Methods and systems for constructng and simulating a digital twin of physical model

A system, comprising:

• a memory storing instructions;
• one or more input/output (I/O) interfaces;
• and
• one or more hardware processors coupled to the memory via the one or more I/O interfaces, wherein the one or more hardware processors are configured by the instructions to:
• receive via the one or more I/O interfaces, a domain knowledge, one or more company specific datasets, and one or more company specific documents associated with a physical model of a company, from a resource repository, wherein the domain knowledge comprises one or more domain knowledge resource documents and one or more insight documents;
• construct an agent meta model of the physical model from the domain knowledge using an agent-based modelling framework, wherein the agent meta model of the physical model comprises (i) one or more components, (ii) one or more subcomponents associated with each of the one or more components, and (iii) one or more relationships between the one or more components and between the one or more subcomponents associated with each of the one or more components;
• construct a digital twin of the physical model by instantiating and populating the agent meta model of the physical model, using the one or more company specific datasets, wherein the digital twin of the physical model comprises (i) the one or more components with one or more component values, (ii) the one or more subcomponents associated with each of the one or more components along with one or more sub component values, and (iii) the one or more relationships between the one or more components and between the one or more subcomponents associated with each of the one or more components along with one or more relationship values;
• validate a structure of the digital twin of the physical model through one or more automated structural validation techniques, to obtain a validated digital twin of the physical model;
• synchronize the validated digital twin of the physical model with a current status data of the physical model, to obtain a synchronized digital twin of the physical model;
• and
• simulate the synchronized digital twin of the physical model with one or more simulation scenarios, using a trained reinforcement learning (RL) agent, wherein the one or more simulation scenarios are associated with one or more target goals of the physical model.

One or more non-transitory machine-readable information storage mediums comprising one or more instructions which when executed by one or more hardware processors cause:

• receiving a domain knowledge, one or more company specific datasets, and one or more company specific documents associated with a physical model of a company, from a resource repository, wherein the domain knowledge comprises one or more domain knowledge resource documents and one or more insight documents;
• constructing an agent meta model of the physical model from the domain knowledge using an agent-based modelling framework, wherein the agent meta model of the physical model comprises (i) one or more components, (ii) one or more subcomponents associated with each of the one or more components, and (iii) one or more relationships between the one or more components and between the one or more subcomponents associated with each of the one or more components;
• constructing a digital twin of the physical model by instantiating and populating the agent meta model of the physical model, using the one or more company specific datasets, wherein the digital twin of the physical model comprises (i) the one or more components with one or more component values, (ii) the one or more subcomponents associated with each of the one or more components along with one or more sub component values, and (iii) the one or more relationships between the one or more components and between the one or more subcomponents associated with each of the one or more components along with one or more relationship values;
• validating a structure of the digital twin of the physical model through one or more automated structural validation techniques, to obtain a validated digital twin of the physical model;
• synchronizing the validated digital twin of the physical model with a current status data of the physical model, to obtain a synchronized digital twin of the physical model;
• and
• simulating the synchronized digital twin of the physical model with one or more simulation scenarios, using a trained reinforcement learning (RL) agent, wherein the one or more simulation scenarios are associated with one or more target goals of the physical model.