Probabilistic models of harmonic currents produced by twelve-pulse AC/DC converters

The probabilistic approach is an efficient mean for harmonic prediction since it allows to take into account the randomly varying operating conditions of static converters. Although the six-pulse ac/dc converter has been the subject of probabilistic modeling, little consideration has been given to the 12-pulse ac/dc converter. In high-power applications, such as large rated dc power supplies and large rated dc and ac drives, 12-pulse ac/dc converters are usually preferred since they allow to reduce system disturbances due to harmonic currents. This paper presents an analysis of the random properties of harmonic currents produced by 12-pulse ac/dc converters that operate under random variations of their dc load and voltage unbalance. The probability density functions of the amplitudes and phase angles of the harmonic currents (of orders 12 k/spl plusmn/1 and 12 k-6/spl plusmn/1) are established and validated by results obtained from Monte Carlo simulations.