Scenario Uncertainty Modeling for Predictive Maintenance with Recurrent Neural Adaptive Processes (RNAPs)

2021-01-0191

04/06/2021

Features
Event
SAE WCX Digital Summit
Authors Abstract
Content
For commercial-vehicle Original Equipment Manufacturers (OEMs), predictive maintenance has drawn attention for the benefits of money saving and increased road safety. Data-driven models have been widely explored and implemented as predictive maintenance solutions. However, the working scenarios for different commercial-vehicles vary a lot, which makes it difficult to build a universal model suitable for all the cases. In this paper, we propose a Recurrent Neural Adaptive Processes (RNAPs) network to adapt to different scenarios by modeling the uncertain at the same time. The ensemble network combines the traits of neural processes, recurrent neural network and meta learning together. Neural processes consider the context information to calculate the uncertainty and improve the prediction results. Meta-learning works well when dealing with few-shot multi-tasks learning, and recurrent networks are utilized as the encoder of the proposed model to process time-series data. We also investigate how to utilize the proposed model for regression and classification tasks of predictive maintains projects. At last we show that the ensemble model could improve the accuracy when the data are highly various and imbalanced by testing the models on two detailed tasks using real world data.
Meta TagsDetails
DOI
https://doi.org/10.4271/2021-01-0191
Pages
8
Citation
Yu, W., Zhao, X., Sun, Y., and Li, X., "Scenario Uncertainty Modeling for Predictive Maintenance with Recurrent Neural Adaptive Processes (RNAPs)," SAE Technical Paper 2021-01-0191, 2021, https://doi.org/10.4271/2021-01-0191.
Additional Details
Publisher
Published
Apr 6, 2021
Product Code
2021-01-0191
Content Type
Technical Paper
Language
English