Objective Power quality is a critical factor for the safe, stable, and efficient operation of urban rail transit (hereinafter referred to as URT) power supply systems. Addressing the data characteristics of existing power quality management systems, such as long collection cycles and large time intervals, a multi-indicator comprehensive evaluation system based on the EWM (entropy weight method) is constructed to achieve precise quantification and dynamic monitoring of power quality in URT distribution systems.

Method Taking the continuous field measurement data from the distribution transformer at an URT station in Hefei City as a case study, six key indicators—system frequency deviation, three-phase THDi (total harmonic distortion of current), three-phase THDu (total harmonic distortion of voltage), three-phase voltage unbalance, power factor deviation, and three-phase voltage fluctuation rate—are selected to build a multi-dimensional evaluation framework covering both steady-state and dynamic characteristics. The EWM is employed to objectively assign weights to each indicator, avoiding bias caused by subjective judgment. A non-linear scoring function is introduced to map raw data into standardized score values, and the power quality levels are categorized based on score intervals, thereby achieving full-process automated assessment from data collection to level determination.

Result & Conclusion  The EWM calculation results indicate that the three-phase voltage fluctuation rate (weight of 0.22) and the three-phase THDi (weight of 0.20) are the core indicators affecting the power quality of the station's distribution transformer. The comprehensive scoring results show that the overall power quality of the research object is at a good level (mean value of 87.2 points), with the "excellent" grade accounting for 27.2% and the "good" grade accounting for 72.0%. Furthermore, a regular pattern of degradation in scores is observed during peak electricity consumption periods.