Multiparameter Real-World System Identification Using Iterative Residual Tuning

Abstract

Abstract In this work we consider the problem of nonlinear system identification, using data to learn multiple and often coupled parameters that allow a simulator to more accurately model a physical system and close the so-called reality gap for more accurate robot control. Our approach uses iterative residual tuning (IRT), a recently-developed derivative-free system identification technique that utilizes neural networks and visual observation to estimate parameter differences between a proposed model and a target model. We develop several modifications to the basic IRT approach and apply it to the system identification of a 5-parameter model of a marble rolling in a robot-controlled labyrinth game mechanism. We validate our technique both in simulation — where we outperform two baselines — and on a real system, where we achieve marble tracking error of 4.02% after just 5 optimization iterations.