Description:PLEASE SEE THE ATTACHMENTS. , Claims:We Claim:
1. A method for predicting intercolumnar cracks in a casting slab, the method comprising:
analysing a plurality of parameters related to the casting slab using a pre-trained Reduced Order Model (ROM) for determining a first set of observations related to the intercolumnar cracks in the casting slab; and
performing a multi-stage analysis of the plurality of parameters using a pre-trained hybrid model for predicting the intercolumnar cracks, the multi-stage analysis comprising:
a first stage analysis of determining a severity index of the intercolumnar cracks based on analysis of the plurality of parameters; and
a second stage analysis of determining a second set of observations related to the intercolumnar cracks based on the plurality of parameters, the first set of observations and the severity index.

2. The method of claim 1, wherein the plurality of parameters comprises at least one of steel composition, grade information, super heat, mold parameters, casting conditions, spray zone parameters, roll force, torque, cooling parameters and casting speed.

3. The method of claim 1, wherein the method comprises pre-processing the plurality of parameters prior to analysis by the pre-trained ROM.

4. The method of claim 3, wherein the pre-processing comprises:
analysing values of each of the plurality of parameters for eliminating outliers in the plurality of parameters; and
structuring the plurality of parameters for analysis by the pre-trained ROM and the pre-trained hybrid model.
5. The method of claim 1, wherein the first set of observations comprises at least one of a thermal profile of the slab, a mechanical profile of the slab and a solidification profile of the slab.

6. The method of claim 1, wherein the second set of observations comprises at least one of a location of the intercolumnar cracks, information of a segment of the slab comprising the of the intercolumnar cracks, and information of parameters causing the formation of intercolumnar cracks.

7. The method of claim 1, wherein the method comprises configuring the pre-trained ROM for:
generating a three-dimensional digital twin of the casting slab to perform a real-time emulation of the casting slab; and
generating one or more visual representations related to a real-time operation of the casting slab.

8. The method of claim 1 further comprises rendering the second set of observations determined by the hybrid model and visual representations generated by the ROM on a User Interface (UI) associated with an expert computing system.

9. A system for predicting intercolumnar cracks in a casting slab, the system comprising:
a data reception module configured to collect a plurality of parameters related to the casting slab; and
a controller configured to:
analyse a plurality of parameters related to the casting slab using a pre-trained Reduced Order Model (ROM) to determine a first set of observations related to the intercolumnar cracks in the casting slab; and
perform a multi-stage analysis of the plurality of parameters using a pre-trained hybrid model to predict the intercolumnar cracks, the multi-stage analysis comprising:
a first stage analysis to determine a severity index of the intercolumnar cracks based on analysis of the plurality of parameters; and
a second stage analysis to determine a second set of observations related to the intercolumnar cracks based on the plurality of parameters, the first set of observations and the severity index.

10. The system of claim 9, wherein the plurality of parameters comprises at least one of steel composition, grade information, super heat, mold parameters, casting conditions, spray zone parameters, roll force, torque, cooling parameters and casting speed.

11. The system of claim 9, wherein the controller is configured to pre-process the plurality of parameters prior to analysis by the pre-trained ROM.

12. The system of claim 11, wherein, to pre-process the plurality of parameters, the controller is configured to:
analyse values of each of the plurality of parameters to eliminate outliers in the plurality of parameters; and
structure the plurality of parameters for analysis by the pre-trained ROM and the pre-trained hybrid model.
13. The system of claim 9, wherein the first set of observations comprises at least one of a thermal profile of the slab, a mechanical profile of the slab and a solidification profile of the slab.

14. The system of claim 9, wherein the second set of observations comprises at least one of a location of the intercolumnar cracks, information of a segment of the slab comprising the of the intercolumnar cracks, and information of parameters causing the formation of intercolumnar cracks.

15. The system of claim 9, wherein the pre-trained ROM is configured to:
generate a three-dimensional digital twin of the casting slab to perform a real-time emulation of the casting slab; and
generate one or more visual representations related to a real-time operation of the casting slab.

16. The system of claim 9, wherein the processor is further configured to render the second set of observations determined by the hybrid model and visual representations generated by the ROM on a User Interface (UI) associated with an expert computing system.