Download Energy Management System for an Hybrid Electric Vehicle, Using Ultracapacitors and Neural Networks I. INTRODUCCION Electric vehicles have improved their performance and made suitable for commercial and domestic use during the last decades. Nevertheless, pure electric vehicles still have not achieved ranges as good as gas powered conventional vehicles. This problem, due to the low energy density and specific energy contained in most electric batteries compared to that of gasoline, is resolved in hybrid vehicles by combining high energy density of gas or hydrogen, with high efficiency of electric drive systems . Still, gas-based or hydrogen-based energy storage systems, unlike batteries, are unable to accept energy from regenerative braking. For this reason, hybrid systems use an auxiliary energy system that is able to receive regeneration from the drive train and helps as a power cache to be used during peak power periods. This allows using energy systems (gas turbines, fuel cells, etc) with lower power ratings, similar to the mean power consumption, and at the same time increase the overall vehicle efficiency . Two topologies can be distinguished regarding auxiliary energy system connection, parallel and serial connections Fig. 1 shows a serial connection of a generic hybrid vehicle, often used in vehicles running on fuel cells, microturbines, internal combustion engines or some kind of battery with poor specific power.