Theoretical Framework

Some system theory terminology must be understood.

Machine learning techniques, such as multilayer perceptron (MLP), and classical system theory are leveraged. Therefore, familiarity with system theory terminology, such as the concepts of states and state variables, is expected.

Given an arbitrary physical problem, modeled with an arbitrary software, the goal of romAI is to turn the problem into a ROM block that includes:
  • Inputs and Outputs
  • State Variables
  • Characteristic Parameters
This is illustrated in the following image.