KR20190078335A - Apparatus for controlling harmonic heat in converter and computer readable recording medium - Google Patents

Abstract

According to an embodiment of the present invention, an apparatus for controlling heat mixing in a converter comprises: a converter heat mixing guide management unit to receive work data of a converter, apply the work data to a plurality of machine learning models in parallel, and select a portion of a plurality of pieces of converter heat mixing guide information generated in parallel by the plurality of machine learning models; a controller driving management unit to drive a controller for controlling heat mixing of the converter based on selected converter heat mixing guide information among the plurality of pieces of converter heat mixing guide information; a work result analysis management unit to receive feedback information in accordance with control of the controller, generate hit rate information of the selected converter heat mixing guide information based on the feedback information, and generate learning compensation weight information based on the hit rate information; and a relearning management unit to generate relearning information of at least a portion of the plurality of pieces of converter heat mixing guide information based on the learning compensation weight information.

Description

TECHNICAL FIELD [0001] The present invention relates to a converter thermal mixing control device and a computer-

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a computer-readable recording medium on which a converter thermal mixing control apparatus and method are recorded.

Generally, the molten steel temperature in the conversion furnace is one of the important operational indexes directly related to the quality of steel products. The converter thermal mixing model used to control the molten steel temperature in the converter operation is a model for controlling the temperature of converter molten steel during the conversion operation and is a model for controlling the target temperature to remove impurities in the molten steel.

Conventional transformer thermal mixing models are based on the metallurgical model and utilize the empirical results of previous operations or controlled using a one - dimensional statistical linear regression model. For both of these methods, we developed a model for each class by dividing the model into several classes by classifying the fixed class for the variables affecting some operations such as steel type. The development of this type of model has the problem that there is no class for a new steel grade at the time of new steel grade change and environment change. Also, as the operating environment changes little by little, the developed model can not be used after a certain period of time, and a new model has to be manually developed.

As the development speed of new types of steel is getting faster in accordance with the changing market conditions in recent years, there is a need to develop models that can be applied to new steel types.

Patent Registration No. 10-0328937

There is provided a computer-readable recording medium having recorded thereon an apparatus and method for controlling thermal transfer blending according to an embodiment of the present invention.

A converter thermal batching control apparatus according to an embodiment of the present invention includes a plurality of machine learning models that are configured to receive operation data of a converter and apply the operation data to a plurality of machine learning models in parallel, A converter thermal mixing guide management unit for selecting a part of the plurality of converter thermal mixing guide information; A controller driving management section for driving a controller for controlling the thermal combination of the converters based on the selected converter thermal mixing guide information among the plurality of converter thermal mixing guide information; An operation result analysis management unit that receives feedback information under the control of the controller, generates hit ratio information of the selected guide bus sequence information based on the feedback information, and generates learning compensation weight information based on the hit ratio information; A re-learning management unit for generating re-learning information of at least a part of the plurality of transducer string combination guide information based on the learning compensation weight information; . ≪ / RTI >

For example, the re-learning management unit may include a learning data configuration management module that determines whether to apply the learning compensation weight information to the re-learning information and manages the configuration of the learning data; A re-learning time calculation module for analyzing trends of learning compensation weights respectively corresponding to the plurality of machine learning models and calculating an application time of the re-learning information based on the analysis results; And a re-learning driving module for each AI model that applies the re-learning information to at least a part of the plurality of machine learning models at an application time determined by the re-learning time calculation module; . ≪ / RTI >

For example, the re-learning management unit may further include a final guide selection reinforcement learning module for generating learning information of a selection model for selecting information on a transfer column combination guide of the converter thermal combination guide management unit based on the learning compensation weight information can do.

For example, the converter thermal mixing guide management unit may include: a business data reception module for receiving the business data of the converter; A work environment information integration module for generating work environment information based on the operation data; A multiple AI model drive management module for managing parallel drive of the plurality of machine learning models based on the operation data; An AI model calculation managing unit for managing calculations for generating information on the number of guide lines of each of the plurality of machine learning models; And a final guide selection AI model drive module for selecting, based on the operating environment information, a part of a plurality of converter train combination guide information respectively generated by the plurality of machine learning models; . ≪ / RTI >

For example, the controller drive management unit may include an oxygen input amount control module that controls an amount of oxygen supplied to the converter based on the selected guide heat composition guide information; A coolant input amount control module for controlling an input amount of the coolant to be introduced into the converter based on the selection converter thermal mixing guide information; And a heat transfer agent injection amount control module for controlling an amount of the heat transfer agent injected into the converter based on the selection information of the heat transfer formulation guide; . ≪ / RTI >

A computer-readable recording medium according to an embodiment of the present invention is a computer-readable storage medium for storing a plurality of machine learning models, the computer-readable recording medium having a plurality of machine learning models, Selecting some of the plurality of converter thermal mixing guide information; Controlling the thermal combination of the converter based on the selected converter thermal mixing guide information among the plurality of converter thermal mixing guide information; Receiving feedback information according to the control on the converter, generating hit ratio information of the selected guide bus sequence information based on the feedback information, and generating learning compensation weight information based on the hit ratio information; And generating re-learning information of at least a part of the plurality of converter array composition guide information based on the learning compensation weight information; The instructions executable by the processor to perform the method comprising the steps of:

The converter thermo combination control device according to an embodiment of the present invention can adaptively control the heat transfer composition for various steel products and various converter operating environments and can be used for a new steel product or a new converter operating environment, The suitability of the control can be improved.

FIG. 1 is a view illustrating a converter thermal blending control apparatus according to an embodiment of the present invention.
FIG. 2 is a diagram specifically illustrating a converter thermal mixing guide management unit and a controller drive management unit. FIG.
FIG. 3 is a diagram illustrating a final guide selection AI model drive module of the converter heat mixing guide management unit in detail.
4 is a flowchart illustrating an operation procedure of the converter thermal mixing guide management unit.
FIG. 5 is a diagram specifically illustrating the operation result analysis management unit.
6 is a diagram specifically illustrating a re-learning management unit.

DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, the embodiments of the present invention can be modified into various other forms, and the scope of the present invention is not limited to the embodiments described below. The shape and the size of the elements in the drawings may be exaggerated for clarity and the same elements are denoted by the same reference numerals in the drawings.

FIG. 1 is a view illustrating a converter thermal blending control apparatus according to an embodiment of the present invention.

1, the converter thermal batch composition control apparatus 100 according to an embodiment of the present invention includes a converter thermal mixing guide management unit 110, a controller drive management unit 120, a operation result analysis management unit 130, And a management unit 140.

The converter thermal compounding guide management unit 110 receives from the steelmaking process computer 10 operating data (for example, molten steel temperature data, carbon concentration data, oxygen concentration data, slag amount data, converter status data, And a method of operating the machine tool, the method comprising: receiving a plurality of machine learning models and component and mass data of the scrap, target steel product data, target operating environment data, weather data, and other process progress data, (For example, decrease of oxygen input, stop of coolant injection, delay of injection of heat transfer agent, change of injection order of coolant and heat exchanger, etc.) generated in parallel by the model can be selected.

The controller drive management unit 120 may drive a controller for controlling the thermal combination of the converters based on the selected guide train combination guide information among the plurality of guide train combination guide information. Here, the controller may include at least a part of the oxygen injector 210, the coolant injector 220, and the injector injector 230.

The operation result analysis management unit 130 receives the feedback information under the control of the controller, generates the hit ratio information of the selected switch train array guide information based on the feedback information, generates the learning compensation weight information based on the hit ratio information can do.

The re-learning management unit 140 can generate re-learning information of at least a part of the plurality of convergence string combination guide information based on the learning compensation weight information. Thus, the plurality of machine learning models can adaptively learn each of the patterns of the plurality of converters' operation data.

Therefore, the converter thermal batch control apparatus 100 according to an embodiment of the present invention can adaptively control the converter heat composition for various steel products and various converter operating environments, and can be applied to a new steel product or a new converter operating environment It is possible to improve the suitability of the converter thermo-mixing control.

FIG. 2 is a diagram specifically illustrating a converter thermal mixing guide management unit and a controller drive management unit. FIG.

2, the converter thermal mixing guide managing unit 110a includes a operation data receiving module 111 for receiving the operation data of the converter, and a control unit 110 for operating environment information (for example, target steel product data, A plurality of AI model drive management modules 113 for managing the parallel drive of the plurality of machine learning models based on the operation data, An AI model calculation management unit (114) that manages calculations for generating information on converter threading guide information of each machine learning model of the plurality of machine learning models; And a final guide selection AI model drive module 115 to be selected based on environment information.

For example, when the operational environment information corresponds to the target steel product data, the final guide selection AI model driving module 115 determines whether or not the target steel product belongs to the first type, Selects heat combination guide information, selects the second converter heat mixing guide information generated by the second machine learning model when the target steel product belongs to the second type, and selects the second converter heat mixing guide information when the target steel product is a new steel product 3 < / RTI > machine learning model.

For example, when the operating environment information corresponds to the target operating environment data, the final guide selection AI model driving module 115 determines whether or not the fourth converter train mixing guide information generated by the fourth machine learning model And the fifth converter train combination guide information generated by the fifth machine learning model can be selected when the converter operation time is long.

For example, when the operation environment information corresponds to the data of the measuring apparatus or the data of the measuring apparatus, the final guide selection AI model driving module 115 generates, by the sixth machine learning model when the molten steel temperature is out of the reference temperature range The seventh converter thermal mixing guide information generated by the seventh machine learning model can be selected when the slag amount is out of the reference range.

For example, the final guide selection AI model driving module 115 selects the eighth transmission line array guide information generated by the eighth machine learning model when the variable of the operation environment information is large, The ninth converter thermoforming guide information generated by the ninth machine learning model can be selected.

As such, selection criteria for the machine learning model of the final guide selection AI model drive module 115 can be variously determined.

2, the controller drive management unit 120a includes an oxygen input amount control module 121 for controlling the input amount of oxygen to be supplied to the converter based on the selection converter thermal mixing guide information, A coolant input amount control module (122) for controlling an input amount of the coolant to be input to the converter based on the information, a heat transfer agent amount control module (122) for controlling an input amount of the heat transfer agent injected into the converter Module 123, as shown in FIG.

That is, the controller drive management unit 120a can drive the various controllers in combination based on the selected converter thermal mixing guide information.

FIG. 3 is a diagram illustrating a final guide selection AI model drive module of the converter heat mixing guide management unit in detail.

3, the final guide selection AI model driving module receives a plurality of converter train combination guide information (AI model result value) and operating environment information (operating environment variable), selects the selected converter train mixing guide information Can be output.

For example, the converter thermal management guide management unit may have a neural network structure including an input layer (2N node), a hidden layer, a full connection layer (N node), and an output layer. Here, the 2N node, the hidden layer, the N node, and the output layer may each include at least one node, and the node may include status information. The state information of some nodes among the 2N nodes of the input layer may be influenced by the input of the plurality of converter array guide information (AI model result value), and the state information of the remaining nodes may be influenced by the operation environment information The input may be affected. The state information of the node of the hidden layer may be determined according to the combination of the state information of the node of the input layer (2N node) connected to the node of the hidden layer, and the state information of the node of the fully connected layer (N node) N nodes of the hidden layer connected to the node of the hidden layer, and the state information of the nodes of the output layer may be determined according to the combination of the state information of the nodes of the N nodes connected to the nodes of the output layer ≪ / RTI >

That is, the final guide selection AI model drive module can progressively learn the selection criteria for the machine learning model. Thus, the roles of a plurality of machine learning models can be efficiently distributed.

On the other hand, the multi-AI model drive management module and / or the AI model calculation management module as well as the converter thermal mix guide management part may have a neural network structure including an input layer, a hidden layer, a complete connection layer and an output layer have. Therefore, the machine learning model and the selection model of the converter management guide can be learned in parallel, and can have a hierarchical structure formed by the upper selection model and the lower machine learning model.

4 is a flowchart illustrating an operation procedure of the converter thermal mixing guide management unit.

Referring to FIG. 4, the converter thermal management guide management unit receives the transfer operation data (environment information) before the start of the bypass refining operation (S110) and transmits the operation data (environmental information) to a plurality of machine learning models (S120). Then, a guide value calculation for each model is performed by driving a plurality of machine learning models (multiple AI models) (S130), and a selection model using optimal computation information and operating environment information as input variables A usage guide model) is driven (S140), and the optimal guide value selection among the calculation information for each model is performed (S150).

FIG. 5 is a diagram specifically illustrating the operation result analysis management unit.

5, the operation result analysis management unit 130a includes a model result hit ratio analysis module 131 (see FIG. 5) that receives feedback information under control of the controller and generates hit ratio information of the selected guide train combination guide information based on the feedback information And a learning compensation weighting module 132 for generating learning compensation weight information based on the hit ratio information.

6 is a diagram specifically illustrating a re-learning management unit.

6, the re-learning management unit 140a includes a learning data configuration management module 141 that determines whether to apply the learning compensation weight information to the re-learning information and manages the configuration of the learning data, A re-learning time calculation module 142 for analyzing trends of learning compensation weights respectively corresponding to the models and calculating the application time of the re-learning information based on the analysis results; A re-learning driving module 143 for applying the re-learning information to at least a part of the plurality of machine learning models at the time of application; And a final guide selection reinforcement learning module 144 for generating learning information based on the learning compensation weight information.

For example, the re-learning time calculation module 142 can advance the application time of the re-learning information when the average hit rate of the selection guide thermal combination guide information of the plurality of machine learning models becomes lower than the reference hit ratio range.

For example, the final guide selection enhancement learning module 144 can generate learning information of the selected model when the average hit ratio of the selected guide thermal combination guide information of a plurality of machine learning models becomes lower than the reference hit ratio range. Thus, the selection criteria for the machine learning model of the converter thermal mixing guide management section can be learned more efficiently. Thus, the roles of the plurality of machine learning models can be more efficiently distributed.

Meanwhile, the machine learning model disclosed herein can be implemented by a Classification And Regression Trees (CART) algorithm, a Linear Discriminant Analysis (LDA) algorithm, a K-Nearest Neighbor (KNN) algorithm, and the like. Further, a plurality of machine learning models can be implemented according to algorithms having different characteristics.

Meanwhile, the converter thermal batch control device disclosed in the present specification can be applied to a personal computer, a server computer, a handheld or laptop device, a mobile device (a mobile phone, a PDA, a media player, etc.), a multiprocessor system, (E.g., peripheral component interconnect (PCI), USB, firmware (IEEE 1394), etc.), which may be implemented in a distributed computing environment, A keyboard, a mouse, a pen, a voice input device, a touch input device, an infrared camera, a video input device, etc.), a storage unit (e.g., a volatile memory ), Non-volatile memory (e.g., ROM, flash memory, etc.) or a combination thereof, or magnetic storage, optical storage, Such as a central processing unit (CPU), a graphics processing unit (GPU), a microprocessor, an application specific integrated circuit (ASIC), field programmable gate arrays (FPGA) , A network interface card (NIC), an integrated network interface, a radio frequency transmitter / receiver, an infrared port, a USB connection, etc.).

Meanwhile, a computer-readable recording medium according to an embodiment of the present invention can be read by a converter thermal mixing control device, and the converter thermal mixing control device can execute a command recorded on the computer readable recording medium.

In the meantime, the term " part " used in this embodiment means a hardware component such as a software or a field-programmable gate array (FPGA) or an ASIC. However, 'part' is not meant to be limited to software or hardware. &Quot; to " may be configured to reside on an addressable storage medium and may be configured to play one or more processors. Thus, by way of example, 'parts' may refer to components such as software components, object-oriented software components, class components and task components, and processes, functions, , Subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays, and variables. The functions provided in the components and components may be further combined with a smaller number of components and components or further components and components. In addition, components and components may be implemented to reproduce one or more CPUs in a device or system.

The present invention is not limited to the above-described embodiments and the accompanying drawings. It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined by the appended claims. It will be self-evident.

100, 100a: converter thermal mixing control device
110, 110a: Conversion heat mixing guide management section
120, 120a: controller drive management section
130, 130a: Operation result analysis management unit
140, 140a: re-learning management unit

Claims (6)

A plurality of machine learning models, and a plurality of machine learning models, the machine learning models including a plurality of machine learning models, a plurality of machine learning models, and a plurality of machine learning models, A guide management unit;
A controller driving management section for driving a controller for controlling the thermal combination of the converters based on the selected converter thermal mixing guide information among the plurality of converter thermal mixing guide information;
An operation result analysis management unit that receives feedback information under the control of the controller, generates hit ratio information of the selected guide bus sequence information based on the feedback information, and generates learning compensation weight information based on the hit ratio information;
A re-learning management unit for generating re-learning information of at least a part of the plurality of transducer string combination guide information based on the learning compensation weight information; And a control device for controlling the temperature of the heat exchanger.
The apparatus according to claim 1, wherein the re-
A learning data configuration management module that determines whether to apply the learning compensation weight information to the re-learning information and manages the configuration of learning data;
A re-learning time calculation module for analyzing trends of learning compensation weights respectively corresponding to the plurality of machine learning models and calculating an application time of the re-learning information based on the analysis results; And
Learning drive module for applying the re-learning information to at least a part of the plurality of machine learning models at an application time determined by the re-learning time calculation module; And a control device for controlling the temperature of the heat exchanger.
The apparatus according to claim 2, wherein the re-
Further comprising a final guide selection reinforcement learning module for generating learning information of a selection model for selecting the information on the travel thermoforming guide of the converter thermal combination guide management section based on the learning compensation weight information.
2. The apparatus according to claim 1,
A operation data receiving module for receiving the operation data of the converter;
A work environment information integration module for generating work environment information based on the operation data;
A multiple AI model drive management module for managing parallel drive of the plurality of machine learning models based on the operation data;
An AI model calculation managing unit for managing calculations for generating information on the number of guide lines of each of the plurality of machine learning models; And
A final guide selection AI model drive module that selects, based on the operating environment information, a part of a plurality of converter train combination guide information generated by each of the plurality of machine learning models; And a control device for controlling the temperature of the heat exchanger.
The apparatus according to claim 1,
An oxygen input amount control module for controlling an input amount of oxygen to be supplied to the converter based on the selection converter thermal mixing guide information;
A coolant input amount control module for controlling an input amount of the coolant to be introduced into the converter based on the selection converter thermal mixing guide information; And
A heat transfer agent injection amount control module for controlling the amount of the heat transfer agent injected into the converter based on the selection information of the heat transfer formulation guide; And a control device for controlling the temperature of the heat exchanger.
The method comprising: receiving operational data of a converter, applying the operational data to a plurality of machine learning models in parallel, and selecting a portion of a plurality of converter train composition guide information generated in parallel by the plurality of machine learning models;
Controlling the thermal combination of the converter based on the selected converter thermal mixing guide information among the plurality of converter thermal mixing guide information;
Receiving feedback information according to the control on the converter, generating hit ratio information of the selected guide bus sequence information based on the feedback information, and generating learning compensation weight information based on the hit ratio information; And
Generating re-learning information of at least a part of the plurality of conversion string combination guide information based on the learning compensation weight information; Readable medium having stored thereon instructions executable by a processor to perform the method.

Images