evaluated,
The accuracy of operation analysis including parameter identification of control model can be improved.

Another aspect of the present invention is that the reliability frequency calculating means of the analysis server in the operation analysis system for the steel refining furnace of the above-mentioned invention calculates the information value of the equilibrium portion regarded as normal in the time series performance information. The reliability degree of the corresponding performance information is obtained from the data score and the deviation of the data.

Still another aspect of the invention is the operation analysis system for a steelmaking and refining furnace according to the above-mentioned aspect, wherein the performance information includes a plurality of performance information items. Further, the process control device has a plurality of control models corresponding to each control target item. Furthermore, the parameter identification means identifies the parameter of each control model for each control model based on the combination of the predetermined performance information items in the performance information and the identification method.

Still another invention is the operation analysis system for a steelmaking and refining furnace of the above-mentioned invention, and the range of the analysis instruction designated from the operation terminal includes a period, a steel type, and a refining classification.

[0025]

BEST MODE FOR CARRYING OUT THE INVENTION An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of an operation analysis system for a steelmaking and refining furnace according to an embodiment.

For example, with respect to a transmission path 5 such as a LAN, a process control device 7 for controlling the steel refining furnace 6 to be controlled, an analysis server 8, and a plurality of operation terminals 9 operated by an analyst.
Are connected.

An information collecting unit 10, a data buffer 11, an information transmitting unit 12, a communication interface 13, and a parameter updating unit 1 are provided in the process control device 7 composed of a computer.
4, a control model memory 15, a control calculation unit 16, an operation output unit 17, and the like are provided.

The information collecting unit 10 processes the process amount (control) such as the time-series exhaust gas flow rate output from the steel refining furnace 6, gas components, furnace temperature, hot metal payout amount, hot metal component result, molten steel temperature ,. And the operation amount such as the time-series input scrap amount, input auxiliary raw material amount, ..., etc. sent from the operation output unit 17 to the steelmaking and refining furnace 6 by itself.

Furthermore, in addition to the above-mentioned process amount and operation amount, the information collecting unit 10 uses (a) a processing start time, an operator operation intervention time, a processing event such as a timing of feeding an auxiliary raw material or ferroalloy, as performance information. Measurement results information such as the time of occurrence, the molten steel temperature and the measured waveform data when measuring the composition during processing (b) Operation target information such as composition and temperature (c) Processes such as the next process and destination Management information (d) Collects all performance information that can be obtained directly or indirectly over the entire refining process in the steelmaking and refining furnace 6, such as plant alarm information from the upper and lower limit monitoring of each sensor.

The information collecting unit 10 sequentially writes the sequentially collected time series performance information to the data buffer 11, and sends the process amount (control amount) of the collected performance information to the control calculation unit 16. A plurality of control models are stored in the control model memory 15. Each control model is formed by a calculation formula simulating the corresponding operation of the control target item given to itself in the steel refining furnace 6.

Then, the control calculation unit 16 uses the respective control models stored in the control model memory 15 to preset the above-mentioned process amounts (control amounts) output from the steel refining furnace 6 by the operator. The above-mentioned manipulated variables for the steelmaking and refining furnace 1 are calculated and sent to the manipulation output unit 17 so as to match the target amount for each control item. The operation output unit 17 applies each operation amount received from the control calculation unit 16 to the steel refining furnace 6.

The time-series performance information composed of the process amount and the manipulated variable written in the data buffer 11 is real time, and after a certain period of time elapses, every time the operation is completed, when the final component is known, and after the operation is completed. The transmission interface 1 is read by the information transmission unit 12 at the timing of operator data input or the like.
3, transmitted to the analysis server 8 via the transmission path 5.

The parameter updating unit 14 updates the corresponding parameter of each control model stored in the control model memory 15 with each parameter input from the analysis server 8 via the transmission path 5 and the communication interface 13.

In the analysis server 8 composed of a computer,
As shown in FIG. 2, the record information database 20, the record information writing unit 21, the communication interface 22, the record information searching unit 23, the reliability number calculating unit 24, and the reference level table 2
5, a facility reliability frequency table 26, an accuracy calculation unit 27, a parameter identification unit 28, an identification calculation method table 29, and the like are provided.

As shown in FIG. 3, in the performance information database 20, a large number of time-series performance information collected by the process control device 7 includes area numbers, IDs (identification information), and information item names. It is attached and written. Area number 1
The detailed data of the relevant information item is stored in each of the areas 20a to 37.

For example, in the area 20a having the area number 7,
Detailed data 2 of actual results by auxiliary material alloy iron hopper shown in Fig. 4
0b is stored. Further, the detailed data 20b of the molten steel temperature waveform data shown in FIG. 5 is stored in the area 20a of the area number 35. FIG. 6 shows a graphic display 20c of a concrete waveform of the molten steel temperature waveform data at each position in the detailed data 20b shown in FIG.

The reference level table 25 is a table for classifying the reliability levels of the input time-series record information into 11 levels from 0 to 10. Specifically, as shown in FIGS. 7 and 6, a data score and a data deviation value (for example, standard deviation) for calculating an information value of the smoothing portion T that can be regarded as normal in time-series performance information. For each of the combinations, 11 levels of reliability degrees from 0 to 10 are set.

For example, when the data of the smoothing portion T has a score of 6 or more and less than 10 and the deviation value of the data is 0.3 or more and less than 0.5, the reliability number of this time series performance information is "5". Is. Furthermore, the number of data points of the smooth portion T is 20 or more and 25.
When the deviation value of the data is less than 0.1 and less than 0.3, the reliability number of the time-series performance information is “9”. Thus, the smaller the data variation (deviation value) in the smoothing section T, the higher the reliability number of the corresponding record information.

In the equipment reliability number table 26, as shown in FIG. 8, the reliability degrees of each equipment such as analyzers and measuring instruments for detecting each performance information are set. The reliability is appropriately changed by the operator depending on the status of occurrence of abnormality and the status of operating period of each equipment.

In the identification calculation method table 29, as shown in FIG. 9, for each control model corresponding to each control target item, a combination of each performance information item used for parameter identification of the corresponding control model, The identification method of the corresponding parameter is stored. For example, for the control model of "Dynamic model temperature estimation" of No. 2, "Middle molten steel temperature", "Middle CD (carbon solidification) temperature (° C)", "Presence or absence of final cooling material", "End lance height" “A mean value” and “end-term oxygen transfer rate average value” are combined, and “least squares method” is registered as a parameter identification calculation method.

Next, the operation of each part of the analysis server 8 will be described in order. The record information writing unit 21 is used by the process control device 7
Each piece of performance information received from is classified and written in the corresponding area 20a of the performance information database 20 shown in FIG. The record information writing unit 21 allocates the area number, time, or time series number as key data or index data when writing each record information, and can extract at high speed at the time of data search.

Here, each operation terminal 9 connected to the transmission path 5 will be described. The operation terminal 9 is a device as a client operated by the analyst. The analyst sends an analysis instruction specifying the period, steel type, and refining classification from the operation terminal 9 to the analysis server 8. Then, the operation terminal 9 obtains the analysis result from the analysis server 8. Further, the routine data extraction processing routinely used by the analyst can also store the processing content in the operation terminal 9 or the analysis server 8.

In this way, the analyst can freely select the survey result items to be surveyed and analyzed and must be easily and quickly stored in the survey information database 20 in the analysis server 8. You can extract detailed performance information.

The performance information search unit 23 of the analysis server 8 reads the performance information requested by the operation terminal 9 from the performance information database 20, and returns it to the operation terminal 9 which is the source of the search request.

Further, the performance information search unit 2 of the analysis server 8
When receiving the analysis instruction designating the period, the steel type, and the refining classification from the operation terminal 9, the device 3 reads out the performance information within the range of the analysis instruction from the performance information database 20 and sends it to the reliability factor calculation unit 24.

The reliability degree calculation unit 24 includes a reference level table 25 shown in FIG. 7 and an equipment reliability table 2 shown in FIG.
6 is used to obtain the reliability number of each record information retrieved from the record information database 20.

Specifically, in the process of obtaining the information value (data value) of the performance information for which the reliability number is calculated, the number of data points required for the equilibrium portion calculation and the deviation value of the data are obtained. Then, the reliability number corresponding to the obtained data score and data deviation value is read from the reference level table 25 shown in FIG. 7, and the read reliability number is used as the reliability of the corresponding record information.

For example, in the steelmaking and refining furnace 6, in a control model for estimating the molten steel temperature at the end of blowing from the amount of oxygen used until the end of blowing, based on information such as the molten steel temperature measured during blowing, this control model The case where the estimation accuracy of the temperature rise model used in the calculation is calculated will be verified.

In this case, the value of the equilibrium portion T of the molten steel temperature waveform, which is the process of obtaining the molten steel temperature on the way to the initial value of the target value and the molten steel temperature measured at the end of blowing (accuracy calculation comparison source performance information) In the calculation of, the number of data points in the used molten steel waveform data and the deviation value of the waveform data are obtained. Then, the reliability level corresponding to the data score and the deviation value is searched from the reference level table 25 shown in FIG.

In addition to the reliability number in the molten steel temperature performance information, the reliability number in each performance information such as the CD (carbon solidification) temperature waveform related to the molten steel C (carbon) and the dissolved oxygen electromotive force waveform related to the dissolved oxygen concentration The calculation is performed similarly.

Further, the reliability degree calculating section 24 calculates the reliability degree of the equipment for obtaining each record information. For example,
When calculating the accuracy of the control model that calculates the manipulated variable in various states using the process amount such as exhaust gas flow rate, exhaust gas component concentration and acid transfer flow rate, the actual information used in those control models is calculated. Searches for abnormal alarms and equipment maintenance records related to the measuring instruments to be collected near the specified accuracy calculation target. Then, based on these abnormality alarms and equipment maintenance results, the target accuracy content and the reliability number of each measuring instrument unit determined in advance for each measuring instrument (equipment) are read from the equipment reliability table 26 of FIG.

Then, the reliability degree calculation unit 24 adds up all the reliability degrees finally obtained for each piece of actual information and each reliability degree obtained for each equipment (measuring instrument) used to obtain each reliability model. The total (total) reliability number for all performance information required for parameter identification is calculated.

The reliability degree calculation unit 24 sends each performance information, the reliability degree for each performance information, and the total reliability degree to the accuracy calculation unit 27 and the parameter identification unit 28.

The accuracy calculation unit 27 calculates the reliability information of each performance information and the reliability information of each performance information received from the reliability frequency calculation unit 24.
The accuracy of each record information is calculated, and the operation terminal 9 of the analysis request source is calculated.
Send to.

The parameter identifying unit 28 stores the same control model as the control model currently set in the control model memory 15 of the process control unit 7. Then, the parameter identification unit 28 uses the reliability frequency threshold value determined in advance for each control model, and this time, the accuracy calculation unit 27.
Compare with the overall confidence number of the corresponding control model received from. Then, when the reliability frequency received this time is lower than the preset reliability frequency threshold value, even if the parameter is identified, the parameter accuracy cannot be expected to be improved, so the parameter identification processing of the corresponding control model is stopped. In this case, the control model that stopped the parameter identification process and the reason for stopping the parameter identification process are notified to the operation requesting operation terminal 9.

Then, the parameter identifying section 28 carries out an identification process for calculating each parameter included in the control model for the control model whose reliability number exceeds the reliability number threshold value. Specifically, the identification calculation method table 2 of FIG.
A combination of performance information used for parameter identification of the corresponding control model and an identification method are read out from 7. Then, using the performance information included in the read combination out of the performance information input from the reliability frequency calculation unit 24, the identification calculation of the parameter of the corresponding control model is performed by the identification method also read.

The parameter identifying section 28 sends each identified parameter of each control model to the process control device 7 via the communication interface 22 and the transmission path 5, and
It is sent to the operation terminal 9 that is the source of the analysis request.

As a result, each parameter of each control model stored in the control model memory 15 of the process control device 7 is updated to the latest parameter identified by the analysis server 8. Further, the analyst can confirm the updated parameters on the operation terminal 9.

In the operation analysis system of the steel-refining furnace thus constructed, the process controller 7 collects time-series detailed execution information including the process amount and the operation amount generated in the steel-refining furnace 6. In addition to being used by the process control device 7 for controlling the steelmaking and refining furnace 6 using the control model, the process control device 7 is transmitted to the analysis server 8 and the analysis server 8
It is stored and held in the performance information database 20 formed therein. Then, in the analysis server 8, the record information in the range of the analysis instruction designated by the operation terminal 9 is searched from the record information database 20, and the control model memory 15 is searched.
The parameters of each control model stored in are identified.

As described above, by providing the analysis server 8 for storing and storing a large amount of performance information and exclusively performing the operation analysis, the operation analysis of the steelmaking and refining furnace 6 including the analysis of the control model can be performed more efficiently and more. It can be performed with high accuracy.

Furthermore, the reliability degree and accuracy of the performance information retrieved from the performance information database 20 are evaluated, and the parameters of the control model are identified using only the performance information having high reliability. Therefore, the accuracy of operation analysis including parameter identification of the control model can be improved.

[0062]

As described above, in the operation analysis system of the steel-refining furnace of the present invention, the series detailed operation information generated in the steel-refining furnace is collected by the process controller and then stored in the analysis server. The analysis server, which is stored and stored, performs operation analysis for the steel refining furnace to be controlled, including parameter identification of the control model.

Therefore, it is possible to stabilize the operation at an early stage and maintain the control model with high accuracy, and to improve the analysis efficiency in developing and maintaining the control model.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an operation analysis system of a steelmaking and refining furnace according to an embodiment of the present invention.

FIG. 2 is a block diagram showing a schematic configuration of an analysis server incorporated in the operation analysis system of the steel refining furnace.

FIG. 3 is a diagram showing storage contents of a performance information database formed in the analysis server.

FIG. 4 is a diagram showing an example of performance information stored in the performance information database.

FIG. 5 is a diagram showing an example of performance information stored in the same performance information database.

FIG. 6 is a diagram showing a graphic display of an example of performance information also stored in the performance information database.

FIG. 7 is a diagram showing stored contents of a reference level table formed in the analysis server.

FIG. 8 is a diagram showing stored contents of a facility reliability frequency table formed in the analysis server.

FIG. 9 is a diagram showing stored contents of an identification calculation method table formed in the analysis server.

FIG. 10 is a diagram for explaining a conventional operation analysis method for a steel refining furnace.

[Explanation of symbols]

6 ... Steelmaking refining furnace 7 ... Process control device 8 ... Analysis server 9 ... Operation terminal 10 ... Information collection section 13, 22 ... Communication interface 14 ... Parameter update unit 15 ... Control model memory 16 ... Control calculation unit 20 ... Achievement information database 21 ... Achievement information writing section 23 ... Achievement information search section 24 ... Reliability frequency calculation unit 25 ... Standard level table 26 ... Equipment reliability table 27 ... Accuracy calculation unit 28 ... Parameter identification unit 29 ... Identification calculation method table

Continued front page    F-term (reference) 4K070 AB20 DA10                 5H004 GA27 GA28 GB03 KC28 KC45                       MA38 MA52                 5H223 AA01 BB01 CC01 DD05 DD07                       DD09 FF05

Claims (6)

[Claims] 1. A steelmaking refining furnace having a control model simulating the operation of the steelmaking refining furnace, collecting time-series performance information of the steelmaking refining furnace, and using the performance information and the control model to control the steelmaking refining furnace. A process control device for operation control, and a process control device connected to the process control device via a transmission line, receiving and storing the performance information transmitted from the process control device, and using the stored performance information. A steel refining furnace comprising: an analysis server for performing an operation analysis of the steel refining furnace; and an operation terminal connected to the analysis server via the transmission path for giving an analysis instruction to the analysis server. Operational analysis system. 2. The analysis server, a performance information database for storing the performance information, performance information writing means for writing performance information received from the process control device to the performance information database, and designated from the operation terminal. A performance information search means for searching the performance information database for performance information in the range of the specified analysis instruction, and a parameter identification means for identifying a parameter of the control model based on the searched performance information. The operation analysis system of the steel refining furnace according to claim 1. 3. The analysis server, a performance information database for storing the performance information, a performance information writing unit for writing performance information received from the process control device to the performance information database, and designated from the operation terminal. A performance information search means for searching the performance information database for performance information in the range of the analyzed analysis instruction, a reliability degree calculation means for calculating a reliability degree of the retrieved performance information, and performance information based on the calculated reliability degree. 2. The operation of the steelmaking and refining furnace according to claim 1, further comprising: accuracy calculation means for calculating the accuracy of the control model, and parameter identification means for identifying the parameter of the control model based on the calculated reliability number and performance information. Analysis system. 4. The reliability level calculation means calculates the reliability level of the corresponding performance information from the data score and the data deviation for calculating the information value of the equilibrium part regarded as normal in the time-series performance information. 4. The method according to claim 3, wherein
Operation analysis system for steel refining furnace described. 5. The performance information includes a plurality of performance information items, the process control device has a plurality of control models corresponding to the respective control target items, and the parameter identification unit is for each of the control models. The steelmaking refining according to any one of claims 2 to 4, characterized in that the parameters of each control model are identified based on a combination and identification method of predetermined performance information items in the performance information. Furnace operation analysis system. 6. The operation analysis system for a steel refining furnace according to claim 1, wherein the range of the analysis instruction designated from the operation terminal includes a period, a steel type, and a refining classification. .