UC Berkeley

This thesis describes the design of control algorithms, modelling, testing, and analysis of use cases of the INSTAR system, a dedicated high-power, low cost energy storage system designed to increase energy efficiency and performance in hybrid vehicles, with special application in urban commercial vehicles. The system is intended to be combined with an electrochemical battery and small energy generator to provide 3 distinct magnitudes of power delivery and absorption to the traction motors. The advantages of such a system would be decreased generator size, decreased component size, and potentially increased battery service life.

Flywheel control algorithms were developed and tested to absorb excessive electrical energy during regenerative braking, as well as to augment electrochemical battery power during vehicle accelerations. During laboratory testing battery charging currents reaching the battery were controlled through modulation of flywheel throttle signals. These algorithms were applied in road driving conditions to decrease vehicle energy use by 8.9 percent over a predetermined acceleration and deceleration event while decreasing the charging rate of the battery. Laboratory testing showed potential to reduce battery discharge rate during acceleration events by pre-charging the flywheel so that it may be discharged during the acceleration event. Road tests demonstrated that energy use could be reduced by 2.6% over a pre-determined acceleration and deceleration event while simultaneously reducing the magnitude of battery discharge current by around 15%.

Such a system has use in urban vehicles where much of the vehicle driving is in stop and go type traffic. These vehicles typically have a large internal combustion engine to accelerate the vehicle from a stop. The INSTAR system may be able to reduce the engine output requirement while still providing the same performance. The INSTAR system is particularly applicable in commercial vehicles where fully electric vehicle may not be practical.