Identification of Dynamical Systems using Symbolic Regression
This work addresses the challenge of model identification for dynamical systems, which is incremental as it enhances existing symbolic regression methods by integrating gradient-based optimization.
The authors tackled the problem of identifying dynamical system models from observational data by combining symbolic regression with gradient-based optimization of ODE parameters, resulting in improved predictive accuracy, with testing on 19 problem instances showing that fine-tuning parameters by fitting to observed values yields the best results.
We describe a method for the identification of models for dynamical systems from observational data. The method is based on the concept of symbolic regression and uses genetic programming to evolve a system of ordinary differential equations (ODE). The novelty is that we add a step of gradient-based optimization of the ODE parameters. For this we calculate the sensitivities of the solution to the initial value problem (IVP) using automatic differentiation. The proposed approach is tested on a set of 19 problem instances taken from the literature which includes datasets from simulated systems as well as datasets captured from mechanical systems. We find that gradient-based optimization of parameters improves predictive accuracy of the models. The best results are obtained when we first fit the individual equations to the numeric differences and then subsequently fine-tune the identified parameter values by fitting the IVP solution to the observed variable values.