DESC:TECHNICAL FIELD
The present disclosure generally relates to optimization systems, and especially relates to a system for real-time polymer properties optimization while polymer formation and a method thereof.
DEFINITIONS
As used in the present disclosure, the following terms are generally intended to have the meaning as set forth below, except to the extent that the context in which they are used indicates otherwise.
Soft Sensor: The term ‘Soft sensor’ refers to software where several measurements are processed together. Commonly soft sensors are based on control theory and also receive the name of state observer.
Melt Flow Index (MFI): The term ‘(MFI)’ refers to a measure of the ease of flow of the melt of a thermoplastic polymer. It is also defined as the mass of polymer, in grams, flowing in ten minutes through a capillary of a specific diameter and length by a pressure applied via prescribed alternative gravimetric weights for alternative prescribed temperatures. The terms Melt Flow Rate (MFR) and Melt Index (MI) can be used interchangeably with Melt Flow Index (MFI).
Xylene Solubility (XS): The term ‘xylene solubility (XS)’ refers to the percentage of soluble species in polypropylene homo- and co-polymers. This value is almost proportional to the amorphous content of the material. In practice, this measure is broadly used for product quality control and monitoring the physical properties of a polymer during synthesis and processing.
Machine learning (ML): The term ‘Machine learning (ML)’ refers to a field of inquiry devoted to understanding and building methods that 'learn', that is, methods that leverage data to improve performance on some set of tasks. It is seen as a part of artificial intelligence (AI). Machine learning algorithms build a model based on sample data, known as training data, in order to make predictions or decisions without being explicitly programmed to do so.
First principle model: The term ‘first principle model’ refers to models that are built on a fundamental understanding of underlying ‘ab initio’ physio-chemical phenomena such as reaction kinetics, mass transfer, heat transfer, and mass flow.
Population balance method: The term ‘population balance method’ generally describes the evolution of a population of particles. The population balance modeling is a widely used approach to describe, crystallization processes, taking into account not only the primary phenomena like nucleation and growth, as well as particle agglomeration and breakage which can be extended to multivariate cases where more internal coordinates i.e. particle properties can be used.
Advanced Process Control (APC): The term advanced process control refers to a broad range of techniques and technologies implemented within industrial process control systems. Advanced process controls are usually deployed optionally and in addition to basic process controls.
Manipulated Variable: A process parameter or variable that is used to control the process.
Set point: Set point is the target value set for a process variable by the operator or an optimizing control system at which the controller attempts to maintain the process variable.
Master-Slave Controller: A master-salve controller is an existing controller element and slave element or elements. The Master controller is a device or a process that will be controlling or adjusting the slave element or elements to attain a desired process state and the slave element or the elements are the process parameters that are being controlled by the Master Controller.
Real-Time Optimization (RTO): The term Real-Time Optimization (RTO) refers to a category of closed-loop process control that aims at optimizing process performance in real-time for systems.
Pre-Set Rules: The term Pre-Set Rules refers to a plurality of rules running a first principle model (FPM).

BACKGROUND
The background information herein below relates to the present disclosure but is not necessarily prior art.
Generally, laboratory analysis techniques are used to determine the properties of polymers. Monitoring of a commercial unit based on intermittent sampling with conventional laboratory analysis methods is not satisfactory for continuously maintaining desired properties of polymers. Typically, the properties of the final polymers are obtained by laboratory analysis which can be completed within 90-120 minutes after drawing the samples at the reactor outlet. Also, this sampling and testing is done at regular time intervals. This time delay plays a crucial role in deciding the time requirements for the transition of a polymer from one grade to another. Usually, the grade transition of polymers is highly dynamic. So, operational adjustments are highly dependent on the properties of polymer samples. While operating a commercial polymer reactor based on lab analysis of the polymer, because of the time delay, a significant amount of undesired grades of the polymer are produced and sometimes the quality of the polymer product drifts out of range. Therefore, a correction/change in the reactor is required to be continuously monitored and adjusted.
In the making of polymers, catalysts, cocatalysts, and additives are often used. Sometimes, a trace amount of catalyst or, cocatalysts /additives remains in the sample at the reactor outlet. This also results in variations in the properties of the polymer during the laboratory analysis of the polymer samples.
Typically, laboratory methods are used to measure the melt flow index (MFI) of the polymer material by melting a polymer sample. Online MFI measurement instruments to monitor the polymer sample quality have limitations because of the instrument’s inaccuracy, frequent choking, and lumps formation by unstable polymer samples.
Several attempts have been made the development of soft sensors for the prediction of the properties of a polymer from first principle models based on the population balance method, reaction kinetics, and the like. A single first principle model is not satisfactory for a commercial plant because it often ignores or simplifies the non-ideal behavior in a commercial reactor. Models are often developed based on laboratory setups with idealistic assumptions and hence often fail to perform in an industrial environment (large scale). Further, these models are not always appropriate/feasible to repeat the polymerization process on a large scale.
ML models based on different available algorithms have also been developed for the prediction of polymer properties based on data from experimental studies. Since most studies have been carried out in laboratories, these ML models developed are tested on limited industrial/commercial data available. As a result, ML models are prone to making false positives or false negatives because of insufficient data/poor quality data. In addition, the ML model’s limited explainability is a barrier to utilizing them in APC and RTO applications of commercial polymer production.
Therefore, there is a need for a system for real-time polymer properties optimization while polymer formation and a method thereof that can mitigate the drawbacks mentioned hereinabove or at least provide an alternative solution.

OBJECTS
Some of the objects of the present disclosure, which at least one embodiment herein satisfies, are as follows:
It is an object of the present disclosure to ameliorate one or more problems of the prior art or to at least provide a useful alternative.
An object of the present disclosure is to provide a system for real-time polymer properties optimization while polymer formation and a method thereof.
Another object of the present disclosure is to provide a system for real-time polymer properties optimization while polymer formation and a method thereof that predicts the upcoming polymer properties along with the extent of change required in reactor parameters to sustain the desired polymer properties.
Still another object of the present disclosure is to provide a system and a method that produce a stable polymer.
Yet another object of the present disclosure is to provide a system and a method that provide a real-time indication of the polymer properties at the reactor outlet and provides a real-time optimization and advanced process control of the polymer reactor based on the predicted properties.
Another object of the present disclosure is to provide a system and a method that use predictive models for Melt Flow Index (MFI) and Xylene Solubility (XS) of a polymer at the reactor outlet.
Yet another object of the present disclosure is to provide a system and a method that use a hybrid model (a combination of the first principle model and machine learning/artificial intelligence model) to optimize the polymerization process on a large scale.
Another object of the present disclosure is to provide a system and a method that are simple and economical.
Still another object of the present disclosure is to provide a system and a method that prevent lump formation in a polymer reactor.
Other objects and advantages of the present disclosure will be more apparent from the following description when read in conjunction with the accompanying figures, which are not intended to limit the scope of the present disclosure.

SUMMARY
The present disclosure envisages a system for real-time polymer properties optimization while a polymer is being produced in a reactor, where the reactor is equipped with a plurality of soft sensors. The system comprises a data capturing unit, a repository unit, an operational model, a first principle model (FPM), and a control system.
The data capturing unit is configured to capture a plurality of real-time process parameters with a time stamp from the plurality of soft sensors and configured to cooperate with a digital user interface to receive a plurality of target polymer properties and a plurality of lab analysis result parameters.
In an aspect, the real-time process parameters include a temperature profile of the reactor, a composition of a gaseous reactor effluent, a power value, a current value, a voltage value, a torque of a polymerization reactor agitator, a dew point or a composition of the gaseous reactor effluent, a hydrogen flow rate into the reactor, a catalyst flow rate to the reactor, co-catalyst(s) flow rate to the reactor, catalyst promoter(s) flow rate to the reactor, a recycle gas flow rate to the reactor, and a silane flow rate to the reactor.
In an aspect, the data capturing unit comprises a real-time reactor parameters collection module, a lab result collection module, and a data sending module.
The real-time reactor parameters collection module is configured to capture a plurality of process parameters and sampling time details in real-time from the plurality of soft sensors installed in the reactor.
The lab result collection module is configured to fetch a plurality of laboratory analysis results digitally received from the user interface or a central remote server.
The data sending module is configured to send the plurality of real-time process parameters with the time stamp and the plurality of lab analysis result parameters to the repository to store.
The repository unit, coupled with the data capturing unit, is configured to receive and store the plurality of real-time process parameters with the time stamp, the plurality of target polymer properties, the plurality of lab analysis result parameters along with a plurality of pre-set rules and a threshold change value.
The operational model, coupled with the data capturing unit and the repository unit, is configured to predict a plurality of polymer properties of the polymer being produced in the reactor, the operational model includes an Artificial Intelligence (AI) unit including a plurality of AI models, each of which being configured to predict a magnitude of change in the polymer properties of the polymer being produced in the reactor, using the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp as input.
In an aspect, the operational model comprises a deviation value calculation unit configured to receive the final magnitude of change and the plurality of target polymer properties, and further configured to calculate a deviation value for each of the real-time process parameters using the FPM magnitude, when the final magnitude of change is greater than a threshold value.
The first principle model (FPM) is configured to derive an FPM magnitude of change in the polymer properties of the polymer being produced in the reactor by using the pre-set rules on the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp. The FPM comprises a validation module.
The validation module is configured to compare the FPM magnitude of change with each magnitude of change predicted by the plurality of AI models to shortlist a final magnitude of change closest to the FPM magnitude of change and further configured to identify whether the final magnitude of change is greater than the threshold change value.
The control system is configured to operate a set of reactor hardware components when the final magnitude of change is identified as greater than the threshold value, so as to optimise the polymer properties of the polymer being produced by changing values of the real-time process parameters to maintain the value of the final magnitude of change below the threshold value.
In an aspect, the control system is configured to receive the deviation value for each of the real-time process parameters and further configured to operate the reactor components to achieve the production of a polymer having the target polymer properties.
In an aspect, the control system transmits a control signal to a master controller, an advanced process control (APC) system, or directly to a slave controller.
In an aspect, the master controller is operable to transmit the control signal received from the control system to the slave controller or a final control element so as to bring the value of polymer properties back to the target value of the polymer properties when the final magnitude of change is greater than a threshold value.
In an aspect, the slave element or the final control element is a single process variable or multiple variables which are manipulated based on the control signal from the master-slave controller, or a control system that controls the process variable or variables to obtain the targeted process parameters.
In an aspect, a computing unit having a comparing unit operable to compare the laboratory analysis results captured by the lab result collection module with the magnitude of change in the polymer properties generated by each of the plurality of AI models for a given time stamp and calculate a gap value for each of the plurality of AI models.
In an aspect, a training unit is configured to train a selected AI model for which the gap value of the AI model continuously crosses a pre-set error tolerance range.
In an aspect, a display unit is operable to display the real-time process parameters for each of the real-time process parameters.
The present disclosure further envisages a method for real-time polymer properties optimization while a polymer is being produced in a reactor.

BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWING
A system for real-time polymer properties optimization while polymer formation and a method thereof of the present disclosure will now be described with the help of the accompanying drawing, in which:
Figure 1 illustrates a block diagram for a system for real-time polymer properties optimization while polymer formation, in accordance with an embodiment of the present disclosure;
Figure 1a illustrates a flow diagram for a method for real-time polymer properties optimization while polymer formation, in accordance with an embodiment of the present disclosure;
Figure 2 illustrates a comparison of real time MFI prediction against laboratory analysis results for a commercial stirred bed gas phase reactor 1, in accordance with an embodiment of the present disclosure;
Figure 3 illustrates a graph showing a comparison between the real-time prediction of the Xylene Solubility (XS) of the polymer and the laboratory analysed XS of the polymer for a commercial stirred bed gas phase reactor 1, in accordance with an embodiment of the present disclosure;
Figure 4 illustrates a graph showing a comparison between the real-time prediction of the Melt flow index (MFI) of the polymer and the laboratory analysed MFI of the polymer for a commercial stirred bed gas phase reactor 2, in accordance with an embodiment of the present disclosure;
Figure 5 illustrates a graph showing a comparison between the real-time prediction of the Xylene Solubility (XS) of the polymer and the laboratory analysed XS of the polymer for a commercial stirred bed gas phase reactor 2, in accordance with an embodiment of the present disclosure;
Figure 6 illustrates a parity plot of predicted MFI vs Laboratory generated MFI for a commercial stirred bed gas phase reactor 1, in accordance with an embodiment of the present disclosure;
Figure 7 illustrates a parity plot of the real-time prediction of predicted XS vs Laboratory generated XS for a commercial stirred bed gas phase reactor 1, in accordance with an embodiment of the present disclosure;
Figure 8 illustrates a parity plot of the real-time prediction of predicted MFI vs Laboratory generated MFI for stirred bed gas phase reactor 2, in accordance with an embodiment of the present disclosure;
Figure 9 illustrates a parity plot of the real-time prediction of predicted XS vs Laboratory generated XS for stirred bed gas phase reactor 2, in accordance with an embodiment of the present disclosure; and
Figure 10 illustrates a typical reactor outlet vapor composition distribution, in accordance with an embodiment of the present disclosure.

DETAILED DESCRIPTION
Embodiments, of the present disclosure, will now be described with reference to the accompanying drawing.
Embodiments are provided so as to thoroughly and fully convey the scope of the present disclosure to the person skilled in the art. Numerous details are set forth, relating to specific components, and methods, to provide a complete understanding of embodiments of the present disclosure. It will be apparent to the person skilled in the art that the details provided in the embodiments should not be construed to limit the scope of the present disclosure. In some embodiments, well-known processes, well-known apparatus structures, and well-known techniques are not described in detail.
The terminology used, in the present disclosure, is only for the purpose of explaining a particular embodiment and such terminology shall not be considered to limit the scope of the present disclosure. As used in the present disclosure, the forms "a,” "an," and "the" may be intended to include the plural forms as well, unless the context clearly suggests otherwise. The terms “including,” and “having,” are open ended transitional phrases and therefore specify the presence of stated features, integers, steps, operations, elements and/or components, but do not forbid the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. The particular order of steps disclosed in the method and process of the present disclosure is not to be construed as necessarily requiring their performance as described or illustrated. It is also to be understood that additional or alternative steps may be employed.
Generally, laboratory analysis techniques are used to determine the properties of polymers. Monitoring of a commercial unit based on intermittent sampling with conventional laboratory analysis methods is not satisfactory for continuously maintaining desired properties of polymers. Typically, the properties of the final polymers are obtained by laboratory analysis which can be completed within 90-120 minutes after drawing the samples at the reactor outlet. Also, this sampling and testing is done at regular time intervals. This time delay plays a crucial role in deciding the time requirements for the transition of a polymer from one grade to another. Usually, the grade transition of polymers is highly dynamic. So, operational adjustments are highly dependent on the properties of polymer samples. While operating a commercial polymer reactor based on lab analysis of the polymer, because of the time delay, a significant amount of undesired grades of the polymer are produced and sometimes the quality of the polymer product drifts out of range. Therefore, a correction/change in the reactor is required to be continuously monitored and adjusted.
In the making of polymers, catalysts, cocatalysts, and additives are often used. Sometimes, a trace amount of catalyst or, cocatalysts /additives remains in the sample at the reactor outlet. This also results in variations in the properties of the polymer during the laboratory analysis of the polymer samples.
Typically, laboratory methods are used to measure the melt flow index (MFI) of the polymer material by melting a polymer sample. Online MFI measurement instruments to monitor the polymer sample quality have limitations because of the instrument’s inaccuracy, frequent choking, and lumps formation by unstable polymer samples.
Several attempts have been made for the development of soft sensors for the prediction of the properties of a polymer from first principle models based on population balance method, reaction kinetics, and the like. A single first principle model is not satisfactory for a commercial plant because it often ignores or simplifies the non-ideal behavior in a commercial reactor. Models are often developed based on laboratory setups with idealistic assumptions and hence often fail to perform in an industrial environment (large scale). Further, these models are not always appropriate/feasible to repeat the polymerization process on a large scale.
ML models based on different available algorithms have also been developed for the prediction of polymer properties based on data from experimental studies. Since most studies have been carried out in laboratories, these ML models developed are tested on limited industrial/commercial data available. As a result, ML models are prone to making false positives or false negatives because of insufficient data/poor quality data. In addition, the ML model’s limited explainability is a barrier to utilizing them in APC and RTO applications of commercial polymer production.
To overcome the above-mentioned drawbacks, the present disclosure envisages a system (hereinafter referred to as “system 100”) for real-time polymer properties optimization while polymer formation and a method (hereinafter referred to as “method 2000”) thereof.
Referring to Figure 1, a block diagram for the system for real-time polymer properties optimization while a polymer being produced in a reactor is shown, wherein the reactor is equipped with a plurality of soft sensors. The system comprises a data capturing unit 200, a repository unit 101, an operational model 300, a first principle model (FPM) 104, and a control system.
The data capturing unit 200 is configured to capture a plurality of real-time process parameters with a time stamp from the plurality of soft sensors and configured to cooperate with a digital user interface to receive a plurality of target polymer properties and a plurality of lab analysis result parameters.
In an aspect, the real-time process parameters include a temperature profile of the reactor, a composition of a gaseous reactor effluent, a power value, a current value, a voltage value, a torque of a polymerization reactor agitator, a dew point or a composition of the gaseous reactor effluent, a hydrogen flow rate into the reactor, a catalyst flow rate to the reactor, co-catalyst(s) flow rate to the reactor, catalyst promoter(s) flow rate to the reactor, a recycle gas flow rate to the reactor, and a silane flow rate to the reactor.
In an aspect, the data capturing unit 200 comprises a real-time reactor parameters collection module 102, a lab result collection module 103, and a data sending module 1111.
The real-time reactor parameters collection module 102 is configured to capture a plurality of process parameters and sampling time details in real-time from the plurality of soft sensors installed in the reactor.
The lab result collection module 103 is configured to fetch a plurality of laboratory analysis results digitally received from the user interface or a central remote server; and
The data sending module 1111 is configured to send the plurality of real-time process parameters with the time stamp and the plurality of lab analysis result parameters to the repository 101 to store.
The repository unit 101, coupled with the data capturing unit 200, is configured to receive and store the plurality of real-time process parameters with the time stamp, the plurality of target polymer properties, the plurality of lab analysis result parameters along with a plurality of pre-set rules and a threshold change value.
The operational model 300, coupled with the data capturing unit 200 and the repository unit 101, configured to predict a plurality of polymer properties of the polymer being produced in the reactor, the operational model 300 includes an Artificial Intelligence (AI) unit including a plurality of AI models, each of which is configured to predict a magnitude of change in the polymer properties of the polymer being produced in the reactor, using the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp as input.
In an aspect, the operational model 300 comprises a deviation value calculation unit configured to receive the final magnitude of change and the plurality of target polymer properties, and further configured to calculate a deviation value for each of the real-time process parameters using the FPM magnitude, when the final magnitude of change is greater than a threshold value.
The first principle model (FPM) 104 is configured to derive an FPM magnitude of change in the polymer properties of the polymer being produced in the reactor by using the pre-set rules on the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp. The FPM comprises a validation module.
The validation module is configured to compare the FPM magnitude of change with each magnitude of change predicted by the plurality of AI models to shortlist a final magnitude of change closest to the FPM magnitude of change and further configured to identify whether the final magnitude of change is greater than the threshold change value.
The control system is configured to operate a set of reactor hardware components when the final magnitude of change is identified as greater than the threshold value, so as to optimise the polymer properties of the polymer being produced by changing values of the real-time process parameters to maintain the value of the final magnitude of change below the threshold value.
In an aspect, the control system is configured to receive the deviation value for each of the real-time process parameters and further configured to operate the reactor components to achieve the production of a polymer having the target polymer properties.
In an aspect, the control system transmits a control signal to a master controller 106, an advanced process control (APC) system, or directly to a slave controller 106a.
In an aspect, the master controller 106 is operable to transmit the control signal received from the control system to the slave controller 106a or a final control element 106b so as to bring the value of polymer properties back to the target value of the polymer properties when the final magnitude of change is greater than a threshold value.
In an aspect, the slave element 106a or the final control element 106b is a single process variable or multiple variables which are manipulated based on the control signal from the master-slave controller 106, or a control system that controls the process variable or variables to obtain the targeted process parameters.
In an aspect, a computing unit 400 having a comparing unit 107 operable to compare the laboratory analysis results captured by the lab result collection module 103 with the magnitude of change in the polymer properties generated by each of the plurality of AI models for a given time stamp and calculate a gap value for each of the plurality of AI models.
In an aspect, a training unit 108 is configured to train a selected AI model for which the gap value of the AI model continuously crosses a pre-set error tolerance range.
In an aspect, a display unit 109 is operable to display the real-time process parameters for each of the real-time process parameters.

Referring to Figure 1a, a flow diagram for the method 2000 for optimization of polymer properties in real–time of a polymer being produced in a reactor is shown, where the reactor is equipped with a plurality of soft sensors. The method comprises steps of:
At step 202, the method 2000 includes the step of capturing, by a data capturing unit 200, a plurality of real-time process parameters with a time stamp from the plurality of soft sensors and configured to cooperate with a digital user interface to receive a plurality of target polymer properties and a plurality of lab analysis result parameters.
At step 204, the method 2000 includes the step of receiving and storing, by a repository unit 101 coupled with the data capturing unit 200, the plurality of real-time process parameters with the time stamp, the plurality of target polymer properties, the plurality of lab analysis result parameters along with a plurality of pre-set rules and a threshold change value.
At step 206, the method 2000 includes the step of predicting, by an operational model 300 coupled with the data capturing unit 200 and the repository unit 101, a plurality of polymer properties of the polymer being produced in the reactor, the operational model 300 includes an Artificial Intelligence (AI) unit including a plurality of AI models, each of which being configured to predict a magnitude of change in the polymer properties of the polymer being produced in the reactor, using the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp as input.
At step 208, the method 2000 includes the step of deriving, by a first principle model (FPM) 104, an FPM magnitude of change in the polymer properties of the polymer being produced in the reactor by using the pre-set rules on the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp.
At step 210, the method 2000 includes the step of comparing, by a validation module of the FPM, the FPM magnitude of change with each magnitude of change predicted by the plurality of AI models to shortlist a final magnitude of change closest to the FPM magnitude of change and further configured to identify whether the final magnitude of change is greater than the threshold change value.
At step 212, the method 2000 includes the step of operating, by a control system, a set of reactor hardware components, when the final magnitude of change is identified greater than the threshold value, so as to optimise the polymer properties of the polymer being produced by changing values of the real-time process parameters to maintain the value of the final magnitude of change below the threshold value.
In accordance with an embodiment of the present disclosure, the input information can include but is not limited to, real-time process parameters and sampling time details from polymer production plant, samples, and laboratory results of polymer properties including but not limited to Melt Flow Index (MFI) or Melt Flow Rate (MFR), Xylene Solubility (XS). Other suitable input information can also be used.
In accordance with an embodiment of the present disclosure, the real-time process parameters and sampling time details from the polymer production plant for the prediction of the polymer properties such as MFI, MFR, XS, and the like. can include, but is not limited to, the temperature profile of the reactor, composition of the gaseous reactor effluent, power, current, voltage and torque of the polymerization reactor agitator, the dew point or composition of the gaseous reactor effluent, hydrogen flow to the reactor, catalyst flow to a reactor, cocatalyst(s) flow to the reactor, catalyst promoter(s) flow to the reactor, recycle gas flow to the reactor and silane flow to the reactor. Other suitable real-time process parameters can also be used. In accordance with an embodiment of the present disclosure, the composition of the gaseous reactor effluent can include but is not limited to, condensable and non-condensable at a temperature in the range of 40 °C to 105 °C and a pressure range of 26 kg/cm2g to 30 kg/cm2g. Other suitable compositions of the gaseous reactor effluent can also be used.
In accordance with an embodiment of the present disclosure, the operational model can include, but is not limited to, first principle model 104 and AI/ML model 105. Other suitable operational models can also be used.
In accordance with the present disclosure, the target property range is set by the plant operator depending on the grade of polymer produced.
In accordance with the present disclosure, the master-slave controller 106 acts on slave elements 106a or final control element 106b like a hydrogen or silane flow control valve in the polymerization reactor inlet or any other suitable process parameters that can be used as the slave control element.
In accordance with an embodiment of the present disclosure, an algorithm for the development of a set of AI/ML model 105 can include, but is not limited to, partial least squares regression (PLS), principal component analysis (PCA), radial basis function (RBF), support vector regression (SVR), ordinary least squares (OLS), multiple linear regression (MLR), range minimum query (RMQ), feedforward neural network (FNN), item response theory (IRT), particle swarm optimization (PSO), simulated annealing (SA), extreme learning machine (ELM), and artificial neural network (ANN). Other suitable algorithms can also be used. These algorithms are well-known to a person skilled in the art.
In another exemplary embodiment, the present disclosure envisages a method for optimizing polymer properties.
In the first step, input information is received and stored at the repository unit 101.
In accordance with an embodiment of the present disclosure, the input information can include, but is not limited to, polymer properties, real-time parameters, sampling time, and laboratory results. Other suitable input information can also be used.
In a second step, the capturing unit 200 captures polymer plant generated real-time process parameters and sampling time details using 102 and laboratory analysis results using 103 In accordance with an embodiment of the present disclosure, the polymer plant generated real-time process parameters and sampling time details can include but is not limited to, the temperature profile of the reactor, composition of the gaseous reactor effluent, power, current, voltage and torque of the polymerization reactor agitator, the dew point or composition of the gaseous reactor effluent (the typical composition of the reactor effluent is as provided in table 1), hydrogen flow to a reactor, catalyst flow to the reactor, cocatalyst(s) flow to the reactor, catalyst promoter(s) flow to the reactor, recycle gas flow to the reactor and silane flow to the reactor. Other suitable real-time process parameters and sampling time details can also be used.
Table 1: Typical composition variation in Reactor outlet gas
Composition Nitrogen
% vol Hydrogen
% vol Ethane
% vol Propane
% vol propylene
% vol
Minimum 1.7 0.15 0.12 2.64 95.39
Average 4.32 1.59 0.04 1.04 93.01
Maximum 4.64 2.58 0.07 1.93 90.73

In accordance with an embodiment of the present disclosure, the composition of the gaseous reactor effluent can include but is not limited to, condensable and non-condensable at a temperature in the range of 40 °C to 105 °C and a pressure range of 26 kg/cm2g to 30 kg/cm2g. Other suitable compositions of the gaseous reactor effluent can also be used.
In accordance with the present disclosure, the process parameters are used to predict the properties of polymers.
In accordance with an embodiment of the present disclosure, polymer properties can include but is not limited to, Melt Flow Index (MFI) and Xylene Solubility (XS). Other suitable polymer properties can also be used.
In a third step, an operational model 300 is executed to determine the real-time predicted polymer properties at the reactor’s outlet by AI/ML model 105 and validate the directional change of the prediction by first principle model 104, a combination of both can determine the magnitude of change required in the manipulated process parameters of a reactor to bring the property value back to the targeted range in case of a deviation with respect to the targeted property value set by the operator, and accordingly, a signal is sent to a master-slave controller 106 or to an APC system or directly to the salve control element or elements.
In accordance with an embodiment of the present disclosure, with AI/ML model 105 or first principle model 104 can determine the magnitude of the change required in the real time process parameters so that the predicted properties can be within the targeted range.
In accordance with an embodiment of the present disclosure, the real-time predicted properties can fall outside the targeted range, wherein the magnitude of change is determined in the manipulated parameters of the reactor, and accordingly, a signal is sent by operating model 300 to the master controller 106 to bring back the property values within the targeted range.
In accordance with an embodiment of the present disclosure, the operational model can include, but is not limited to, a combination of the first principle model 104 and AI/ML model 105.
In accordance with the present disclosure, the target property range is set by the plant operator depending on the grade of polymer produced.
In accordance with the present disclosure, the master controller 106 acts on a final control element such as a hydrogen, silane or catalyst flow in the polymerization reactor inlet or other suitable process parameters in the polymerization reactor process can be used.
In accordance with an embodiment of the present disclosure, an algorithm for the development of AI/ML model 105 can include, but is not limited to, partial least squares regression (PLS), principal component analysis (PCA), radial basis function (RBF), support vector regression (SVR), ordinary least squares (OLS), multiple linear regression (MLR), range minimum query (RMQ), feedforward neural network (FNN), item response theory (IRT), particle swarm optimization (PSO), simulated annealing (SA), extreme learning machine (ELM), and artificial neural network (ANN). Other suitable algorithms can also be used.
While the ML model predicts the properties of polymers, the first principle model validates the directional change in the predictions. This aspect not only brings in the explainability of the model’s response to input parameters but also minimizes the false positives and negatives in ML’s prediction.
The model of the present disclosure provides a real-time indication of the properties of polymers at the reactor outlet and provides a real-time optimization (RTO) and advanced process control (APC) method by identifying the extent of change required in manipulated variables to attain or be within the targeted polymer property range of a polymerization reactor based on the predicted properties.
In a fourth step, the master-slave controller 106 operates to transmit the signal to the slave element 106a to bring the property value back to the targeted range in case of a deviation.
In a fifth step, a computing unit 400, the comparing unit 107 operated to compare the laboratory analysis results 103 with the predicted values of the AI/ML model 105 for a given sampling time captured by 102 by executing for the same timestamp.
In a sixth step, the AI/ML model 105 is retrained by a training unit 108 and substituted with a better model if the gap between any of the laboratory values and predicted values for the same polymer property continuously exceeds the desired levels.
For the training purpose, the training dataset is retrieved from the repository unit 100 by the training unit 108 and the data set is divided into training and testing sets. The models are accordingly tested. The tested models are screened based on their performance for the different training-testing sets chosen. The best performing model selected is then used to replace the existing AI/ML model 105.
In a seventh step, the desired results are displayed by the display unit 109.
The present disclosure cannot be limited to a form or a type of polymer.
In an exemplary embodiment, the present disclosure provides a system and a method for optimizing polypropylene properties.
The prediction accuracy of the model so developed for two commercial reactors of 220,000 tons per annum polypropylene (polymer) production capacity each is being observed to be within the repeatability and reproducibility range of the MFI and XS laboratory analysis as per ASTM testing method.
The model developed in the present disclosure enables building computer-implemented APC and RTO applications over it. The explainability of the model will aid in generating control actions on a large-scale industrial /commercial reactor input parameters to target the desired polymer grade’s MFI and XS.
The method of the present disclosure will bring up the prime grade material production to greater than 95%. This will account for per year savings in the range of 0.6 million USD to 1 million USD for a 440 KTPA plant.
The final model is chosen and applied for the real-time prediction of the polypropylene properties for two vertical stirred bed gas phase reactors (Reactor 1 and Reactor 2) for Melt Flow Index (MFI) or Melt Flow Rate (MFR), and Xylene Solubility (XS) separately. The predictions are within the lab analysis repeatability and reproducibility range as per ASTM testing methods for the properties MFI, and XS. The predicted values vs lab values for both the properties for reactors 1 and 2 are shown in Figures 2-5.
Referring to Figure 2, a graph of comparison of real time MFI prediction against laboratory analysis results for a commercial stirred bed gas phase reactor 1 is shown, in accordance with an embodiment of the present disclosure.
Referring to Figure 3, a graph of comparison between the real-time prediction of the Xylene Solubility (XS) of the polymer and the laboratory analysed XS of the polymer for a commercial stirred bed gas phase reactor 1 is shown, in accordance with an embodiment of the present disclosure.
Referring to Figure 4, a graph of comparison between the real-time prediction of the Melt flow index (MFI) of the polymer and the laboratory analysed MFI of the polymer for a commercial stirred bed gas phase reactor 2 is shown, in accordance with an embodiment of the present disclosure;
Referring to Figure 5, a graph of comparison between the real-time prediction of the Xylene Solubility (XS) of the polymer and the laboratory analysed XS of the polymer for a commercial stirred bed gas phase reactor 2 is shown, in accordance with an embodiment of the present disclosure;
Referring to Figure 6, a parity plot between the real-time prediction of predicted MFI vs Laboratory generated MFI for a commercial stirred bed gas phase reactor 1 is shown, in accordance with an embodiment of the present disclosure;
Referring to Figure 7, a parity plot between the real-time prediction of predicted XS vs Laboratory generated XS for a commercial stirred bed gas phase reactor 1 is shown, in accordance with an embodiment of the present disclosure;
Referring to Figure 8, a parity plot between the real-time prediction of predicted MFI vs Laboratory generated MFI for stirred bed gas phase reactor 2 is shown, in accordance with an embodiment of the present disclosure;
Referring to Figure 9, a parity plot between the real-time prediction of predicted XS vs Laboratory generated XS for stirred bed gas phase reactor 2 is shown, in accordance with an embodiment of the present disclosure;
Referring to Figure 10, a reactor outlet vapor composition distribution is shown, in accordance with an embodiment of the present disclosure.
The functions described herein may be implemented in hardware, software executed by a processor, firmware, or any combination thereof. If implemented in software executed by a processor, the functions may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Other examples and implementations are within the scope and spirit of the disclosure and appended claims. For example, due to the nature of software, functions described above can be implemented using software executed by a processor, hardware, firmware, hardwiring, or combinations of any of these. Features implementing functions may also be physically located at various positions, including being distributed such that portions of functions are implemented at different physical locations.
The foregoing description of the embodiments has been provided for purposes of illustration and is not intended to limit the scope of the present disclosure. Individual components of a particular embodiment are generally not limited to that particular embodiment, but, are interchangeable. Such variations are not to be regarded as a departure from the present disclosure, and all such modifications are considered to be within the scope of the present disclosure.
PSEUDO CODE
An exemplary pseudo code depicting the function of the present disclosure is as follows –
Class detect
{
Repository()
{
A=algorithm;
P=polymer properties;
R= lab results;
I= plant parameters;
}
Capturing unit ()
{
Call.repository();
Rt= capturing real-time parameter and sampling time details from the repository;
Lr= capturing lab results from the repository;
}
Operational model()
{
Call.Capturing unit ();
Call.Computing model();
Call.Training model();
FPM= Select first principal model 104 and select the parameters based on Rt;
ML= Select AI/ML algorithm 105 and select the parameters based on Rt;
ML model performs the predictionCompute FPM model for direction and magnitude of change in polymer properties with respect to previous timestamp value;
FPM.Computing model();
Validate the change in direction of the prediction;
}
Else
{
Select ML;
If(select algorithm from ML) then
{
Compute ML model for prediction of polymer properties;
ML.Computing model();
ML.Training model();
Train the ML model for accuracy;
}
Else {
Combination of FPM and ML can determine the magnitude of change required in the manipulated process parameters of a reactor to bring the property value back to the targeted range;
}
}
}
Computing model ()
{
Call.Capturing unit ();
Call.Operational model();
Cu=comparing unit;
Tu=Training unit;
If (FPM is detect the desired properties) then
{
Compute the FPM model for selection;
System. out.println(“Select and final FPM Model for determine the magnitude of change required in the manipulated process parameters of a reactor to bring the property value back to the targeted range”);
FPM.Training unit ();
System. out.println(“final FPM Model determine the magnitude of change required in the manipulated process parameters of a reactor to bring the property value back to the targeted range”);
}
Else
{
If (ML is detect the desired properties) then
{
Compute the ML model for selection;
System.out.println(“Select and final AL/ML Model determine the magnitude of change required in the manipulated process parameters of a reactor to bring the property value back to the targeted range”);
ML.Training unit ();
If (ML!= desired model)
{
substitute 105 with a better model;
}
Else
{
System. out.println(“final ML Model determine the magnitude of change required in the manipulated process parameters of a reactor to bring the property value back to the targeted range”);
}
}
}
Master-slave controller()
{
Call.Computing model();
Call.Capturing Unit ();
Call.Operational model();
Check for another runtime parameter from Capturing Unit ;
Select and Return parameters to the operational model;
Transmit the signal to the final control element 106a to bring the property value back to the targeted range in case of a deviation;
}
Display unit()
{
Call.Operational model();
Display the desired result;
System.out.println(“Result”);
}
}
TECHNICAL ADVANCEMENTS
The present disclosure described herein above has several technical advantages including, but not limited to, a system optimization of polymer properties in real–time of a polymer being produced in a reactor and a method thereof that:
• predicts the upcoming polymer properties along with the extent of change required in reactor parameters to sustain the desired polymer properties;
• produce a stable polymer;
• a real-time indication of the polymer properties at the reactor outlet and provides a real-time optimization and advanced process control of the polymer reactor based on the predicted properties;
• uses predictive models for Melt Flow Index (MFI) and Xylene Solubility (XS) of a polymer at the reactor outlet;
• implements a hybrid model (a combination of the first principle model and machine learning/artificial intelligence model) to optimize polymerization process on a large scale;
• a simple and economical method for optimizing the properties of polymer by using a model;
• prevents lump formation in a reactor;
• is scaled for commercial production; and
• provide desired properties of the obtained polymer.

LIST OF REFERENCE NUMERALS
100 - System
200 - Data Capturing Unit
101 - Repository Unit
102 - Real-Time Reactor Parameters Collection Module
103 - Lab Result Collection Module
104 - First Principle Model
105 – Artificial Intelligence Model
106 - Master Controller
106a - Slave Controller
106b - Final Control Element
107 - Comparing Unit
108 - Training Unit
109 - Display Unit
110 - Validation Module
112 - Charging Module
300 - Operational Model
400 - Control System/ Control Unit
1111 - Data Sending Module
2000 – Method
Equivalents
The embodiments herein and the various features and advantageous details thereof are explained with reference to the non-limiting embodiments in the following description. Descriptions of well-known components and processing techniques are omitted so as to not unnecessarily obscure the embodiments herein. The examples used herein are intended merely to facilitate an understanding of ways in which the embodiments herein may be practiced and to further enable those of skill in the art to practice the embodiments herein. Accordingly, the examples should not be construed as limiting the scope of the embodiments herein.
The foregoing description of the specific embodiments so fully reveal the general nature of the embodiments herein that others can, by applying current knowledge, readily modify and/or adapt for various applications such specific embodiments without departing from the generic concept, and, therefore, such adaptations and modifications should and are intended to be comprehended within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology employed herein is for the purpose of description and not of limitation. Therefore, while the embodiments herein have been described in terms of preferred embodiments, those skilled in the art will recognize that the embodiments herein can be practiced with modification within the spirit and scope of the embodiments as described herein.
The use of the expression “at least” or “at least one” suggests the use of one or more elements or ingredients or quantities, as the use may be in the embodiment of the disclosure to achieve one or more of the desired objects or results.
Any discussion of documents, acts, materials, devices, articles or the like that has been included in this specification is solely for the purpose of providing a context for the disclosure. It is not to be taken as an admission that any or all of these matters form a part of the prior art base or were common general knowledge in the field relevant to the disclosure as it existed anywhere before the priority date of this
While considerable emphasis has been placed herein on the components and component parts of the preferred embodiments, it will be appreciated that many embodiments can be made and that many changes can be made in the preferred embodiments without departing from the principles of the disclosure. These and other changes in the preferred embodiment as well as other embodiments of the disclosure will be apparent to those skilled in the art from the disclosure herein, whereby it is to be distinctly understood that the foregoing descriptive matter is to be interpreted merely as illustrative of the disclosure and not as a limitation. ,CLAIMS:WE CLAIM:

1. A system (100) for optimization of polymer properties in real–time of a polymer being produced in a reactor, wherein the reactor being equipped with a plurality of soft sensors, the system comprises:
a data capturing unit (200) configured to capture a plurality of real-time process parameters with a time stamp from the plurality of soft sensors and configured to cooperate with a digital user interface to receive a plurality of target polymer properties and a plurality of lab analysis result parameters;
a repository unit (101), coupled with the data capturing unit 200, configured to receive and store the plurality of real-time process parameters with the time stamp, the plurality of target polymer properties, the plurality of lab analysis result parameters along with a plurality of pre-set rules and a threshold change value;
an operational model (300), coupled with the data capturing unit (200) and the repository unit (101), configured to predict a plurality of polymer properties of the polymer being produced in the reactor, the operational model (300) includes an Artificial Intelligence (AI) unit including a plurality of AI models, each of which being configured to predict a magnitude of change in the polymer properties of the polymer being produced in the reactor, using the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp as input;
a first principle model (FPM) (104) configured to derive an FPM magnitude of change in the polymer properties of the polymer being produced in the reactor by using the pre-set rules on the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp, wherein the FPM comprises:
a validation module configured to compare the FPM magnitude of change with each magnitude of change predicted by the plurality of AI models to shortlist a final magnitude of change closest to the FPM magnitude of change and further configured to identify whether the final magnitude of change is greater than the threshold change value; and
a control system (400) configured to operate a set of reactor hardware components, when the final magnitude of change is identified greater than the threshold value, so as to optimise the polymer properties of the polymer being produced by changing values of the real-time process parameters to maintain the value of the final magnitude of change below the threshold value.
2. The system (100) as claimed in claim 1, wherein the data capturing unit (200) comprises:
a real-time reactor parameters collection module (102) configured to capture a plurality of process parameters and sampling time details in real-time from the plurality of soft sensors installed in the reactor;
a lab result collection module (103) configured to fetch a plurality of laboratory analysis results digitally received from the user interface or a central remote server; and
a data sending module (1111) configured to send the plurality of real-time process parameters with the time stamp and the plurality of lab analysis result parameters to the repository (101) to store.
3. The system (100) as claimed in claim 1, wherein the real-time process parameters include a temperature profile of the reactor, a composition of a gaseous reactor effluent, a power value, a current value, a voltage value, a torque of a polymerization reactor agitator, a dew point or a composition of the gaseous reactor effluent, a hydrogen flow rate in to the reactor, a catalyst flow rate to the reactor, co-catalyst(s) flow rate to the reactor, catalyst promoter(s) flow rate to the reactor, a recycle gas flow rate to the reactor, and a silane flow rate to the reactor.
4. The system (100) as claimed in claim 1, wherein the operational model (300) comprises a deviation value calculation unit configured to receive the final magnitude of change and the plurality of target polymer properties, and further configured to calculate a deviation value for each of the real-time process parameters using the FPM magnitude, when the final magnitude of change is greater than a threshold value.
5. The system (100) as claimed in claim 4, wherein the control system is configured to receive the deviation value for each of the real-time process parameters and further configured to operate the reactor components to achieve the production of a polymer having the target polymer properties.
6. The system (100) as claimed in claim 1, wherein the control system transmits a control signal to a master controller (106), an advanced process control (APC) system, or directly to a slave controller (106a).
7. The system (100) as claimed in claim 6, wherein the master controller (106) is operable to transmit the control signal received from the control system to the slave controller (106a) or a final control element (106b) so as to bring the value of polymer properties back to the target value of the polymer properties when the final magnitude of change is greater than a threshold value.
8. The system (100) as claimed in claim 7, wherein the slave element (106a) or the final control element (106b) is a single process variable or multiple variables which are manipulated based on the control signal from the master-slave controller 106, or a control system that controls the process variable or variables to obtain the targeted process parameters.
9. The system (100) as claimed in claim 2, wherein the system (100) comprises a computing unit (400) having a comparing unit (107) operable to compare the laboratory analysis results captured by the lab result collection module (103) with the magnitude of change in the polymer properties generated by each of the plurality of AI models for a given time stamp and calculate a gap value for each of the plurality of AI models.
10. The system (100) as claimed in claim 9, wherein the system (100) comprises a training unit (108) configured to train a selected AI model for which the gap value of the AI model continuously crosses a pre-set error tolerance range.
11. The system (100) as claimed in claim 1, wherein the system (100) comprises a display unit (109) operable to display the real-time process parameters for each of the real-time process parameters.
12. A method for optimization of polymer properties in real–time of a polymer being produced in a reactor(s), wherein the reactor being equipped with a plurality of soft sensors, the method comprising steps of:
capturing, by a data capturing unit (200), a plurality of real-time process parameters with a time stamp from the plurality of soft sensors and configured to cooperate with a digital user interface to receive a plurality of target polymer properties and a plurality of lab analysis result parameters;

receiving and storing, by a repository unit (101) coupled with the data capturing unit 200, the plurality of real-time process parameters with the time stamp, the plurality of target polymer properties, the plurality of lab analysis result parameters along with a plurality of pre-set rules and a threshold change value;
predicting, by an operational model (300) coupled with the data capturing unit (200) and the repository unit (101), a plurality of polymer properties of the polymer being produced in the reactor, the operational model (300) includes an Artificial Intelligence (AI) unit including a plurality of AI models, each of which being configured to predict a magnitude of change in the polymer properties of the polymer being produced in the reactor, using the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp as input;
deriving, by a first principle model (FPM) (104), an FPM magnitude of change in the polymer properties of the polymer being produced in the reactor by using the pre-set rules on the real-time process parameters with a current time stamp and the real-time process parameters with a previous time stamp;
comparing, by a validation module of the FPM, the FPM magnitude of change with each magnitude of change predicted by the plurality of AI models to shortlist a final magnitude of change closest to the FPM magnitude of change and further configured to identify whether the final magnitude of change is greater than the threshold change value; and
operating, by a control system, a set of reactor hardware components, when the final magnitude of change is identified greater than the threshold value, so as to optimise the polymer properties of the polymer being produced by changing values of the real-time process parameters to maintain the value of the final magnitude of change below the threshold value.

Dated this 16th day of August, 2023

_______________________________
MOHAN RAJKUMAR DEWAN, IN/PA – 25
of R.K.DEWAN & CO.
Authorized Agent of Applicant

TO,
THE CONTROLLER OF PATENTS
THE PATENT OFFICE, AT CHENNAI