Claims:We claim:
1. A method (400) to calculating mechanical properties of hot rolled steel sheet along length from hot strip mill (100), the method comprises:
receiving (401) process parameter, chemical composition, and run-out table (ROT) snapshot data from a plurality of sensor (102) of the hot strip mill (100);
cleaning (402) the received process parameter, chemical composition, and run-out table (ROT) snapshot data to remove inconsistent data;
processing (403) the cleaned data for normalization and storing the processed data into data (310);
calculating (404) F6 speed, coil length, and ROT snapshot data based on the processed data and storing calculated F6 speed, coil length, and ROT snapshot data into dictionaries, D1, D2, D3;
generating (405) three cooling sequence zones based on the calculated F6 speed data, the coil length data, and the ROT snapshot data; and
calculating (406) mechanical properties the hot rolled steel sheet based on the three cooling sequence zones, the process parameter and chemical composition using deep neural network methodology.

2. The method (400) as claimed in claim 1, wherein the method further comprises displaying (407) calculated mechanical properties of the hot rolled steel sheet on a display device (200).

3. The method (400) as claimed in claim 1, wherein the calculating (407) further comprises:
normalizing the process parameter, chemical composition , and the three cooling sequences in standard deviation normalization method to obtain normalized values; and
processing the normalized values by a multilayer neural network model (605) using a Feed Forward Back Propagation (FFBP) algorithm to create relation between the process parameter, chemical composition, and the three cooling sequences and the mechanical properties.

4. The method (4000) as claimed in claim 1, wherein the mechanical properties are Yield strength (YS), Ultimate Tensile Strength (UTS), and percentage elongation (%el).
5. The method (400) as claimed in claim 1, wherein the process parameters are Coil Thickness, Coil Width, Coil weight, SDOT, Slab Retention Time, Slab Soaking Time, FRT Average, CT Average, F6 Speed Average, RMET Average.
6. The method (400) as claimed in claim 1, wherein the chemical composition parameters are Al, C, Cr, Cu, Mn, N, Nb, Ti, V, Ni, P,S, B Si,

7. A prediction and control system (300) to predict mechanical properties of hot rolled steel sheet coming from a hot strip mill (100), the prediction and control system (300) comprises:
a processor (301) coupled with the memory (303), the memory (303) received process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data from a plurality of sensors (102);
a cleaning module (305), coupled with the processor (301) and the memory (303), cleans the received process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data to remove inconsistent data;
a ROT processing module (306), coupled with the processor (301) and the memory (303), process the cleaned data and creates three data dictionaries (D1, D2, D3), the ROT processing module (306) calculates F6 speed, coil length, and ROT Snapshot data and stores the calculated data into the three data dictionaries (D1, D2, D3);
a sequence generation module (307) generates three cooling sequences for the ROT based on the calculated F6 speed, coil length, and ROT Snapshot data; and
a modeling module (308) calculates the mechanical properties of the hot rolled steel sheet based on the three cooling sequences, the process parameter, and the chemical composition using deep neural network methodology.

8. The prediction and control system (300) as claimed in claim 7, wherein the prediction and control system (300) further comprises a display module (309) to display mechanical properties of the hot rolled steel sheet on a display unit (200).

9. The prediction and control system (300) as claimed in claim 7, wherein the modeling module (308) coupled with processor (301) to
normalize the process parameter, the chemical composition, and the three cooling sequences in standard deviation normalization method to obtain normalized values; and
process the normalized values by a multilayer deep neural network model (605) using a ADAM optimization algorithm to create relation between the process parameter, the chemical composition, and the three cooling sequences and the mechanical properties.

10. The prediction and control system (300) as claimed in claim 7, wherein the mechanical properties are Yield strength (YS), Ultimate Tensile Strength (UTS), and percentage elongation (%el).

11. The prediction and control system (300) as claimed in claim 7, wherein the process parameters are Coil Thickness, Coil Width, Coil weight, SDOT, Slab Retention Time, Slab Soaking Time, FRT Average, CT Average, F6 Speed Average, RMET Average.

12. The prediction and control system (300) as claimed in claim 7, wherein the compositional parameters are Al, C, Cr, Cu, Mn, N, Nb, Ti, V, Ni, P,S, B, Si.

13. A hot strip mill (700) for producing hot rolled steel sheet with defined mechanical properties, the hot strip mill (700) comprises:
a Run-Out Table (ROT) (701) having a plurality of nozzles at top and bottom side to cool the hot rolled steel sheet;
a prediction and control system (300) coupled with the Run-Out Table (ROT) (701) for controlling cooling of the hot rolled steel sheet, the prediction and control system (800) comprises:
a processor (301) coupled with the memory (303), the memory (303) received process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data from a plurality of sensors (102);
a cleaning module (305), coupled with the processor (301) and the memory (303), cleans the received process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data to remove inconsistent data;
a ROT processing module (306), coupled with the processor (301) and the memory (303), process the cleaned data and creates three data dictionaries (D1, D2, D3), the ROT processing module (306) calculates F6 speed, coil length, and ROT Snapshot data and stores the calculated data into the three data dictionaries (D1, D2, D3);
a sequence generation module (307) generates three cooling sequences for the ROT based on the calculated F6 speed, coil length, and ROT Snapshot data;
a modeling module (308) calculates the mechanical properties of the hot rolled steel sheet based on the three cooling sequences, process parameter data, chemical composition data using deep neural network methodology; and
a updating module (320) to update the three cooling sequences of the ROT to modify the mechanical properties of the hot rolled steel sheet.

14. A method (800) for controlling mechanical properties of hot rolled steel sheet in a hot strip mill, the method comprising:
receiving (801) and implementing generated cooling sequences from a prediction and control system (300) into run-out table (ROT) of the hot strip mill;
controlling (802) cooling temperature of the ROT based on the cooling sequences;
receiving (803) updating inputs via display unit (200) from plurality of input buttons to update mechanical properties of the hot rolled steel sheet by changing the cooling sequence; and
controlling (804) the ROT based on the updated cooling sequences.
, Description:SYSTEM AND METHOD FOR PREDICTING AND CONTROLLING MECHANICAL PROPERTIES USING ARTIFICIAL NEURAL NETWORKS
FIELD OF INVENTION:
[001] The present subject matter described herein, relates to a system and a method for predicting mechanical properties of hot rolled coil in a hot strip mill and, in particular, to a system and a method for predicting mechanical properties along the length of the hot rolled coil in the hot strip mill using deep learning techniques of Artificial Neural Networks. Further, the present subject matter describes a system which guides run-out table (ROT) of the hot strip mill for obtaining required hot rolled coil with mechanical properties.
BACKGROUND AND PRIOR ART AND PROBLEM IN PRIOR ART:
[002] Steel is one of the most important green materials consumed worldwide due to its high tensile strength, low cost and recyclable life-span. Its versatility makes it a crucial material for construction and engineering purposes. According to the end product requirements, steel products can be broadly classified into bars, rods and sheets.
[003] In the hot strip mill the slabs are heated and soaked at an elevated temperature approx 1200° C in the reheat furnace, and are subjected to subsequent reduction in the roughing and finishing mill. All reductions are completed in the austenitic phase (˜890° C) before the strip enters the run-out table (ROT). The strips are cooled down to ˜600° C by using laminar water jets installed on the ROT, before being cooled in the down coiler.
[004] Existing practice for determining the mechanical properties of a hot rolled coil from the hot strip mill is to perform tensile tests of the specimen in a tensile testing machine. The specimen used for tensile testing is prepared from a cut-out sample of the outer wrap of the coil produced in the mill. The cut-out sample is then machined to prepare the specimen for tensile testing.
[005] From the stress-strain graph generated from the tensile testing machine, the mechanical properties like Yield Strength (YS), Ultimate Tensile Strengths (UTS) and Percentage Elongation (EL) can be obtained. The test results are posted in the Test Certificate (TC) before the coil is shipped to the customer.
[006] One drawback of this existing method is that there is only one sample per coil that can be tested since the coil cannot be cut from the middle for taking the samples.
[007] Another technical problem with the existing approach is to cut a sample from each coil to determine the mechanical properties. Cutting of samples from the coils (sheets) create wastage problem along with labor cost.
[008] Another technical problem associated with the hot rolled coil is that the variation in properties along length of the hot rolled coil because the sample from the outer wrap of the coil does not represent the entire length of the coil. Therefore, there is need for a system and a method which can determine properties along length of the hot rolled coil.
[009] The mechanical properties of hot-rolled steels (coils) are governed by a complex interaction of various process parameters and chemical composition. The convoluted non-linear relationship between these parameters makes it difficult to efficiently control and predict the required mechanical properties (Yield Strength, Ultimate Tensile Strength and Percentage Elongation) via traditional mathematical models. Thus, the challenge is to efficiently model these non-linear features and identify an optimum setting of the parameter values for achieving the target specifications.
[0010] The output mechanical properties of hot-rolled steel have a high correlation with the finish rolling temperature (FRT) and the coiling temperature (CT). While the FRT depends on the type of cooling process, the CT is chosen based on the desired properties of the output. Fig.1a shows the effect of CT on the micro-structure of steel. Generally, in the hot strip mill specifies an FRT of about 810-925°C and a CT of about 550-700°C. However, for a particular combination of FRT and CT, it is observed that the mechanical properties also depend on the way the coil was cooled, that is, the cooling rate (CR) of the coil as shown in the Fig. 1b. For a hot-strip mill, the cooling rate is decided by the run-out table (ROT). The ROT is equipped with laminar water-cooling facility consisting of approximately 24 top nozzles and approximately 24 bottom nozzles. The combination in which these top and bottom nozzles cool the coil directly affects the mechanical properties of the coil. Therefore, the ROT plays an important role in the mechanical properties of the hot rolled coils (sheets).
[0011] Therefore, there is a need for a system which can determine or predict mechanical properties of the hot rolled sheet without any mechanical or stress testing of a sample. Further, there is a need for a system which can operate the hot strip mill, specifically, ROT in predefined manner to control the mechanical properties of the sheet in a predefined range along length of the sheet. The inventors of the present invention develop a system based on artificial neural networks to solve the above mentioned technical problems.
OBJECTS OF THE INVENTION:
[0012] The principal objective of the present invention is to provide a system and a method for estimating or predicting mechanical properties of hot rolled sheets segment wise using deep neural network.
[0013] Another object of the present invention is to provide a system and a method for controlling mechanical properties of hot rolled sheets segment wise in hot strip mill.
[0014] Another object of the present invention is to provide a system and a method for controlling operations of run out table (ROT) for obtaining required mechanical properties of the hot rolled steel sheet.
[0015] Yet another object of the present invention is to provide a system and a method for predicting mechanical properties of steel sheet along the length of the steel coil by modeling the process parameters, chemical composition, and ROT cooling sequence as input features.
SUMMARY OF THE INVENTION:
[0016] The present subject matter relates a method and system for predicting mechanical properties of hot rolled steel sheet based on artificial neural network. The prediction and control system comprise a processor which is coupled with the memory. The memory receives process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data from a plurality of sensors located at various locations in the hot strip mill. Further, the system has a cleaning module coupled with the processor and the memory to clean the received process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data to remove inconsistent data. A ROT processing module processes the cleaned data and creates three data dictionaries (D1, D2, D3). Further, the ROT processing module calculates F6 speed, coil length, and ROT Snapshot data and stores the calculated data into the three data dictionaries (D1, D2, D3). Upon calculation, sequence generation module of the system generates three cooling sequences for the ROT based on the calculated F6 speed, coil length, and ROT Snapshot data. Further the system has a modeling module which calculates the mechanical properties of the hot rolled steel sheet based on the three cooling sequences, the process parameter, and the chemical composition using deep neural network methodology. The modeling module coupled with processor to normalize the process parameter, the chemical parameter, and the three cooling sequences in standard deviation normalization method to obtain normalized values and process the normalized values by a multilayer neural network model using a Feed Forward Back Propagation (FFBP) algorithm to create relation between the process parameter, the chemical composition, and the three cooling sequences and the mechanical properties. The system has a display module to display mechanical properties of the hot rolled steel sheet on a display unit.
[0017] In another embodiment of the present subject matter, a method for predicting mechanical properties of hot rolled steel sheet is described.
[0018] In another embodiment, the present subject matter explained a hot strip mill with the prediction and control system for controlling production of hot rolled steel sheets with predefined range of mechanical properties.
[0019] In yet another embodiment of the present subject matter explained updating input parameters to modify mechanical properties of the hot rolled steel sheets.
[0020] In order to further understand the characteristics and technical contents of the present subject matter, a description relating thereto will be made with reference to the accompanying drawings. However, the drawings are illustrative only but not used to limit scope of the present subject matter.
BRIEF DESCRIPTION OF THE DRAWINGS
[0021] It is to be noted, however, that the appended drawings illustrate only typical embodiments of the present subject matter and are therefore not to be considered for limiting of its scope, for the invention may admit to other equally effective embodiments. The detailed description is described with reference to the accompanying figures. In the figures, a reference number identifies the figure in which the reference number first appears. The same numbers are used throughout the figures to reference like features and components. Some embodiments of system or methods or structure in accordance with embodiments of the present subject matter are now described, by way of example, and with reference to the accompanying figures, in which:
[0022] Fig. 1a illustrates effect of coiling temperature on the micro-structure of hot rolled sheet;
[0023] Fig. 2 illustrates effect of cooling rate of the hot rolled coil;
[0024] Fig. 3 illustrates system block diagram of prediction and control system for predicting mechanical properties of hot rolled steel sheets, in accordance with an embodiment of the present subject matter;
[0025] Fig. 4 illustrates method for predicting or estimating mechanical properties of hot rolled steel sheet in hot strip mill, in accordance with an embodiment of the present subject matter;
[0026] Fig. 5a illustrates calculation of valid duration of hot rolled steel sheet on ROT in hot strip mill, in accordance with an embodiment of the present subject matter;
[0027] Fig. 5b illustrates reference direction for cooling sequence generation in the ROT of the hot strip mill, in accordance with an embodiment of the present subject matter;
[0028] Fig. 6a illustrates architecture of the modeling module having deep learning techniques of artificial neural networks for calculating mechanical properties, in accordance with an embodiment of the present subject matter;
[0029] Fig. 6b illustrates multilayer neural network type model having more than one hidden layer for training the data, in accordance with an embodiment of the present subject matter;
[0030] Fig. 7 illustrates implementation of prediction and controlling system on the hot strip mill for producing hot rolled steel sheets with desired mechanical properties, in accordance with an embodiment of the present subject matter; and
[0031] Fig. 8 illustrates implementation method of the prediction and controlling system on the hot strip mill for producing hot rolled steel sheets with desired mechanical properties, in accordance with an embodiment of the present subject matter.
[0032] The figures depict embodiments of the present subject matter for the purposes of illustration only. A person skilled in the art will easily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.
DESCRIPTION OF THE PREFERRED EMBODIMENTS:
[0033] The present subject matter relates a method and system for predicting mechanical properties of hot rolled steel sheet based on artificial neural network. The prediction and control system comprise a processor which is coupled with the memory. The memory receives process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data from a plurality of sensors located at various locations in the hot strip mill. Further, the system has a cleaning module coupled with the processor and the memory to clean the received process parameter data, chemical composition data, and Run-Out Table (ROT) snapshot data to remove inconsistent data. A ROT processing module processes the cleaned data and creates three data dictionaries (D1, D2, D3). Further, the ROT processing module calculates F6 speed, coil length, and ROT Snapshot data and stores the calculated data into the three data dictionaries (D1, D2, D3). Upon calculation, sequence generation module of the system generates three cooling sequences for the ROT based on the calculated F6 speed, coil length, and ROT Snapshot data. Further the system has a modeling module which calculates the mechanical properties of the hot rolled steel sheet based on the three cooling sequences, the process parameter, and the chemical composition using deep neural network methodology. The modeling module coupled with processor to normalize the process parameter, the chemical composition, and the three cooling sequences in standard deviation normalization method to obtain normalized values and process the normalized values by a deep neural network model using a Feed Forward Back Propagation (FFBP) algorithm to create relation between the process parameter, the chemical composition, and the three cooling sequences and the mechanical properties. The system has a display module to display mechanical properties of the hot rolled steel sheet on a display unit.
[0034] In another embodiment of the present subject matter, a method for predicting mechanical properties of hot rolled steel sheet is described.
[0035] In another embodiment, the present subject matter explained a hot strip mill with the prediction and control system for controlling production of hot rolled steel sheets with predefined range of mechanical properties.
[0036] In yet another embodiment of the present subject matter explained updating input parameters to modify mechanical properties of the hot rolled steel sheets.
[0037] As explained in the background, the ROT cooling sequence for a particular coil affects the mechanical properties of that coil. Since the coil spreads over a relatively long distance, it has to be taken into consideration that the cooling pattern may vary along the length of the coil (from its head-end to its tail-end). This difference in cooling pattern whilst having the same process and chemical composition, results in a difference in the mechanical properties along the length of the coil. Hence, mechanical properties like the Yield Strength (LYS), Ultimate Tensile Strength (UTS) and the Percentage Elongation (EL) at one end of the coil will the considerably different from the LYS, UTS, and EL at the other end of the coil. Linear regression, like all mathematical regression analysis, is not deemed accurate for modeling steel data with noise as it inevitably introduces a significant degree of error. Furthermore, it is seen that high accuracy of the training set does not necessarily guarantee coherently predicted values for all test sets. The Artificial Neural Networks (ANNs) and other Deep Learning models can use distinctive training algorithms which are able to overcome these difficulties and successfully relate these sundry non-linear inputs to the mechanical output properties of steel. The present subject matter uses the Deep learning models based on the artificial neural networks to calculate mechanical properties based on input parameters along length of the hot rolled steel sheet.
[0038] It should be noted that the description and figures merely illustrate the principles of the present subject matter. It should be appreciated by those skilled in the art that conception and specific embodiment disclosed may be readily utilized as a basis for modifying or designing other structures for carrying out the same purposes of the present subject matter. It should also be appreciated by those skilled in the art that by devising various arrangements that, although not explicitly described or shown herein, embody the principles of the present subject matter and are included within its spirit and scope. Furthermore, all examples recited herein are principally intended expressly to be for pedagogical purposes to aid the reader in understanding the principles of the present subject matter and the concepts contributed by the inventor(s) to furthering the art, and are to be construed as being without limitation to such specifically recited examples and conditions. The novel features which are believed to be characteristic of the present subject matter, both as to its organization and method of operation, together with further objects and advantages will be better understood from the following description when considered in connection with the accompanying figures.
[0039] The terms “comprises”, “comprising”, or any other variations thereof used in the disclosure, are intended to cover a non-exclusive inclusion, such that a device, system, assembly that comprises a list of components does not include only those components but may include other components not expressly listed or inherent to such system, or assembly, or device. In other words, one or more elements in a system or device proceeded by “comprises… a” does not, without more constraints, preclude the existence of other elements or additional elements in the system or device.
[0040] These and other advantages of the present subject matter would be described in greater detail with reference to the following figures. It should be noted that the description merely illustrates the principles of the present subject matter. It will thus be appreciated that those skilled in the art will be able to devise various arrangements that, although not explicitly described herein, embody the principles of the present subject matter and are included within its scope.
[0041] Fig. 1 illustrates schematic block diagram of prediction and control system 300 for prediction or estimating mechanical properties of hot rolled steel sheet from hot strip mill, in accordance with an embodiment of the present subject matter. The present prediction and control system 300 (can be referred as system 300 hereinafter) can be used with the hot strip mill to control mechanical properties of new hot rolled steel sheets. The system 300 includes one or more processor(s) 301, interface(s) 302, memory 303 coupled to the processor 301, modules 304, and data 310. The processor(s) 301, may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, logic circuitries, and/or any devices manipulate signals based on operational instructions. Among other capabilities, the processor(s) 301 is configured to fetch and execute computer-readable instructions stored in the memory 303. Further processor 301 is a hardware device which communicates with the other software and hardware and processes the data and provides the results.
[0042] The interface(s) 302 may include a variety of software and hardware interfaces, for example, interfaces for peripheral device(s), such as a keyboard, a mouse, an external memory, and display device. Further, the interfaces 302 is used to receive inputs from a plurality of sensors 102 and to send output to the hot strip mill 100 and other display devices 200 connected with the system 300. Further, the interfaces 302 may facilitate multiple communications within a wide variety of protocol types including, operating system to application communication, inter process communication, client server communication, etc.
[0043] The memory 303 can include any computer-readable medium known in the art including, for example, volatile memory, such as static random access memory (SRAM) and dynamic random access memory (DRAM), and/or non-volatile memory, such as read only memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and magnetic tapes.
[0044] Further modules 304 and data 309 may be coupled with the processor(s) 301. The modules 304, amongst other things, include routines, programs, objects, components, data structures, etc., which perform particular tasks or implement particular data types. The modules 304 may also be implemented as, signal processor(s), state machine(s), logic circuitries, and/or any other device or component that manipulate signals based on operational instructions. In another aspect of the present subject matter, the modules 304 may be computer-readable instructions which, when executed by a processor/processing unit, perform any of the described functionalities. The machine-readable instructions may be stored on an electronic memory device, hard disk, optical disk or other machine-readable storage medium or non-transitory medium. These instructions are executed by the processor 301 along with input values to determine the desire output value.
[0045] In an implementation, the module(s) 304 include data cleaning module 305, ROT processing module 306, sequence generation module 307, modeling module 308, and display module 309.
[0046] The data 310 includes cleaned data 311, ROT processed data 312, parameter data 313, sequence data 314, modeled data 315 and other data 316. The other data 316 amongst other things, may serve as a repository for storing data that is processed, received, or generated as a result of the execution of one or more modules in the module(s) 304. Although the data 310 is shown internal to the system 300, it may be understood that the data 310 can reside in an external repository, such as server, which may be coupled to the system 300 over network. The system 300 may communicate with the external repository through the interface(s) 302 to obtain information from the data 310.
[0047] As explained above, the system 300 estimates mechanical properties of the hot rolled steel sheets obtained from the hot strip mill based on the process parameters, chemical composition, and cooling sequence in the run out table (ROT). The present system 300 is coupled with a hot strip mill 100 via network, such as LAN, WIFI, WLAN, GSM. The hot strip mill 100 has a plurality of sensors 102 which are placed at various locations in the hot strip mill 100. The plurality of sensors 102 being configured to sense values corresponding to parameters of Length, width, thickness, weight, slab drop out temperature, soaking time, slab retention time, roughing mill exit temperature, finish rolling temperature, coiling temperature, last stand speed in finishing mill and composition of liquid steel, header opening and closing snapshot at ROT and forwarded the sensed values to the memory 303 via interface 302 to the system 300. Furthermore, all parameter data are stored in the parameter data 313 of the system 300.
[0048] In one implementation, the data cleaning module 305 receives the parameter data 313 and the ROT data for pre-processing which helps in smoothening the noisy, inconsistent data, easy data transformation. The optimal data cleaning and preprocessing helps the modeling module 308 to work efficiently and effectively to predict the mechanical properties. The cleaning module 305 prepares two clean data sheets one for the process parameters and chemical composition along with process parameters and second for ROT snapshot data. The data cleaning module 305 cleans and transforms the ROT data and stores the ROT snapshot in predefined format. The ROT snapshot data has information about the opening and closing of the nozzles in the ROT along with time duration.
[0049] In another implementation, the ROT processing module 306 receives the clean data sheets from the cleaned data 311 and processes the clean data sheets of the process parameters, chemical composition and the ROT snapshot data sheet. The ROT processing module 306 creates three data dictionaries or data storage or data structures for F6 speed, sheet length or coil length, and ROT snapshot data. The plurality of sensor 102 provides speed of the rollers in the hot strip mill 100 which provide information about coil speed, i.e., F6 speed in the hot strip mill. In the present embodiment, the sensors particularly provide information about the coil speed, i.e., F6 speed in the ROT of the hot strip mill. The plurality of sensors provides information about the coil length to the ROT processing module 306. The ROT processing module 306 determines length of the coil in the ROT of the hot strip mill (as shown in the figure 5a, 5b). The ROT processing module 306 associate the different data with speed and length. For example, the ROT processing module 306 determines the head, middle, and tail of the steel sheet in the ROT. Based on the location whether head, middle, or tail the ROT processing module 306 associates opening and closing of nozzles of the ROT. The opening and closing of the nozzle of the ROT depends on the speed of the sheet.
[0050] The sequence generation module 307 combines all three data dictionaries and creates at least three cooling sequences for the particular section of the steel sheet which is inside the ROT. Further, the sequence generation module 307 receives the inputs from the plurality of sensors 102 to generate the cooling sequences for the last hot rolled steel sheet. This generated cooling sequence can be used as an input value along with other process parameters and chemical composition to control mechanical properties of the steel sheet in required range (as explained in figure 7 and 8).
[0051] In another embodiment, the Modeling module 309 receive process parameter data 313, chemical composition 313, and ROT cooling sequence data 314 from the data 310 and applies deep neural network (DNN) techniques to predict or estimate or calculate mechanical properties of the steel sheet along length of the steel sheet.
[0052] Referring to figure 6a and 6b to understand working of modeling module 309 in the present system to predict mechanical properties based on process, chemical composition and ROT cooling sequence data. As shown in the figure 6b, the Artificial Neural Network (ANN) is a mesh of artificial neurons (fundamental building blocks of the computational model) that are inspired by the highly-flexible neurons present in the human brain. The entire model imitates the flow of knowledge gathering, information processing, and information retrieval methods observed by the human nervous system. For computational purposes, a bias is set to each neuron and each pair of neurons is assigned a weight. This set of weights and biases lays the foundation of the neural network and defines the function which relates the input parameters to the output targets. After performing a linear combination of inputs, weights and biases, each neuron may or may not perform a small non-linear transformation. Mathematically, the action of an artificial neuron, j, can be described as follows:
oj = F ( S ( wijxi + bi ) ) …………………………………………Equation 1
[0053] where, oj is the output weight-age of neuron j; xi and bi are the input weight-age and bias value of the previous neuron i; and wij is the weight between neurons i and j. The non-linear transformation function F(ReLU) is usually an abstract representation of the rate of action potential firing in the neurons. The mechanical properties of the hot rolled steel sheet are predicted by using the Feed Forward with Back Propagation Algorithm for training the data sets. The Feed Forward with Back Propagation Algorithm has three layers, such as input layer, hidden layer, and output layer. In this case the no of hidden layer is 6 with varying nodes. In the algorithm, the Back Propagation supersedes its predecessor, the perceptron, in its ability to train hidden layers thereby escaping the restricted capabilities of single layer networks. Fig 6b illustrates architecture of a fully connected Artificial Neural Network with two hidden layers. In the present architecture as explained in above equation 1, the multiplication of the input values and the weights are summed and then processed by the activation function. Then the function is processed, and the signal is passed to the output layers. The output layer compares the predicted value against the measured value and calculates the error. Then, training of the data sets is repeated until the error is minimized. Data sets are trained using the adam optimization algorithm.
[0054] Referring to figure 6a, a standard deviation normalize device 604 collects or receives a plurality of process parameters 601, a plurality of chemical composition parameter 602 and three cooling sequences 603 of the hot rolled steel sheet. The received data is normalized to obtain values in the range of -1 to 1. After normalization, the normalized values are processed by the multi layer neural network model 605 as shown in the figure 6a. The model 605 uses the Feed forward back propagation algorithm to create relationship between the input values and the output values by training. At the end of the training, input values of the hot rolled steel sheet are given to the model 605 and the output values are simulated. The simulated output values are mechanical properties 606 of the hot rolled steel sheet along length of the sheet. The present model predicts the mechanical properties of the hot rolled steel sheet using process parameters, chemical composition and cooling sequence pattern. Therefore, there is no need to cut a sample from the sheet to determine the mechanical properties. Further, the present model determines the mechanical properties along length of the sheet coil. Therefore, mechanical property of each segment is available. Based on the mechanical properties, the hot rolled steel sheet can be accepted or rejected.
[0055] In an implementation, the modeling module 308 stores all data received from the model 605 into the modeled data 315 from where this data can be accessed by the display module 309 to represent the data in to required format.
[0056] In an implementation, the system 300 has display module 309 which is connected to a display unit 200 via interface 302 to display the predicted mechanical properties of the hot rolled steel sheet. The display unit 200 can be a computer screen or any other display device. The display module 309 displays the mechanical properties in different format as desired. The display module 309 displays the mechanical properties in to table format, in graph format or any particular report format. The display module 309 stores all the results into the other data 316 of the memory 303 so that it can be retrieved later course of action. The display module 309 is coupled with a printing device to print the report of the mechanical properties.
[0057] In other implementation, the system 300 may be communicatively coupled with the server over the network. The network may be internet, Intranet, LAN, WAN, WLAN, Wi-Fi and any other mode of communication. The system 300 may stores the data on the server. The system 300 may be implemented on the local machine at the location of the hot strip mill 100. The server may be any communicating device and any operating machine which can store and perform actions. The person skilled in the art is very well aware about the server communication structure.
[0058] Fig. 2 illustrates a method for predicting mechanical properties of the hot rolled steel sheet in the hot strip mill, in accordance with an embodiment of the present subject matter. Present methodology is not only limited to the described steps and combination of the steps. A person skilled in the art will understand the steps of the method and can modify the steps and achieve the end results.
[0059] The method may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, functions, etc., that perform functions or implement particular abstract data types. The method may also be practiced in a distributed computing environment where functions are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, computer executable instructions may be located in both local and remote computer storage media, including memory storage devices.
[0060] At step 401, plurality of process parameters, a plurality of chemical composition parameters and a ROT snapshot data is received. The system receives the data and arranges the data is prescribed format for further processing. The data is received from a plurality of sensors placed at various locations in the hot strip mill.
[0061] At step 402, the system 300 cleans the received parameter data and the chemical composition data. The system also cleans the ROT snapshot data also. The cleaning of data helps the model 605 to work efficiently and effectively with less data training time. The cleaning of data removes noisy and inconsistent data which leads to efficient data transformation and reduction in transformation time.
[0062] At step 403, the cleaned data is processed. The process and chemical composition data is processed separately from the ROT snapshot data. After cleaning and processing the ROT snapshot data, the certain attributes of the data are calculated in order to store in the processed ROT snapshot data in desirable format. After processing of the data, the data can be used for further processing.
[0063] At step 404, the system 300 calculates the F6 speed which is speed of the sheet in the hot strip mill, coil length and opening and closing of top and bottom nozzle in the ROT based on the processed data. The plurality of sensors provides information about the rolling speed and coil length. Further, the plurality of sensors provided in the ROT provides information about opening and closing of the top and bottom nozzle along with time duration. The system correlates the data with each other for further processing. The calculated data is stored in three different data dictionaries, such as D1 for F6 speed, D2 for coil length, and D3 for ROT snapshot. The ROT snapshot data has information about opening and closing of the top and bottom nozzles in the ROT.
[0064] At step 405, the three data dictionaries D1, D2, D3 are used to generate cooling sequences for a particular section of the steel sheet.
[0065] Referring to figure 5a and 5b, sheet length in the ROT is calculated with three reference points, such as head end, tail end and middle end. The length of the ROT is available with the system. When a sheet enters the ROT, cooling of the sheet starts with valid time duration, when the sheet leaves the ROT, the cooling time ends. There is invalid time which does not require further cooling of the sheet. The invalid time is calculated and based on the invalid time, the nozzles of the ROT are controlled.
[0066] Invalid time = tail end length/ f6 speed……………………. Equation 2
[0067] As shown in the figure 5b, three reference points, such as tail end element length, head end element length, and middle section element length. The opening and closing of the top and bottom nozzles along with duration of opening is controlled by the generated cooling sequence to obtain desired mechanical properties. Further, to predict mechanical properties of the last rolled steel sheet, the opening and closing of the top and bottom nozzles are captured and feed to the system for calculating the mechanical properties of the hot rolled steel sheet.
[0068] In the ROT, steel sheet is divided into three zones for three cooling sequence such as head zone, tail zone and middle zone. Based on the three zones, the cooling sequence, i.e., opening and closing of top and bottom nozzle is generated based on required temperature for cooling. For example, the head zone, i.e., first entering into the ROT requires more cooling as it has highest temperature then middle zone. Tail end zone requires less cooling as compared to head and middle zone.
[0069] At step 406, the generated cooling sequence along with process parameters and chemical composition of the steel sheet are given to model for predicting the mechanical properties, such as YS, UTS and %EL. The model (as shown in the figure 6a, 6b) applies standard normalization to normalize all input values. After normalizing the input values, the normalized values are processed by the multilayer neural network model (as shown in the figure 6b) to train the input values. The model has feed forward back propagation algorithm to create relationship between the input values and output values by training. After training in the model, the model provides simulated output values, i.e., mechanical values of the fresh steel sheet. Based on the experiments, it has been observed that the simulated mechanical values obtained from the present system are similar to the measured values of the steel sheet.
[0070] At step 408, the calculated or predicted mechanical properties are displayed on a display device. The predicted mechanical properties can be shown in any format, such as table, graphs.
[0071] Figure 7 illustrates implementation of the prediction and controlling system 800 with the hot strip mill 700 to control the mechanical properties of the hot rolled steel sheet into desirable range. The prediction and controlling system 800 is implemented with the hot strip mill 700 along with real time monitoring and modifying features which allows user to feed cooling sequence into the hot strip mill 700 and receiving inputs of the mechanical properties in real time. Based on the mechanical properties, the user is allowed to modified cooling sequence in accordance with process parameters and chemical composition to change the mechanical properties of the steel sheet. The prediction and controlling system 300 generates a ROT cooling sequence (as explained in figure 3 and 4) for each segment of the steel sheet to control opening and closing of ROT nozzle in the ROT. Further, the ROT mechanical design and speed of steel sheet in the ROT is also given as input in the system to control the opening and closing of the ROT nozzles for effective and required cooling of the hot rolled steel sheet. The generated ROT cooling sequence, ROT design, sheet length, speed of sheet in the ROT along with process parameters and chemical compositions are given as input parameters to the prediction and controlling system 300 to control mechanical properties of the steel sheet in an effective way. Further, the prediction and controlling system 300 has an updating module 320 along with other modules are described in the figure 3. The updating module 320 is coupled with the display module 309 to receive user inputs from the display device about updates. Based on the requirements, the user can update the cooling sequence by the updating module 320. Upon receiving the user inputs, the updating module 320 updates the cooling sequence and display resulting mechanical properties to the user. Once the user accepts the resulting mechanical properties, the system 300 applies the updated cooling sequence into the hot strip mill, specifically, in ROT.
[0072] Figure 8 illustrates working of the hot strip mill in combination with the cooling sequence generated by the prediction and controlling system, in accordance with an embodiment of the present subject matter.
[0073] At step 801, the generated cooling sequence along with process parameters and chemical composition is implemented in the system of the hot strip mill.
[0074] At step 802, the system controls cooling temperature of the hot rolled steel sheet in the ROT based on the generated sequence.
[0075] At step 803, the user can update the cooling sequence by a plurality of input buttons. After receiving the inputs from the user, the system calculates the mechanical properties based on the updated cooling sequence, process parameters and chemical composition and displays the predicted mechanical properties to the user for his approval. Further, the mechanical properties may be varied by changing other parameters, such as process parameters and chemical compositions.
[0076] At step 804, the system controls cooling temperature of the hot rolled steel sheet based on the updated cooling sequence of the ROT.
[0077] Due to the above implementation of the system along with the hot strip mill, the rejection of the hot rolled steel sheets can be minimized by controlling property variation along the length of the steel sheet. Further, the present system can be used as guidance system for development of new product where non-linear cooling profile is employed.
[0078] The present system predicts mechanical properties similar to measured values without cutting any sample from the sheet and without performing any stress-strain experiment. Further, the present system is efficient and effective in predicting or estimating or calculating mechanical properties of the hot rolled steel sheet without variation. Further, the present system is based on the training of the input values; therefore, it becomes more accurate and effective with more number of input values. When data sets are near to accurate data sets, the training of data sets is efficient and effective with less error of margin.
[0079] It will be understood by those within the art that, in general, terms used herein, and especially in the appended claims (e.g., bodies of the appended claims) are generally intended as “open” terms (e.g., the term “including” should be interpreted as “including but not limited to,” the term “having” should be interpreted as “having at least,” the term “includes” should be interpreted as “includes but is not limited to,” etc.). It will be further understood by those within the art that if a specific number of an introduced claim recitation is intended, such an intent will be explicitly recited in the claim, and in the absence of such recitation no such intent is present. For example, as an aid to understanding, the following appended claims may contain usage of the introductory phrases “at least one” and “one or more” to introduce claim recitations. However, the use of such phrases should not be construed to imply that the introduction of a claim recitation by the indefinite articles “a” or “an” limits any particular claim containing such introduced claim recitation to inventions containing only one such recitation, even when the same claim includes the introductory phrases “one or more” or “at least one” and indefinite articles such as “a” or “an” (e.g., “a” and/or “an” should typically be interpreted to mean “at least one” or “one or more”); the same holds true for the use of definite articles used to introduce claim recitations. In addition, even if a specific number of an introduced claim recitation is explicitly recited, those skilled in the art will recognize that such recitation should typically be interpreted to mean at least the recited number (e.g., the bare recitation of “two recitations,” without other modifiers, typically means at least two recitations, or two or more recitations). Furthermore, in those instances where a convention analogous to “at least one of A, B, and C, etc.” is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., “a system having at least one of A, B, and C” would include but not be limited to systems that have A alone, B alone, C alone, A and B together, A and C together, B and C together, and/or A, B, and C together, etc.). In those instances where a convention analogous to “at least one of A, B, or C, etc.” is used, in general such a construction is intended in the sense one having skill in the art would understand the convention (e.g., “a system having at least one of A, B, or C” would include but not be limited to systems that have A alone, B alone, C alone, A and B together, A and C together, B and C together, and/or A, B, and C together, etc.). It will be further understood by those within the art that virtually any disjunctive word and/or phrase presenting two or more alternative terms, whether in the description, claims, or drawings, should be understood to contemplate the possibilities of including one of the terms, either of the terms, or both terms. For example, the phrase “A or B” will be understood to include the possibilities of “A” or “B” or “A and B.”
[0080] Although implementations for the method and the system for predicting mechanical properties of the hot rolled steel sheet have been described in language specific to structural features and/or method, it is to be understood that the present subject matter is not necessarily limited to the specific features described. Rather, the specific features and methods are disclosed as embodiments for the present subject matter. Numerous modifications and adaptations of the system/device/structure of the present invention will be apparent to those skilled in the art, and thus it is intended by the appended claims to cover all such modifications and adaptations which fall within the scope of the present subject matter.