# High impedance faults detection in power distribution networks using rogowski coils, kalman filtering, least-squares and non-recursive DFT computation engines

## TL;DR

This paper introduces a new method for detecting high-resistance faults in power networks using a non-recursive DFT approach, which outperforms existing techniques.

## Contribution

The novel non-recursive DFT computational engine for high-impedance fault detection is introduced and validated.

## Key findings

- The non-recursive DFT method showed a fixed amplitude error and rotated output angle during transient errors.
- The proposed technique outperformed current reconstruction, Kalman filtering, and least-squares methods in fault detection.
- Simulations using MATLAB confirmed the effectiveness of the new method under various arc models and test conditions.

## Abstract

This paper presents a novel computational engine based on non-recursive discrete Fourier transform (DFT) technology to detect high-resistance faults (HRFs) in power distribution networks. The non-recursive DFT approach utilizes the interconnection between a sliding window for current signals and a foundation function window during transient error. This non-recursive DFT technology is characterized by a fixed error in amplitude calculation and a rotated output angle. The proposed technique is compared against several established methods for high-resistance fault detection in distribution systems, including current reconstruction (CR) using Rogowski coils, Kalman filtering, and least-squares computational engines. The performance of each technique is evaluated by assessing the estimated percentage error in the calculation of fundamental and harmonic amplitudes. To study the proposed technique, the aforementioned methods were carefully modeled and simulated using MATLAB software for the IEEE 33-bus test feeder, simulated arcing faults, and Rogowski coils under various test conditions. The comparison is conducted under the influence of different arc models in the distribution system to assess the performance of the proposed technology. The comparative results demonstrate the effectiveness of the proposed non-recursive DFT-based technique in detecting high-resistance faults in power distribution networks, outperforming the other established methods considered in this study.