Distance Until Charge prediction and fuel economy impact for Plug-in Hybrid Vehicles

For a Plug-in Hybrid-Electric Vehicle (PHEV) a close to linear battery depletion profile is near optimal in terms of fuel economy, given that the driven Distance Until Charge (DUC) is known. However the intended driven distance until charge is not always known. We propose a pattern recognition method to predict the DUC based on the key-on time and day observations and the on-board stored data in vehicle. We review the energy management strategy and we show a fuel economy improvement despite prediction errors.