Generating Synthetic Vehicle Speed Records Using LSTM

Vrany, Jiri; Krepelka, Michal; Chumlen, Matej

doi:10.1007/978-3-031-34111-3_12

Cited by 1 publication

(4 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, Vanilla RNNs suffer from both exploding and vanishing gradients when dealing with long-term dependencies. For this purpose, it is better to use a special type of RNN called an LSTM neural network [6,17].…”

Section: Lstmmentioning

confidence: 99%

“…LSTM neural networks' success in time series forecasting includes fields such as travel time [19] and vehicle speed [6,17] prediction, making them a reasonable choice for solving our problem.…”

Section: Lstmmentioning

confidence: 99%

“…Yufang et al [5] showed an approach to predict long-term vehicle speed. Work published in [6] extended this idea, allowing us to generate a vehicle speed profile for an arbitrary trip where an Open Street Map (OSM) and elevation data are available.…”

Section: Introductionmentioning

confidence: 99%

“…This work aims to extend further the work of [6], who suggests the possibility of predicting a vehicle's Controller Area Network (CAN) bus data, such as brake pressure and the percentage of the pressed throttle pedal, along an a priori known trip with available OSM data, elevation data, and synthetic speed.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Synthesizing Vehicle Speed-Related Features with Neural Networks

Krepelka,

Vrany

2023

Vehicles

View full text Add to dashboard Cite

In today’s automotive industry, digital technology trends such as Big Data, Digital Twin, and Hardware-in-the-loop simulations using synthetic data offer opportunities that have the potential to transform the entire industry towards being more software-oriented and thus more effective and environmentally friendly. In this paper, we propose generative models to synthesize car features related to vehicle speed: brake pressure, percentage of the pressed throttle pedal, engaged gear, and engine RPM. Synthetic data are essential to digitize Hardware-in-the-loop integration testing of the vehicle’s dashboard, navigation, or infotainment and for Digital Twin simulations. We trained models based on Multilayer Perceptron and bidirectional Long-Short Term Memory neural network for each feature. These models were evaluated on a real-world dataset and demonstrated sufficient accuracy in predicting the desired features. Combining our current research with previous work on generating a speed profile for an arbitrary trip, where Open Street Map data and elevation data are available, allows us to digitally drive this trip. At the time of writing, we are unaware of any similar data-driven approach for generating desired speed-related features.

show abstract

Section: Lstmmentioning

confidence: 99%