Preventing Catastrophic Forgetting using Prior Transfer in Physics Informed Bayesian Neural Networks

Abstract

Predictive models can be integrated in the sensing and monitoring methodologies of mechatronic systems in operation. When systems change or are subject to varying operating conditions, adaptivity of the models is needed. The goal of this paper is to enable this adaptivity by presenting a framework for continual learning. The framework aims to transfer and remember information from previously learned systems when a model is updated to new operating conditions. We achieve this by means of the following three key mechanisms. We first include physical information about the system, heavily regularizing the model output. Secondly, the usage of epistemic uncertainty, used as an indicator of the changing system, shows to what extend a transfer is desired. Last but not least the usage of a prior within a Bayesian framework allows to regularize models further according to previously obtained information. The last two principles are enabled thanks to the use of Bayesian neural networks. The methodology will be applied to a camfollower system in a simulation environment, where results show that previously trained systems are better remembered with an increase of 72% compared to normal training procedures.

Optimized Magnetic Particle Navigation Based on Data-Driven Model for Enhanced Drug Targeting

“…We need to apply the regression models to realistic settings in 2D and 3D, taking into account flow rates and wall effects. Advanced learning algorithms can be incorporated to transfer a prior model to more realistic setups and conditions [65]. Finally, we should make the control algorithm compatible with particle imaging techniques to enhance the goalreaching over longer distances (for example through model-based predictive control [66]).…”