• cell thermal model
• eledtrical equivalent cell model
• hybrid power-train
• hybrid-electric vehicle
• lithium battery

Batteries are used on-board vehicles for broadly two applications – starting-lighting-ignition (SLI) and vehicle traction. This thesis examines the suitability of the rechargeable lithium battery for both these applications, and develops algorithms for runtime prediction of the remaining battery charge.
The largest market-share of rechargeable batteries is for the SLI application. Lead-acid batteries rule this market presently, although a handful of lithium SLI batteries have recently appeared on the market. The practicality of different lithium battery chemistries has been evaluated for this application over wide-ranging criteria and it has been found that the batteries based on lithium iron phosphate and lithium titanate oxide chemistries commercially available in the market are the most suitable. Lithium SLI batteries would require a higher initial cost and additional electronic hardware in the form of battery management and thermal management systems, but would last the life-time of the vehicle. In fact, with the decrease in the cost of lithium SLI batteries with higher volumes, over the life-time of the vehicle, the total costs of the existing lead-acid battery and the lithium battery would be about the same.
The electric traction application is probably the most demanding of all battery applications and imposes the harshest requirements on the battery cells and the battery management system. Algorithms to manage the battery cells for consumer power electronics, for example, do not perform satisfactorily for the electric traction application. This thesis presents algorithms to accurately determine the remaining charge of a lithium battery cell during runtime on-board the vehicle. The algorithm changes slightly depending upon the type of lithium chemistry and could be used in conjunction with different power management strategies on a vehicle with electric traction — whether a pure electric vehicle, hybrid electric vehicle (HEV) or a plug-in hybrid electric vehicle. An accurate estimate of battery charge is important for the battery management system; allows the battery pack to be used more efficiently, reliably and safely; and also provides a reasonably accurate estimate of the remaining distance that could be travelled to the driver. It also prevents over-charging or over-discharging the battery, which are detrimental to its life, and provides an indication when the battery would need to be replaced.
The central contribution of this thesis is in developing an algorithm based on an electrical equivalent circuit model of a rechargeable lithium cell that includes thermal dependence, is accurate, yet simple enough to require low on-board processing capacity. The algorithm has been validated through extensive experimental tests for the lithium nickel-manganese-cobalt and the lithium iron phosphate chemistries at the University of Pisa labs. The algorithm was also successfully implemented using an adaptive state estimator (extended Kalman filter) for overcoming the difficulties imposed by the lithium iron phosphate chemistry. The algorithm was also developed into a model in collaboration with Mathworks for their toolbox and shall be commercially launched later this year. The model algorithm also forms the core of the battery management algorithm for the European Union’s Hybrid Commercial Vehicle (HCV) Project for future HEV trucks and buses by Volvo, Iveco, Daf and Solaris.
The model was also used (as part of a battery model) for hybridizing the power-train of passenger buses for Bredamenarinibus (an Italian bus manufacturer) through modelling and simulation. The conventional power trains of three different buses representing different market segments were hybridised using a series-hybrid electric architecture and simulated with different power management strategies over different types of duty cycles, including real-life duty cycles provided by the manufacturer. Even with the increased weight of the hybrid buses (due to additional batteries and electrical equipment) the simulation predicts fuel savings between 22 to 25% depending upon the power management strategy for the hybrid buses. The prototypes of these series-hybrid buses are under production and would be tested in different Italian cities this year, before entering commercial production.