Signal clustering of power disturbance by using chaos synchronization

This study develops and applies a chaos synchronization-based detection method for an engineering application of monitoring power quality disturbance. The new method can detect minor dynamic changes in signals. Likewise, prominent characteristics of system signal disturbance can be extracted by this technique. The method is then combined with the extension recognition algorithm to accurately apply to signal clustering of power disturbance. According to extensive computer simulation results and a comparison among three typical chaotic systems, it is confirmed that the proposed method is well applicable using various chaotic systems, mostly with very high accuracies. As compared with other traditional methods, the new method is shown to have higher accuracy, faster computing speed and better expandability. It is foreseen that if the method can be implemented by system-on-chip in the near future, it will find many real engineering applications such as hand-held power quality analyzers and auxiliary means for on-line real-time detection, among others.