Energy Management System For Controlling Series Hybrid Electric Motorcycle

Pollution issues and scarcity of fossil fuel inspire the development of hybrid electric vehicle. Motorcycles are widely used in developing countries and Asia for their size, cost, and maneuverability. They create enormous pollutants due to the lack of viable pollution prevention technologies. There are plenty of research on hybrid cars, but very limited literature on hybrid motorcycle, thus, the behavior and performance of hybrid motorcycle are not completely known. Hybridizing conventional motorcycle is necessary because of the increasing usage due to the population growth and rising living standard and these can bring about disastrous climate change if current habit persisted. One of the problems that remain unsolved in hybrid motorcycle is the prediction of the future trip. Various techniques have been used for the prediction, but these are either too complex, expensive, or performed poorly. This research improves the performance of an electric motorcycle by hybridization where the performance of the building blocks for hybrid motorcycle were studied and characterized. Via dynamic programming simulation, efficient use of hybrid motorcycle was found. The characteristics identified from the dynamic programming were then used for the formulation of the energy management system. Kalman filtering was applied to the energy management system to pretreat the signals measured from the traffic. Kalman filter requires only 2 kB when implemented with Atmel ATmega328p compared to 10 kB required by simple moving average filter. The series hybrid electric motorcycle embedded with the energy management system achieves 89.58 km per charging compared to 19.30 km per charging for the electric motorcycle under the modified ECE-R40 drive cycle. In addition, the energy management system outperformed the conventional thermostat control strategy in terms of traveling distance and it has more optimized fuel usage. The energy management system proposed achieves above 80 % performance of the dynamic programming approach, for long traveling distance, it achieves as high as 98.06 %. Tuning and adaptation of the control algorithm had been demonstrated so developers can make use of them for their applications. Several contributions are made: electromagnetic torque of brushless DC motor can be estimated based on the single-phase current sensing. The mathematical models developed for subsystem components and the experimental techniques are invaluable for hybrid motorcycle developers. Besides, efficient series hybrid electric motorcycle performance is obtainable with simple and efficient control algorithm developed