{"4cb785e55420811695b8536b20a6cd38alexbaier":{"DOI":"","ISBN":"","ISSN":"","URL":"http://arxiv.org/abs/2103.06727","abstract":"Physical motion models offer interpretable predictions for the motion of\r\nvehicles. However, some model parameters, such as those related to aero- and\r\nhydrodynamics, are expensive to measure and are often only roughly approximated\r\nreducing prediction accuracy. Recurrent neural networks achieve high prediction\r\naccuracy at low cost, as they can use cheap measurements collected during\r\nroutine operation of the vehicle, but their results are hard to interpret. To\r\nprecisely predict vehicle states without expensive measurements of physical\r\nparameters, we propose a hybrid approach combining deep learning and physical\r\nmotion models including a novel two-phase training procedure. We achieve\r\ninterpretability by restricting the output range of the deep neural network as\r\npart of the hybrid model, which limits the uncertainty introduced by the neural\r\nnetwork to a known quantity. We have evaluated our approach for the use case of\r\nship and quadcopter motion. The results show that our hybrid model can improve\r\nmodel interpretability with no decrease in accuracy compared to existing deep\r\nlearning approaches.","annote":"","author":[{"family":"Baier","given":"Alexandra"},{"family":"Boukhers","given":"Zeyd"},{"family":"Staab","given":"Steffen"}],"citation-label":"baier2021hybrid","collection-editor":[],"collection-title":"","container-author":[],"container-title":"","documents":[],"edition":"","editor":[],"event-date":{"date-parts":[["2021"]],"literal":"2021"},"event-place":"","id":"4cb785e55420811695b8536b20a6cd38alexbaier","interhash":"a135e9dacfd0b251d72027ff6e0bc0c1","intrahash":"4cb785e55420811695b8536b20a6cd38","issue":"","issued":{"date-parts":[["2021"]],"literal":"2021"},"keyword":"EXC2075 deeplearning machinelearning myown pn4 prePrint simtech","misc":{"language":"English"},"note":"cite arxiv:2103.06727","number":"","page":"","page-first":"","publisher":"","publisher-place":"","status":"","title":"Hybrid Physics and Deep Learning Model for Interpretable Vehicle State Prediction","type":"manuscript","username":"alexbaier","version":"","volume":""}}