摘要 |
<P>PROBLEM TO BE SOLVED: To solve the problem that divergence occurs between an average value and a command value of a controlled variable when model prediction control is performed in an overmodulation region. <P>SOLUTION: An evaluation function J of an operation state deciding portion 34 evaluates an applied voltage vector to be higher as a difference between predicted currents ide and iqe corresponding to the voltage vectors Vi (i=0 to 7) and command currents idr and iqr is smaller. The voltage vector whose evaluation of the evaluation function J is the highest is set to a next operation state. In phase delay compensators 50 and 60, a processing for attenuating harmonic components of the predicted currents ide and iqe is performed. In the overmodulation region, an input parameter of the evaluation function J is switched from the predicted currents ide and iqe to outputs of the phase delay compensators 50 and 60. <P>COPYRIGHT: (C)2012,JPO&INPIT |