This library provides the implementation in PyTorch of the Neural Integration for Constitutive Equations (NICE) method. The algorithms in this repository are implemented using torch and torchdiffeq libraries, thus are fully supported to run on GPU.
The NICE (Neural Integration for Constitutive Equations) method is a novel deep learning tool for the automatic discovery of constitutive equations from small data - partial and incomplete material state observations. The approach builds upon the solution of the initial value problem describing the time evolution of the material state and leverages the framework provided by neural differentials equations. NICE can learn accurate, consistent, and robust constitutive models from incomplete, sparse, and noisy data collecting simple conventional experimental protocols.