Abstract
Electrical faults can change the power quality parameters of power systems. A numerical protection technique for fault detection and imbalance assessment based the Durbin-Watson (DW) factors for three phase currents is proposed in this paper. The approach integrates two protection functions based on the Durbin-Watson and Pearson similarity algorithms into one protection scheme. This strategy can figure out online faults located on the three-phase power transformer windings, such as turn-to-turn, winding-to-neutral, and winding-to-winding. Moreover, it can distinguish between balanced and imbalanced currents. To assess the validity of the protection scheme, it is practically examined on a three-phase power transformer with tapped windings connected to a three-phase load. Comprehensive tests are conducted to investigate the efficacy and efficiency of the suggested scheme. The analog-to-digital converter is integrated with LABVIEW software to process and analyze the two algorithms of the suggested scheme. The results of the experiments reveal that the security, dependability and precision ratios of the developed protection are almost 99%. Additionally, the protection system can immediately identify electrical faults, triggering a tripping signal to both the annunciator panel and the circuit breaker trip coil of the equipment, but it remains inactive under normal operating conditions and acceptable current unbalance. In the fault events, the numerical approach can respond quickly using a limited data set within a single cycle of the foundation frequency, and operate effectively using a pair of algorithms based on DW and Pearson similarity. It is also robust against the condition of sound transformer windings. Besides, it can determine and estimate the severity rate of perturbation and unbalance in power transformer currents, and it has a protection redundancy. Furthermore, the scheme is extremely sensitive to light fault currents, and has a unique set of tripping curves.