Improved fuel efficiency in hybrid electric vehicles requires a delicate balance between the internal combustion engine usage and battery energy, using a carefully designed energy management control algorithm. Numerous energy management strategies for hybrid electric vehicles have been proposed in literature, with many of these centered on the equivalent consumption minimisation strategy (ECMS) owing to its potential for online implementation. The key challenge with the equivalent consumption minimisation strategy lies in estimating or adapting the equivalence factor in real-time so that reasonable fuel savings are achieved without over-depleting the battery state of charge at the end of the defined driving cycle. To address the challenge, this paper proposes a novel state of charge feedback ECMS controller which simultaneously optimises and selects the adaption factors (proportional controller gain and initial equivalence factor) as single parameters which can be applied in real time, over any driving cycle. Unlike other existing state of charge feedback methods, this approach solves a conflicting multiple-objective optimisation control problem, thus ensuring that the obtained adaptation factors are optimised for robustness, charge sustenance and fuel reduction. The potential of the proposed approach was thoroughly explored over a number of legislative and real-world driving cycles with varying vehicle power requirements. The results showed that, whilst achieving fuel savings in the range of 8.40 −19.68% depending on the cycle, final battery state of charge can be optimally controlled to within ±5% of the target battery state of charge.
Machine Learning Applied to Electrified Vehicle Battery State of Charge and State of Health Estimation: State-of-the-Art
