Sensorless control of a permanent magnet machine/gasoline engine using an extended Kalman filter

An extended Kalman filter is used to estimate shaft position and speed of a gasoline engine/permanent magnet generator during the running mode of operation. Simulations are performed to derive pertinent filter coefficients and to verify satisfactory operation. As an intermediate step to full hardware implementation, the Kalman estimator is realized in digital hardware and is interfaced to the software plant model. The systematic migration from software simulation to hardware highlights the influence of the various error sources. Finally, results are presented which verify operation of the hardware and which examine the assumption that the error sources are white noise