Probabilistic analysis of energy exchange processes for automotive hybrid powertrain

In this paper a statistical method for establishing the minimum energy reservoir needed for a hybrid-electric vehicle (HEV) is proposed. This method is based on real world data and investigates the statistical properties of the charge and discharge events for a variety of driving profiles. The distribution of the magnitude of discharge and charge events was found to be exponentially distributed. Using an exponential distribution assumption, the probability distribution for the energy stored in the battery was calculated. This distribution is a function of only two parameters, the average discharge energy (μd), and the ratio of the average discharge energy to the average charge energy(ζ). These parameters are functions of the drive profiles of interest and the power-level of the HEV power supply. Based on this, a strategy HEV battery capacity and the power-level of the primary power supply was proposed. This strategy is of particular importance for fuel cell based HEVs because the cost of the fuel cell stack directly scales with the power level.

► We analyzed HEV battery charge/discharge needs using real world data.
► We developed a statistical model of HEV battery energy levels.
► HEV battery energy distribution becomes unreliable below an engine size threshold.
► Above this threshold required battery size is independent of engine size.
► We propose a strategy for sizing HEV power train components.