Instantaneous vehicle fuel consumption estimation using smartphones and recurrent neural networks

作者：

Highlights：

• Estimation of instantaneous vehicle fuel consumption using smartphone measurements.

• Deep Neural Network estimation assessment and comparison to other methods.

• Deep Neural Network training evaluation and proposal of a novel hybrid training technique.

• Performance loss investigation for less information-rich inputs.

摘要

•Estimation of instantaneous vehicle fuel consumption using smartphone measurements.•Deep Neural Network estimation assessment and comparison to other methods.•Deep Neural Network training evaluation and proposal of a novel hybrid training technique.•Performance loss investigation for less information-rich inputs.

论文关键词：Recurrent neural networks,Vehicle fuel consumption,Smartphone data,Training,Optimisation,Soft sensor

论文评审过程：Received 28 March 2018, Revised 10 November 2018, Accepted 2 December 2018, Available online 4 December 2018, Version of Record 8 December 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.12.006