LabVIEW FPGA Control of an Electric Vehicle Converter with Ultracaps and Batteries

[BMac]

Learn about the design of a high-frequency multi-phase DC-to-DC converter for an electric vehicle hybrid energy storage system (HESS) that combines the high specific power (kW/kg) of ultra-capacitors with the high specific energy (Wh/kg) of Lithium-Ion batteries. The 30 kW bidirectional DC-DC converter operates in uses advanced system optimization algorithms executing on the real-time processor of the NI CompactRIO system to determine the reference current for optimal charge and discharge of the ultra-capacitors and batteries during the drive cycle. A liquid cooled, 4-phase, two quadrant DC-DC converter with interleaving for reduced ripple current and high frequency operation is designed and simulated using NI Multisim using LabVIEW FPGA co-simulation. To maximize the energy efficiency, a variable frequency PWM control scheme is implemented along with phase shedding that enables the FPGA-based control system to drop out half-bridge phases during low power operation.

Watch video now.

The new NI Multisim co-simulation tools are used to design the LabVIEW FPGA power electronics control code including the logic for startup and fault mode conditions. Unlike the traditional approach to model based design, this avoids the need to create a separate model of the control system for the purposes of simulation, which then must be converted to register level code for implementation in the FPGA. Instead, the LabVIEW FPGA controller code is co-simulated with the schematics in Multisim. LabVIEW graphical programming tools for the FPGA completely eliminate the need for any knowledge of VHDL or Verilog programming.

The Multisim power electronics simulation schematic is a high fidelity model that includes parasitics including capacitor, battery and cabling ESR. Using the new variable-timestep co-simulation tools, the converter schematic is modeled in continuous time along with the actual discrete time, fixed-point LabVIEW FPGA code, including all quantization effects.

During co-simulation, a user friendly graphical user interface (GUI) is developed for the FPGA-based control system. The LabVIEW front panel GUI and LabVIEW FPGA control code is completely re-used when transferred to the NI CompactRIO system for operation of the physical power converter. The NI toolchain enables the identical embedded code to exist simultaneously in both a desktop simulation and embedded system deployment context. The user interface enables the PWM switching frequency to be adjusted on the fly during operation along with control gain tuning and transfer between converter operating modes. All kinds of diagnostic information about the currents, temperature, voltages and fault conditions are displayed on a line-diagram of the converter circuit.

The LabVIEW FPGA control system includes digital PI compensator, phase balancing algorithms, and frequency control look up table (LUT) logic.  Co-simulation results show single-phase and four-phase load step responses from 0 to 50 Amps, based on a range of ultracapacitor voltages. We see a very good match between co-simulation results and physical measurements from the experimental system.

A 3D rendering of the converter assembly mechanical design is shown, including the logic board, capacitors, IGBT models and drivers, and the NI CompactRIO control system. The CompactRIO control system also acts as a data logger, enabling analysis of the experimental converter dynamic response give slow (10 s) and fast (250 ms) step-changes to the ultra-capacitor charge/discharge current commands. The speed of the data logging system enables us to zoom in on the time-response measurements to verify control stability in all operating conditions.

The Multisim and LabVIEW FPGA co-simulation platform was very useful for developing and validate the control software in a short period of time by a student with no knowledge of HDL and no prior LabVIEW programming experience, and eliminated the need to create a separate model of the control software. However, due to the inherently long simulation times for high fidelity modeling of SMPS and 80 MHz FPGA control loops, we did find some challenges in using the exact same control code. Next step is to port the control to the NI sbRIO GPIC.

View the entire IEEE Spectrum webcast: http://spectrum.ieee.org/webinar/2118882

Visit the Project EVE website: http://www.projecteve.ca/

About the Speaker:

Dr. Olivier Trescases (B.A.Sc, M.A.Sc., Ph.D.) received his Ph.D. degree in electrical engineering at the University of Toronto in 2007. He has published over 40 papers in IEEE conferences and power electronics journals. He received two IEEE best-paper awards, one IEEE Vehicular Technology award, as well as the 2010 Green Innovation Award from the City of Toronto. From 2007 to 2009, he worked as a concept engineer and mixed-signal integrated circuit designer at Infineon Technologies in Austria. Dr. Trescases joined the University of Toronto as an Assistant Professor in January 2009, where he conducts research on power electronics for automotive, industrial, aerospace and renewable energy applications.