Low-Cost High-Speed Techniques for Real-Time Simulation of Power Electronic Systems; Final rept. 1 Mar 2004-31 Dec 2007

摘要

A research team at the McLeod Institute of Simulation Sciences (MISS) at California State University, Chico (CSUC) has been developing high-speed, real-time simulations of power electronic systems since 1999. At that time, real-time simulators were available mainly for large-scale power utility applications and offered minimum frame times in the region of 50 microS These simulators were customized to specific applications and were expensive (typically from hundreds of thousands to millions of dollars). With the rapid growth of power electronic applications along with the increasing use of higher- frequency, pulse-width modulation (PWM) controllers a need arose for higher- speed, but lower cost, real-time simulators. This was the motivation for the original research effort at Chico. The principal factors offering the potential for success were the availability of alternative computing architectures capable of achieving fast real-time operation, and the recognition that the algorithms commonly used for power system simulation, including those used in real-time simulation, had changed very little over almost four decades. It was felt that a fresh approach to the choice of numerical techniques offered the potential for significant improvement. Real-time simulation, in which simulated time is exactly equal to real time, is required whenever there is a need to interface the simulation to real hardware, to embedded software running at normal speed, or to a human operator. Hardware-in-the-loop (HIL) simulation, which is interpreted here to include embedded controllers in the loop, provides a convenient, safe and economical test environment for hardware components and subsystems. In the power electronics field, real controllers may be interfaced to a real-time simulator of the actual power system for test and evaluation. The controller will normally incorporate one or more embedded processors that execute the control algorithm. A real-time simulator might also be connected.