Research on Electromagnetic Interference Signal Characteristics and Suppression Methods for High Frequency Partial Discharge Monitoring in Renewable Energy Stations

As China's energy transformation advances, the proportion of renewable energy power generation in the power system continues to increase. 35 kV power cables are widely used as collector lines in renewable energy stations. High frequency (HF) partial discharge (PD) online monitoring is an important means to discover potential insulation defects in power cables and evaluate their operational status. Compared with traditional power stations, renewable energy stations use a large number of power electronic equipment, which can generate severe HF and high intensity pulse interference. These interferences are prone to submerging weak PD signals, posing new and severe challenges to high-frequency PD detection. Based on the engineering practice of Ulanqab Renewable Energy Station, this paper conducts actual measurement research on high-frequency electromagnetic interference signals in the station. According to the topological wiring of the station, the source and characteristics of the interference signals are studied through multi-point synchronous comparison testing. According to the time-frequency characteristics of the signal, a pulse extraction method based on OTSU and local energy threshold is proposed. Then, according to the extracted large number of pulses, a repetitive pulse interference separation method based on spectral clustering was proposed, which successfully separated several similar pulses from field data. On this basis, combined with stencil test, each pulse interference signal is identified and separated one by one, thus achieving the strategy of eliminating high-frequency and strong repetitive pulse interference signals. Verified with field data, the interference suppression rate reached 93.67%, proving the feasibility of this strategy.