Wavelet-artificial immune system algorithm applied to voltage disturbance diagnosis in electrical distribution systems

Abstract

This study presents a new approach to detecting and classifying voltage disturbances in electrical distribution systems based on wavelet transform and artificial immune algorithm. This proposal unifies the negative selection artificial immune algorithm with the discrete wavelet transform concept. Thus, the measurements obtained in a distribution substation by the supervisory control and data acquisition acquisition system are transformed into the wavelet domain. Afterward, a negative selection artificial immune system realises the diagnosis, identifying and classifying the abnormalities. The principal application of this tool is to aid the system operation during faults as well as to supervise the protection system. To evaluate the performance of the proposed method, two distribution systems were modelled in EMTP software: an 84-bus test system and a 134-bus real system. The results show a good performance, emphasising the precision of the diagnosis.