Development of an optimal, real-time implementable, energy management strategy for a parallel hybrid two-wheeler : innovation report

Concerns over rising CO2 emissions, poor air quality and energy sustainability have led the transportation sector to investigate technologies for a reduction of emissions. EV, HEV and PHEV are all potential solutions with PHEV gathering attention in certain sec-tors. However, there are significant barriers to wider penetration of this technology, pri-marily variability in fuel economy, the effect on the reliability of other subsystems, and battery lifetime. One important method to address these concerns is through the develop-ment of an intelligent energy management strategy of a PHEV which optimises the trade-offs between fuel economy, factors affecting battery lifetime and engine stop-start reliabil-ity which up until now have not been considered. This work investigates the effect of EMS on p., engine on/off transitions and factors influencing battery ageing. This research programme has the objective of laying down a foundation for future improvements in energy management development and its optimization and provides aid in establishing a basis for comparing energy management controllers, which are real-time capable and thus applicable to current vehicles. Firstly, a modified rule-based strategy based on load levelling was developed which showed a considerable benefit in fuel economy over the basic rule-based strategy, which had not previously been investigated in-depth for a 2-wheeler application. Variations in the system parameters had a significant influence on the fuel economy benefit through load levelling. Secondly, a novel modified Equivalent fuel Consumption Minimization Strategy was developed considering the engine operating point deviation from the optimum line. The modified ECMS strategy referred to as ECMS_LL showed enhanced benefit in fuel economy when compared to the basic ECMS strategy. The third avenue of investigation was focused on an all-inclusive evaluation of this modified ECMS_LL with other state-of-art energy -management strategies concerning battery ageing, engine on/off switching along with fuel economy and charge sustainability. There was considerable improvement in fuel economy of up to 12.98 % and a reduction in the engine on/off transitions of up to 55% while using modified ECMS compared to rule-based strategies. However, this re-sulted in negative potential effects on critical battery parameters influencing battery ageing. There was a detrimental effect on absolute battery temperature and absolute battery throughput, which increased by 11.4 ºC and 8.32 Ah respectively. The results were ana-lysed and found consistent for two different drive cycles and three different powertrain component configurations. Lastly, a unique cost-effective solution of equivalence factor adaptation using learning vector quantitative neural network-based drive cycle recognition was developed for the real-time implementation of ECMS_LL. A maximum of 10.70% fuel benefit was seen by using the LVQNN based estimation of equivalence factors in AECMS_LL over an ECMS_LL using equivalence factor optimized for a single drive cy-cle. In conclusion, the work presented provides a detailed insight into the systematic development of a real-time implementable optimal-control based energy management strategy for the full parallel hybrid two-wheeler application considered. This research also defines a robust evaluation process for energy management strategies, including the effects of fuel economy optimisation on battery life and engine switching while comparing energy management strategies