Physics-Informed Neural State Space Models via Learning and Evolution
This work addresses the challenge of determining optimal structure and optimization for physics-informed neural networks when complete prior physical knowledge is unavailable, which is a problem for engineers and scientists modeling complex dynamical systems.
This paper explores methods for discovering neural state space dynamics models for system identification, specifically for physical systems. It uses an asynchronous genetic search algorithm to find accurate and physically consistent models for an aerodynamics body, a continuous stirred tank reactor, and a two-tank interacting system.
Recent works exploring deep learning application to dynamical systems modeling have demonstrated that embedding physical priors into neural networks can yield more effective, physically-realistic, and data-efficient models. However, in the absence of complete prior knowledge of a dynamical system's physical characteristics, determining the optimal structure and optimization strategy for these models can be difficult. In this work, we explore methods for discovering neural state space dynamics models for system identification. Starting with a design space of block-oriented state space models and structured linear maps with strong physical priors, we encode these components into a model genome alongside network structure, penalty constraints, and optimization hyperparameters. Demonstrating the overall utility of the design space, we employ an asynchronous genetic search algorithm that alternates between model selection and optimization and obtains accurate physically consistent models of three physical systems: an aerodynamics body, a continuous stirred tank reactor, and a two tank interacting system.