Search

Due to the repeated switching, the detailed switch-level models of electrical machines coupled with power-electronic converters are computationally expensive and hard to linearize for small-signal frequency-domain analysis. Average-value modeling (AVM) has become an effective tool for small-signal analysis of power electronic systems and time-domain transient studies where the details of switching are not important and can be neglected. Recently, a parametric AVM (PAVM) approach has been developed for machine/diode rectifier systems. This paper extends the parametric approach to the machine/thyristor-controlled-rectifier systems, where the thyristor firing may be referenced to either the line voltages or the rotor position. An analytical average model for this system is also developed based on the recently proposed constant-parameter voltage-behind-reactance synchronous machine model. The new PAVM is compared against the original switching system, as well as the analytical AVM. It is shown that the PAVM can accurately predict both small-signal characteristics and large-signal transients of the original switching system in light and heavy modes, which represents an advantage over the analytical models which are typically implicit. [ABSTRACT FROM AUTHOR]

The 120° voltage source inverter driven brushless dc (BLDC) motors are very common in many applications. This paper extends the previous work and presents an improved dynamic average-value model for such BLDC motor-drive systems. The new model is explicit and uses a proper qd model of the permanent magnet synchronous machine with nonsinusoidal rotor flux. The model utilizes multiple reference frame theory to properly include the back EMF harmonics as well as commutation and conduction intervals into the averaged voltage and torque relationships. The commutation angle is readily obtained from the detailed simulation. The proposed model is demonstrated on a typical industrial BLDC motor with trapezoidal back EMF waveforms. The results of studies are compared with experimental measurements as well as previously established models, whereas the new model is shown to provide appreciable improvement. [ABSTRACT FROM PUBLISHER]