A real-time method for time–frequency detection of transient disturbances in voltage supply systems

• Detection transient disturbances in time–frequency domain.
• Capture of a transient in the time domain, detection in the frequency domain.
• Detection in real-time of impulsive, oscillatory, transient harmonics, short duration rms variations.
• The method is insensitive to harmonic distortion.
• The method is low sensitivity to noise.
• The method allows the reconstruction of the transient signal in the time domain.

This paper presents a new method for real-time detection, extraction and analysis of transient disturbances in low-voltage supply systems. The algorithm developed is based on the sliding window method and applies the discrete wavelet transform to a difference signal computed in the time domain, comparing each voltage sample with its corresponding sample in the previous period. In the case that the difference signal surpasses a defined threshold, the algorithm proposed computes the energy of the different frequency bands of the difference signal corresponding to one fundamental period, detecting a transient disturbance in the frequency domain within the next period of the input signal when a specific threshold is exceeded. Other voltage events, such as short duration root-mean-square variations can also be detected using the method proposed. The results obtained demonstrate the good performance of the method.