FIELD OF THE INVENTION
The present invention relates to a system for predicting strip temperature in finishing
stands of hot strip mill involving hybrid thermal model based on input data relating to
temperature at the entry of first finish stand and other input parameters affecting heat
transfer. More particularly, the present invention relates to a system involving
mathematical-Statistical-Artificial Neural Network(ANN) based hybrid thermal model for
finishing stands of hot strip mill for online acquisition, analysis, control and display of
process data accurately independent of operator's intervention to predict temperature
profile along strip length before and after each stand of finishing stands of hot strip mill.
The system software is adapted to receive the input data from pyrometer located at the
entry of first finishing stand and exit of last finish stand and other data relating to roll
gap setting, speed setting and diameter of work rolls via a PLC system connected with
OPC network to a computer system termed Process Work Station(PWS). The output is
displayed to operator in another computer system viz Operator Workstation(OWS).
Operator controls the finishing stand setting based on this feedback. Importantly, the
hybrid thermal model based system of the invention is adapted for accurate temperature
prediction wherein mathematical model is based on there equations derived from
fundamental principles of heat transfer considering heat transfer in finishing stands
consisting of radiation heat transfer, convection heat transfer with roll cooling water,
conduction heat loss to work rolls and heat generation due to plastic deformation. A
linear regression based statistical model has been developed. The slope and bias of the
regression equation have been considered as model calibration factors. The output of
mathematical and statistical models along with other input are input to the feed-forward
ANN model with back-propagation algorithm for training. The weights and biases of ANN
model is also considered as model calibration factors. The model software is developed
for on-line automatic calibration of hybrid model for different steel groups. The present
invention is thus having the advantageous wide scale application of the hybrid model in
determining the output of calibrated model predicting temperatures in finishing stands
with an error less than 1.5%, thus improving product quality avoiding inaccuracy
involved in human estimation based operation.

BACKGROUND ART
It is known in the existing art that strips of different thickness and width are rolled from
slab in a Hot Strip Mill. The Hot Strip Mill of Rourkela Steel Plant of applicants consisting
of 3 Roughing stands and 6 Finishing stands. The first roughing stand (Ro/V0) is a
combination of a horizontal stand and a vertical stand. The other two roughing stands (R1
and R2) are 4 high horizontal stands. There is a delay table after R2 stand, one coil box
and a crop shear at the end of the delay table. There are six numbers of 4 high finishing
stands (F1 to F6) and two hydraulic down coilers. The operator of finishing stand sets roll
gap on the basis of temperature of first finishing stand entry temperature. In
conventional practice, the operator predicted finishing entry temperature based on
measured temperature at the exit of second roughing stand R2 on the basis of his
experience. There was however no practical basis for prediction of temperature before
and after each stand of finishing stands. As a result the finish rolling operation Involved
indeterministic control on process parameters including finishing stand temperatures. The
inaccuracy and uncertainty involved in the determination of finish rolling parameters thus
involved inaccuracy and uncertainty which In turn adversely affected the quality of the
finish rolled steel product.
There has been thus a continuing need in the rolling process in hot strip mill to
developing an online computer based system involving a hybrid thermal model for
determination, control and display of the finish rolling temperatures at entry and exit of
each finish stand so as to enable setting of process parameters of finishing stands in a
much deterministic manner without reliance on operator's experience and expertise in
order to eliminate the inaccuracy and undeslred variations in parameter values to carry
out finish rolling in a controlled manner resulting in superior and consistent quality of end
product.
In the Indian patent application number 1306/KOL/2009 of the applicants , there is
disclosed a system and method for predicting the temperature along the length of strip
after delay table based on a hybrid thermal model considering heat transfer in radiation
mode only. In case there is no pyrometer at the entry of first finishing stand, the delay
table model can be used to predict the temperature at entry of finishing stand and this
can be used as input to the thermal model for the finishing stand model of the present
invention.

OBJECTS OF THE INVENTION
The basic object of the present Invention is thus directed to developing a Mathematical-
statistical-ANN based hybrid model for accurate prediction of strip temperature at entry
and exit of each stand of finishing stands and display to operator through PLC-Server
system.
Another object of the present invention is directed to developing computer system
involving a Mathematical-statistical-ANN based hybrid software model wherein said
thermal model is based on radiation heat transfer, convection heat transfer with roll
cooling water, conduction heat loss to work rolls and heat generation due to plastic
deformation at the finishing stands.
Yet another object of the present invention is directed to developing a computer system
involving said Mathematical-statistical-ANN based hybrid model for accurate prediction of
temperatures in each finish rolling stands to enable the operator to select and set the
operating parameters In a reliable manner to ensure desired characteristics/quality of the
finish rolled product.
A further object of the present invention Is directed to developing a computer system
involving said Mathematical-statistical-ANN based hybrid model for accurate prediction of
temperatures in each finish rolling stands wherein said system computes the required
temperatures at entry/exit of finishing stands utilizing Input data relating to temperature
after R2, before Fl and after F6 and other input data such as roll gap setting, speed
setting and diameter of work rolls coming to the process work station(PWS).
A still further object of the present invention is directed to developing software enabled
computer system wherein said hybrid thermal model based software is loaded on process
work station that works on continuous basis so that the model output is shown to the
operator.
A still further object of the present invention is directed to developing a software enabled
computer system adapted to provide output feed back displayed to operator in another
computer system termed Operator Workstation (OW5).

A still further object of the present invention is directed to developing a software enabled
computer system wherein the PWS/OWS system has been connected with the PLC
system using OPC network software.
A still further object of the present invention Is directed to developing a software enabled
computer system wherein the hardware arrangement comprising the servers,
workstations, PLC system and the sensors/actuators that are deployed in three different
levels for implementation of the software.
SUMMARY OF THE INVENTION
The basic aspect of the present invention is thus directed to a system for online
prediction of strip temperature at the finishing stands of hot strip mill comprising,
means for acquiring online input data after last roughing stand(R2) , before entry of
first finishing stand(Fl) and after last finishing stand(F6) as well as other input data
comprising roll gap setting, speed setting and diameter of work rolls;
computer means adapted for predicting said finish stand temperatures along length of
strip based on said input data and output derived based on mathematical-Statistical-
Artificial Neural Network(ANN) based hybrid thermal model and generating a feed
back output;
means for controlling the finish stand setting based on said feed back output.
Another aspect of the present invention is directed to said system wherein said hybrid
thermal model comprising said mathematical model is based on equations derived from
fundamental principles of heat transfer comprising
(a) Temperature fall due to radiation heat loss:


where, ΔTrl = Radiation temperature fall of strip from previous stand, Tj-1 =
Temperature of strip at previous stand, 1= interstand distance In m, Vl-1= speed of
previous stand (m/s), tCB= time delay in coll box, Kr = Coefficient of radiation heat
transfer loss. Kr, a factor which is dependent upon the view factor of radiation has been
considered as a mathematical calibration factor in the model.
(b) Temperature fail due to roll cooling water loss

where, ΔTql = temperature fall due to roll cooling water, qD = roll cooling water flow rate
(m3/hr), b= width of water cooling header (mm), v = speed of stand (m/s), h=strip
thickness, KQ1, KQ2 = Coefficients of heat transfer due to water cooling loss. KQ1 and KQ2,
are factors considered as a mathematical calibration factors in the model.
(c) Temperature fall due to conduction to work rolls

where, ΔTCl = temperature fall due to conduction to work rolls, Tl-1 = Temperature
before entry into stand, h1i and h2i are thickness of strip before and after the stand, vi =
strip speed in the stand, roll cooling water flow rate (m3/hr), b= width of water cooling
header (mm), v = speed of stand (m/s), Ldi = projected arc of contact, KC = Coefficients
of heat transfer due to conduction to work rolls. Kc is a factor considered as a
mathematical calibration factors in the model.
(d) Temperature rise due to conduction to work rolls

where, ΔTD = temperature rise due to plastic deformation, h1 and h2 are thickness of
strip before and after the stand, P is the roll force and KD is a mathematical calibration
factor in the model.

A further aspect of the present invention is directed to said system wherein said
computer means comprises a process work station for feeding and storing of input data
and a separate operator work station for display of feedback output to the operator to
control the finishing stand settings based thereon.
A still further aspect of the present invention is directed to said system wherein said
hybrid model is adapted to predict temperature accurately at entry and exit of each stand
in the finish rolling stand and display the same to the operator through a PLC-Server
system.
Yet another aspect of the present invention directed to said system comprising OPC
networked PLC-Server system.
A still further aspect of the present invention directed to said system comprising,
hot strip mill comprising 3 roughing stands and 6 finishing stands, the first roughing
stand(R0/VO) being a combination horizontal stand and a vertical stand, the other two
roughing stands (R1 and R2) being 4 high horizontal stands, a delay table after R2 stand
, one coil box and a crop shear at the end of the delay table, six numbers of 4 high
finishing stands (F1 to F6) and two hydraulic down coilers;
process work station(PWS) adapted to execute the said hybrid model based prediction ;
operator work station(OWS) adapted to display model output to the operator; and
PLC based system operatively connected involving OPC network ,
wherein the temperature after R2, before Fl and after F6 along with data relating to roll
gap setting, speed setting and diameter of work rolls are provided to the process work
station and In the said process work station the input data are stored in a MS Access
database , said mathematical-statistical-Artlficial Neural Network (ANN) based hybrid
model adapted to read said input data from the database and generate the output data
passing through the PLC system and output displayed in the operators work station.

According to yet another aspect of the present invention directed to said system a linear
regression based statistical model, the slope and bias of the regression equation being
considered as model calibration factors, the output of mathematical and statistical
models along with other input data are input to the feed-forward ANN model with back-
propagation algorithm for training, the weights and biases of ANN model is also
considered as model calibration factors, said system adapted for on-line automatic
calibration of hybrid model for different steel groups.
A still further aspect of the present invention is directed to said system comprising
hierarchical input data structure for calibration purposes, selective initial ANN training
weights and input data sampling done on average basis and normalization of inputs data
segments based on linear interpolation.
The present invention and its objects and advantages are described in greater details
with reference to the following accompanying non limiting illustrative drawings.
BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS
Figure 1: is the schematic illustration of the existing layout arrangement of the roughing
and finishing stand and delay table, coil box and crop shear located in between, where
the hybrid model of the present invention is applied for on line process control.
Figure 2: is the schematic illustration of the computer system according to the present
invention involving the hardware deployment in three separate levels comprising the
PWS/OWS/Servers, PLC system and the sensors/actuators/ pyrometers for
implementation of the software.
Figure 3: is the illustration of the screenshot of the model calibration according to the
present Invention.
Figure 4: is the illustration of the conceptual diagram of the hybrid model of the present
invention.
8

Figure 5: is the schematic illustration of the automatic online data exchange for training
according to the present invention.
Figure 6: is the schematic Illustration of the hierarchy of different data group comprising
grade, coil width, coll thickness, coil box status and segment number for input data
structure for calibration purpose.
Figure 7: is the schematic illustration of the selection of initial training weights of ANN
instead of random weights.
Figure 8: is the illustration of the screenshot of typical model output to operator to
facilitate setting process parameters at finish rolling stage.
DETAILED DESCRIPTION OF THE INVENTION WITH REFERENCE TO THE
ACCOMPANYING DRAWINGS
The present invention is directed to providing an online software enabled system for
accurate prediction of temperature profile along the strip length at entry and exit of each
finishing stands in hot strip mill Involving a mathematical-Statistical-Artificial Neural
Network (ANN) based hybrid thermal model and display of the system output to operator
through PLC-Server system to facilitate setting of the finish rolling parameters including
the roll gaps.
Reference is first invited to the accompanying Figure 1, that schematically illustrates
existing hot strip mill facility in one of the steel plants of the applicant that comprised
three roughing stands (RS) and six numbers horizontal finish rolling stands. There is
delay table (DT) of specified length and a coil box (CB) in between the roughing stand
and the finishing stand. It is clearly apparent from the Figure 1 that the first roughing
stand (R0/V0) is a combination of horizontal stand and a vertical stand, whereas the other
two roughing stands e.g. R1 and R2 next in a row after R0/V0, are 4 high horizontal
stands. The delay table (DT) is located after R2 stand. There are also one coil box (CB)
and a crop shear (CS) at the end of the delay table. There are six numbers of 4 high
finishing stands numbered as F1 at entry to F6 at the exit of finish rolling stage and two
hydraulic down coilers. The operator of finishing stand sets roll gap on the basis of

temperature of first finishing stand entry temperature. In conventional practice, the
operator predicted finishing entry temperature based on measured temperature at the
exit of second roughing stand R2 on the basis of his experience. There has been however
no practical basis for prediction of temperature before and after each stand of finishing
stands.
Accompanying Figure 2 schematically illustrate the deployment of the hardware of the
PLC-Server based computer system in different levels. The system of the Invention
comprises a process work station (PWS) in which the hybrid model is loaded and works
on continuous basis and a Operator Work Station (OWS) which shows model output to
the operator. This system has been connected with the PLC system using OPC network
software. Through this system the temperature data after R2, before F1 and after F6
comes to the PWS. Other input data such as roll gap setting, speed setting and diameter
of work rolls also comes to the PWS. In PWS, all these input parameters are stored in a
MS Access database. The model software reads all these input data from the database
and calculates temperature before and after each finishing stand. The output data passes
through the PLC system in similar fashion and output is shown to the operator. It is seen
from the Figure 2, that the system comprises OWS and PWS alongwith the PDI server in
level 2 which are integrated with PLC and level 1 server while the PLC is coupled with the
sensors and actuators (Pyrometers) to establish bothway communication for input and
processed data with the level 0 at the operating level so that the above hardware
arrangement favours implementation of the hybrid software based model for desired
controlled setting of finish roll parameters. The output is displayed to operator in another
computer system called Operator Workstation. Operator controls the finishing stand
setting based on this feedback.
It has been described that the present invention Is directed to a hybrid thermal model
based system for predicting the finishing stand temperatures accurately.
The hybrid model according to the invention consists of mathematical, statistical and
artificial neural network based model. The mathematical model is based on there
equations derived from fundamental principles of heat transfer. These equations are
described below:
(a) Temperature fall due to radiation heat loss:

where, ΔTri = Radiation temperature fall of strip from previous stand, Tl-1 =
Temperature of strip at previous stand, 1= interstand distance in m, Vl-1= speed of
previous stand (m/s), tCB= time delay in coil box, Kr = Coefficient of radiation heat
transfer loss. Kr, a factor which is dependent upon the view factor of radiation has been
considered as a mathematical calibration factor in the model.
(b) Temperature fall due to roll cooling water loss

where, ΔTq, = temperature fall due to roll cooling water, qD = roll cooling water flow rate
(m3/hr), b= width of water cooling header (mm), v = speed of stand (m/s), h=strip
thickness, KQ1, KQ2 = Coefficients of heat transfer due to water cooling loss. KQ1 and KQ2,
are factors considered as a mathematical calibration factors in the model.
(c) Temperature fall due to conduction to work rolls

where, ΔTCi = temperature fall due to conduction to work rolls, Ti-1 = Temperature
before entry into stand, h1i and h2i are thickness of strip before and after the stand, vi =
strip speed in the stand, roll cooling water flow rate (m3/hr), b= width of water cooling
header (mm), v = speed of stand (m/s), U = projected arc of contact, KC = Coefficients
of heat transfer due to conduction to work rolls. Kc is a factor considered as a
mathematical calibration factors in the model.
(d) Temperature rise due to conduction to work rolls


where, ΔTD = temperature rise due to plastic deformation, h1 and h2 are thickness of
strip before and after the stand, P is the roll force and KD is a mathematical calibration
factor in the model.
A linear regression based statistical model has been developed. The slope and bias of the
regression equation have been considered as model calibration factors. The output of
mathematical and statistical models along with other input are input to the feed-forward
ANN model with back-propagation algorithm for training.
The weights and biases of ANN model is also considered as model calibration factors. The
software has been developed for on-line automatic calibration of hybrid model for
different steel groups (a steel group consisting of a grade, a target width, a target
thickness and a coilbox use). The model Is automatically calibrated for different steel
groups. It has been found that the output of calibrated model predicts temperature very
accurately. The screenshot of the hybrid model output is illustrated in the accompanying
Figure 3. The model also compute the error % for each data set.
The working of the hybrid thermal model for prediction of temperature along strip length
at the entry and exit of each rolling stand in the finishing stands, implemented by way of
the hardware configuration and the software Implementation comprising the steps of
(a) Development of mathematical model for heat transfer in finishing stands consists
of radiation heat transfer, convection heat transfer with roll cooling water,
conduction heat loss to work rolls and heat generation due to plastic deformation;
(b) Identification of key calibration factor for the mathematical model;
(c) Development Statistical model using linear regression technique;
(d) Development of artificial neural network based model for finishing stands in which
output of mathematical and statistical model has been used as additional input
parameters along with other measurable input parameters affecting heat transfer;
(e) Development of data communication system between pyrometers, PLCs, Process
Work Station and Operator Work Station for segment wise collection of data and
segment-wise prediction of temperature of coils at each finishing stand;

(f) Design of database structure for Primary Data Inputs of Coil, Pyrometer
temperatures, Coilbox status and model calibration factors;
(g) Development of software for on-line automatic calibration/training of hybrid model
for different steel groups (a steel group consists of a grade, a target width, a
target thickness and a coilbox use);
(h) The initial weights and bias of ANN model has been chosen in a particular fashion
against the traditional procedure of selecting random weights.
(I) On-line trial of the model has been done and it is observed that the output of
calibrated model predicts temperature with an error less than 1.5%.
Reference is now invited to the accompanying Figure 4 that schematically illustrate the
conceptual diagram for the above described hybrid thermal model for operation of the
model on the input data set for different steel group. Accompanying Figure 5 illustrates
automatic online data exchange for training. Accompanying Figure 6 shows the
hierarchy of data group used for computing the finish temperature by the hybrid model.
Reference is also invited to the accompanying Figure 7 that shows the use of selected
initial weights 1.0, 0.5 and 0.0 for operation on a data segment to predict the
temperatures at finishing stands. Accompanying Figure 8 illustrates screenshot of typical
model output to operator to facilitate setting of process parameters at finish rolling stage.
It is observed that the model error % corresponding to actual/measured F1 and F6
temperatures are 0.5% and 1.5% respectively.
It is thus possible by way of the present invention to developing an online computer
system and method for accurate and precise prediction of the entry and exit
temperatures at each finish stand and temperature profile along strip length at the finish
rolling stands, based on input data relating to temperature data after R2, before Fl and
after F6 and also other input data such as roll gap setting, speed setting and diameter of
work rolls etc., by involving a mathematical, statistical and Artificial Neural Network
based hybrid thermal model based software. The system allow display of the results of
prediction to the operator to favour setting the finishing roll parameters in a reliable and
continuous manner so as to improve productivity and quality of out put steel strip
products in Hot Strip mill. The system and the process of the invention are adapted to
predict temperature limiting the error to less than 1.5% by involving an automatic on-
line calibration module. The system of the invention is thus having prospect of wide scale
application in Hot Strip Mills (HSM) with improved efficiency and economy in HSM.

We claim:
1. A system for online prediction of strip temperature at the finishing stands of hot strip
mill comprising,
means for acquiring online input data after last roughing stand(R2) , before entry of
first finishing stand(Fl) and after last finishing stand(F6) as well as other input data
comprising roll gap setting, speed setting and diameter of work rolls;
computer means adapted for predicting said finish stand temperatures along length of
strip based on said input data and output derived based on mathematical-Statistical-
Artificial Neural Network(ANN) based hybrid thermal model and generating a feed
back output;
means for controlling the finish stand setting based on said feed back output.
2. A system as claimed in claim 1 wherein said hybrid thermal model comprising said
mathematical model is based on there equations derived from fundamental principles
of heat transfer comprising
(a) Temperature fall due to radiation heat loss:

where, ΔTri = Radiation temperature fall of strip from previous stand, Ti-1 =
Temperature of strip at previous stand, l= interstand distance in m, vi-1= speed of
previous stand (m/s), tCB= time delay in coil box, Kr = Coefficient of radiation heat
transfer loss. Kr, a factor which is dependent upon the view factor of radiation has been
considered as a mathematical calibration factor in the model.
(b) Temperature fall due to roll cooling water loss


where, ΔTql = temperature fall due to roll cooling water, qD = roll cooling water flow rate
(m3/nr), b= width of water cooling header (mm), v = speed of stand (m/s), h=strip
thickness, KQ1, KQ2 = Coefficients of heat transfer due to water cooling loss. KQ1 and KQ2,
are factors considered as a mathematical calibration factors in the model.
(c) Temperature fall due to conduction to work rolls

where, ΔTCi = temperature fall due to conduction to work rolls, Ti-1 = Temperature
before entry into stand, h1i, and h2i are thickness of strip before and after the stand, vi =
strip speed in the stand, roll cooling water flow rate (m3/hr), b= width of water cooling
header (mm), v = speed of stand (m/s), Ldi = projected arc of contact, KC = Coefficients
of heat transfer due to conduction to work rolls. Kc is a factor considered as a
mathematical calibration factors In the model.
(d) Temperature rise due to conduction to work rolls

where, ΔTD = temperature rise due to plastic deformation, h1 and h2 are thickness of
strip before and after the stand, P is the roll force and KD is a mathematical calibration
factor in the model.
3. A system as claimed in claim 1 or 2 wherein said computer means comprises a process
work station for feeding and storing of input data and a separate operator work station
for display of feedback output to the operator to control the finishing stand settings
based thereon.
4. A system as claimed in anyone of claims 1 to 3 wherein said hybrid model is adapted
to predict temperature accurately at entry and exit of each stand in the finish rolling
stand and display the same to the operator through a PLC-Server system.

5. A system as claimed in anyone of claims 1 to 4 comprising OPC networked PLC-Server
system.
6. A system as claimed in anyone of claims 1 to 5 comprising
hot strip mill comprising 3 roughing stands and 6 finishing stands, the first roughing
stand(R0/V0) being a combination horizontal stand and a vertical stand, the other two
roughing stands (R1 and R2) being 4 high horizontal stands, a delay table after R2 stand
, one coil box and a crop shear at the end of the delay table, six numbers of 4 high
finishing stands (F1 to F6) and two hydraulic down coilers;
process work station(PWS) adapted to execute the said hybrid model based prediction ;
operator work station(OWS) adapted to display model output to the operator; and
PLC based system operatively connected involving OPC network ,
wherein the temperature after R2, before Fl and after F6 along with data relating to roll
gap setting, speed setting and diameter of work rolls are provided to the process work
station and in the said process work station the input data are stored in a MS Access
database , said mathematical-Statistical-Artificial Neural Network (ANN) based hybrid
model adapted to read said input data from the database and generate the output data
passing through the PLC system and output displayed in the operators work station.
7. A system as claimed in claim 5 comprising a linear regression based statistical model,
the slope and bias of the regression equation being considered as model calibration
factors, the output of mathematical and statistical models along with other input data are
input to the feed-forward ANN model with back-propagation algorithm for training, the
weights and biases of ANN model is also considered as model calibration factors, said
system adapted for on-line automatic calibration of hybrid model for different steel
groups.

8. A system as claimed in anyone of claims 1 to 7 comprising hierarchical input data
structure for calibration purposes, selective initial ANN training weights and input data
sampling done on average basis and normalization of inputs data segments based on
linear interpolation.
9. A system for prediction of temperature profile along strip length at the finishing
stands of a hot strip mill substantially as hereindescrlbed and illustrated with reference to
the accompanying figures.

A system for predicting strip temperature in finishing stands of hot strip mill involving
mathematical-Statistical-Artificial Neural Network(ANN) based hybrid thermal model for
online acquisition, analysis, and predict temperature along strip length before and after
each stand of finishing stands of hot strip mill. The system soft ware is adapted to
receive the input data from pyrometer located before F1 and after F6, and other data
relating to roll gap setting, speed setting and diameter of work rolls via a PLC system
connected with OPC network to a computer system termed Process Work Station(PWS).
The output is displayed to operator in another computer system viz Operator
Workstation(OWS). Operator controls the finishing stand setting based on this feedback.
The hybrid thermal model wherein mathematical model is based on heat transfer in
finishing stands consisting of radiation heat transfer, convection heat transfer with roll
cooling water, conduction heat loss to work rolls and heat generation due to plastic
deformation. The model predicts finishing stand temperatures with less than 1.5% error.