A quantitative comparison approach for different voltage dip characterization methods

• A systematic approach has been developed to quantitatively compare different methods for characterizing voltage dips.
• The proposed approach is illustrated by comparing seven different characterization methods and their ability to predict the minimum dc-bus voltage of a three-phase adjustable-speed drive.
• It is shown that the most-commonly used characterization method gives a much larger prediction error than methods that include unbalance and phase-angle jump.

This paper presents a systematic approach to compare different methods for characterizing voltage dips in a quantitative way. A prediction error is calculated between measured and synthetic dips (reproduced from single-event characteristics for the measured dips) with respect to the way they impact the performance of a generic device. The proposed approach is illustrated by comparing seven different characterization methods and their ability to predict the minimum dc-bus voltage of a three-phase adjustable-speed drive. A generic model of such a drive is used for this. Based in this comparison it is concluded that characterization method for dips in three-phase systems should include unbalance and phase-angle jump.