A fast protection of hybrid AC/DC distribution network based on feature extraction of DWT

This paper presents a hybrid distribution network protection scheme based on digital signal processing using discrete wavelet transform and feature extraction. The proposed protection scheme can detect all types of faults in both AC and DC sides taking into consideration the disturbances affecting the system. Wavelet transform is used to decompose the signals into different frequency components (approximate and detailed coefficients); then the feature extraction technique is used to extract the characteristics of a certain coefficient. The extracted features are then used to identify different types of faults in the system. Furthermore, a threshold-based decision-making method is employed for fault detection in both AC and DC sides. Simulation results demonstrate that the proposed protection scheme can accurately detect all types of faults in either in AC or DC systems with high accuracy and low computational burden while eliminating the need for a complex training module that requires a lot of data. Also, the proposed scheme has the advantage of depending on local measurement, so that communication infrastructure is not required.