energy storage system technolo- ... energy sources  are urging many ... The University of Rome âROMA TREâ and ENEA, the Italian National Agency for New.
E l e c t r o n i c s
Multi Input Power Electronic Converter The University of Rome “ROMA TRE” and ENEA, the Italian National Agency for New Technologies, Energy and the Environment, have jointly developed and tested a Multi Input Power Electronic Converter (MIPEC) that is able to manage the bi-directional electric power flowing from a maximum of three different sources to a common output. A Multi Input Power Electronic Converter (MIPEC) able to manage the bidirectional electric power flow from a maximum of three different sources to a common output.
New energy storage system technologies
In order to improve the energyconversion efficiencies in the vehicle, vehicle manufacturers are looking with great interest at new energy storage system technologies and fuel converters. Continued advances and standardization in fuel-cell technology  and emerging methods to generate hydrogen by exploiting renewable energy sources  are urging many researchers to design efficient Zero Emission Vehicles (ZEVs) provided with fuel cells. The main aim is to optimise the conversion chain from the fuel (e.g. hydrogen) to mechanical energy. In order to minimize the conversion losses and recover braking energy, it is common practice to connect the fuel cell (the unique fuel converter in the vehicle) to a pack of batteries. As fuel cells a relative have poor efficiency at low and very high loads, the fuel cell is forced most of the time to operate over its maximum efficiency region and
the batteries have to supply or absorb the energy difference requested by the traction drive. In such a system, the batteries have to deal with power peaks during acceleration or braking phases.
Power Peaks Peak currents are supplied at the expense of lower efficiency (relative to efficiency at lower currents) and an increased heating rate. They can also cause accelerated ageing. Moreover, the operating regions in which the battery is forced to operate are of paramount importance for reliable, efficient operation over a long service life. Narrow operating ranges of the state of charge (SOC) are associated with longer cycling life. In order to minimize the power peaks, additional storage devices may be introduced. Ultracapacitors [3,4] are able to support these peaks due to their high power
density, by storing energy at high efficiency during braking and supplying it during acceleration. Therefore, the vehicle’s energy storage system is a hybrid , as it is formed by storage devices with different power and energy densities and different high-efficiency operating ranges. Ultracapacitors (double layer capacitors, supercapacitors) can be charged and discharged at very high currents rates and are more efficient than batteries at these rates. The two main reasons an ultracapacitor can accept and supply high-rate currents more efficiently  are its lower series resistance (ESR) relative to batteries and its non-Faradic method of storing charge through a fairly ideal double layer capacitance. Thus, ultracapacitors may handle challenging power transients in vehicle power profiles more efficiently and with a more suitable energy content. A pack of batteries and an ultracapacitor AutoTechnology 6/ 2004
E l e c t r o n i c s
tank are able to bridge the gap between the power requested by the traction drive and that supplied by the fuel cell. The sizing of the ultracapacitor tank and the battery pack must be accomplished by carefully evaluating the variable power-to-energy (P/E) ratio requirement, which is closely related to the event’s time constant (its units are 1/s), and the P/E-associated duty cycle of the load profiles over the chosen driving schedules. This approach allows hybrid storage elements to be sized for handling the types of loads for which they are best suited. Figure 1 shows the drive train layout proposed in accordance with this approach: a Multi Input Power Electronic Converter (MIPEC) feeds the traction drive and manages the power flowing from three different sources.
The Power Converter The power converter is the parallel connection of three bi-directional step-up/step-down (boost/ buck) DC-to-DC converters, Figure 2. The joint output terminals provide a voltage-regulated DC power supply to the traction drive. A fuel cell (FC), a battery pack (B) and an ultracapacitor tank (UC) separately feed the three input ter-
two power diodes. Such an assembly is commercially available at reasonable cost and is commonly called a “dual-IGBT power module” or “intelligent power module” (IPM). Each step-up/step-down converter contains an input inductor/capacitor filter, which is required to damp the current ripple in the output circuit of each power source, and an output capacitive filter, which is necessary to limit the voltage ripple at the input terminals of the traction drive (Vlink). Although the MIPEC has been tested with a fuel cell, the fuel converter can be a conventional internal combustion engine (ICE) or a gas turbine that drives an electric generator, as in a series hybrid electric vehicle (HEV). The aim of the MIPEC is to provide • Power sharing among the three power sources in order to meet the power demand of the traction drive • A voltage-regulated DC power supply to the traction drive • Accurate control of the fuel cell and batteries: selection of operating points on the fuel cell and battery voltage/current space • Regulation of the state of
Figure 2: MIPEC circuitry layout.
Figure 1: Vehicle configuration. The MIPEC manages the power flow from a battery storage unit (BSU), an ultracapacitor tank (UC) and a fuel cell (FC) to the traction drive. minals. Of course, the fuel cell supports a unidirectional power flow only, and the step-down topology in its DC-to-DC converter is never driven. Each stepup/step-down converter contains two power switches (IGBTs) and AutoTechnology 6/ 2004
charge (SOC) of the ultracapacitors and batteries The IGBT duty cycles are controlled in order to meet the power demand of the traction drive. In doing this, the MIPEC shares the required power flow among the
Figure 3: Power flow sharing; a) vehicle acceleration, b) vehicle braking.
E l e c t r o n i c s
Figure 4: MIPEC power configuration.
pacitors are continuously monitored to ensure that the current SOC is always kept within preset upper and lower limits. In addition, each of the three input currents and voltages is controlled. Each input current is prevented from exceeding a maximum value and rate. Each input voltage must not fall below a threshold limit or rise above the nominal one. Figures 3a and 3b are a qualitative description of the power flow sharing when the vehicle is accelerating (3a) or braking with a deactivated fuel cell (3b).
On the Test Bench
Figure 5: FC (green, 10A/div) and battery (blue, 10A/div) current transients corresponding to power traction step variations (blue, 20A/div); time 2s/div (zoom on left: 200ms/div, zoom on right: 100ms/div).
The highest possible efficiency
three sources. Such a control strategy is accomplished by taking into account the SOC of the ultracapacitors and the batteries, as well as both the maximum admissible power flow variation and the efficiency map of each power source. The IGBT duty cycle is adjusted for both the fuel cell-fed and the battery-fed DC-to-DC converters, in order to achieve the desired control of the input currents. As a result, the desired operating points over the V-I (voltage/current) fuel cell generator and battery characteristic may be selected. On the other hand, the MIPEC output voltage is adjusted at the desired value by controlling the duty cycle of the ultracapacitor-fed DC-to-DC converter, which assures adequate dynamic response during either acceleration or braking operations of the vehicle. In order to achieve the highest possible efficiency, the fuel cell – if possible in its high-efficiency region – supplies the average energy
required by the vehicle, and the ultracapacitor tank assures load levelling of the peak power demand. The battery pack provides the remaining power required to meet the driving profile. SOC values of the batteries and ultraca-
In order to verify the effectiveness of the DSP-based control strategy, the MIPEC has been tested in a laboratory facility at ENEA Research Centre “Casaccia”, near Rome. The converter inputs have been linked to three sources: a pack of lead/acid batteries, an ultracapacitor tank and a PEM (proton exchange membrane) fuel cell. The converter fed the traction drive, an inverter that supplies a three-phase induction motor. The drive train was sized on the basis of a hypothetical small city car with a total mass of 1350 kg. The traction drive was coupled to a four-quadrant dynamometer whose torque can be freely adjusted to generate the desired load profiles. This complete layout allowed the propulsion system to be tested over any driving cycle. Table 1 shows the main parameters of the fuel cell and hy-
Figure 6: Consecutive DC link current variation from 0A to 60A and from 60A to 0A: Vlink (CH1 blue trace, 50V/div), IUC (CH2 cyan trace, 40A/div), IB (CH3 red trace, 20A/div), IFC (CH4 green trace, 40A/div); time 2s/div. AutoTechnology 6/ 2004
E l e c t r o n i c s
brid storage measured during the testing activity. Table 2 provides data of the traction drive. Figure 5 shows the fuel cell and battery current responses to a step variation of the traction power request. Figure 6 shows the three input currents and the output link voltage, as a result of two consecutive load current variations from 0 to 60 A and back to 0 A. The link voltage is kept constant to the nominal value (216 V) and the three sources provide the current requested, each with its own rate. The New European Driving Cycle is used for the emission certification of light-duty vehicles in Europe. The entire cycle includes four ECE (UDC) segments repeated without interruption, followed by one EUDC segment. The ECE+EUDC test cycle is performed on a chassis dynamometer. The maximum speed of the EUDC cycle is 120 km/h and that of the UDC cycle is 50 km/h. We slightly modified the UDC cycle by forcing the drive train to meet more aggressive accelerations and decelerations in urban driving conditions. For this purpose, we shortened the acceleration and the deceleration time from 42 to 28 seconds and from 34 to 24 seconds, respectively. Figure 7 shows how the power flow is shared in one of these modified urban schedules. by Augusto Di Napoli, Fabio Crescimbini, Luca Solero, Alessandro Lidozzi, University of Rome “ROMA TRE”, Giovanni Pede, Marco Santoro, Manlio Pasquali, ENEA Research Center “Casaccia”.
 P. Van den Bossche, J. Van Mierlo, G. Maggetto “The Fuel Cell Vehicle: Shaping the Future with Standardization”, 20th International Electric Vehicle Symposium and Exposition (EVS 20), November 15 – 19, 2003, Long Beach, CA, U.S.A.  www.enea.it/com/ingl/solar/index.html  A. Burke, M. Miller “Ultracapacitor and Fuel Cell Applications", 20th International Electric Vehicle Symposium and Exposition (EVS 20), November 15 – 19, 2003, Long Beach, CA, U.S.A.  P. Barrade, A. Rufer “High-Power Fast Energy Exchange between Storage Systems: Supercapacitors as AutoTechnology 6/ 2004
energy buffer in transportation systems”, 18th International Electric, Fuel Cell and Hybrid Vehicle Symposium and Exhibition (EVS 18), October 20 - 24, 2001, Berlin, Germany  M. Zolot “Dual-Source Energy Storage – Control and Performance Advantages in Advanced Vehicles”, 20th International Electric Vehicle Symposium and Exposition (EVS 20), November 15 – 19, 2003, Long Beach, CA, U.S.A.  S. Buller, E. Karden, D. Kok, R.W. De Doncker “Modeling the Dynamic Behavior of Supercapacitors Using Impedance Spectroscopy", IEEE Transactions on Industry Applications, Vol. 38, No. 6, November/December 2002, pp. 1622-1626
Figure 7: Power flow sharing over one urban section of the New European Driving Cycle (NEUDC).
Table 1: Operating values of source and storage components during the testing activity
Table 2: Traction drive data