This invention describes a system for a plug-in hybrid electric vehicle that uses internet cloud connectivity to improve its energy efficiency. It works by collecting vehicle and trip data, sending it to the cloud for offline training of an automated driving system to predict vehicle movements. This prediction helps an automated powertrain control system manage the engine, electric motor, and transmission to minimize wasted energy during a trip, specifically using a non-linear autoregressive recurrent neural network for training.
Why it matters: Since 2020, advancements in AI/ML model efficiency and cloud computing infrastructure have made the large-scale, offline training of complex neural networks for vehicle control more practical. This could simplify the development and deployment of the specified non-linear autoregressive recurrent neural network for energy optimization.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI