Time Figure 1. Interpretable Physics Models. Consider the sequences shown above. Not only we can predict the future frames of collisions but we can also predict the underlying factors that lead to such an inference. For example, we can infer the mass of cylinder is much higher in second sequence and therefore it hardly moves in the image. Our ability to infer meaningful underlying latent factors inspires us in this paper to learn an interpretable intuitive physics model.
AbstractHumans have a remarkable ability to use physical commonsense and predict the effect of collisions. But do they understand the underlying factors? Can they predict if the underlying factors have changed? Interestingly, in most cases humans can predict the effects of similar collisions with different conditions such as changes in mass, friction, etc. It is postulated this is primarily because we learn to model physics with meaningful latent variables. This does not imply we can estimate the precise values of these meaningful variables (estimate exact values of mass or friction). Inspired by this observation, we propose an interpretable intuitive physics model where specific dimensions in the bottleneck layers correspond to different physical properties. In order to demonstrate that our system models these underlying physical properties, we train our model on collisions of different shapes (cube, cone, cylinder, spheres etc.) and test on collisions of unseen combinations of shapes. Furthermore, we demonstrate our model generalizes well even when similar scenes are simulated with different underlying properties.
This paper presents the locomotive traction controller performance with respect to the track wear under different operation conditions. In particular, an investigation into the dynamic response of a locomotive under changing wheel-rail friction conditions is performed with an aim to determine the effect of controller setting on track wear. Simulation using a full-scale longitudinal-vertical locomotive dynamic model shows that the appropriately designed creep threshold, controller, settings can effectively maintain a high tractive effort while avoiding excessive rail damage due to wear, especially during acceleration under low speed.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.