基于线路零序相频特性的谐振接地系统主动区段定位OA

The single-phase-to-ground faults account for approximately 80％of all faults in the distribution network,necessitating accurate and reliable fault location to ensure power system security and stability.Due to the influence of different grounding modes,traditional passive location methods based on transient or steady-state features often lack location sensitivity under high-resistance grounding or small fault initial angles,owing to weak characteristic signals.Existing active signal injection methods typically require additional signal-injection equipment and are limited by monotonous signal features and a lack of anti-interference capability.Considering the development of the double-high power system,a reliable identification of the section affected by a ground fault can be achieved by a characteristic controllable current signal generated by the power electronics employed to switch the damping resistor in the arc-suppression coil.This method requires no additional signal-injection equipment,maintains identification accuracy under interference,and eliminates the need for potential transformers at terminal units. When the single-phase-to-ground fault enters a steady state,the damping resistor is periodically switched for short durations by controlling the delay triggering angle δ of the anti-parallel thyristor applied to the damping resistor in the arc suppression coil,generating a transient current disturbance in the zero-sequence circuit.The characteristic signal is then accurately extracted from the zero-sequence current—which contains power frequency components and background noise—by employing a full-cycle subtraction combined with a multi-signal polarity correction mean algorithm.Furthermore,establishing the zero-sequence harmonic impedance at the beginning and end of the section,and constructing the phase information indicator θav based on the characteristic frequency band.Based on the indicator above,the double-end and single-end section location criteria are established to cope with the various operating conditions. Results from Matlab simulations based on the topology of an actual 10 kV distribution network line,as well as simulated experiments employing the method of reducing the entire network voltage,indicate that the disturbance signal is concentrated in the low-frequency band below 100 Hz,and the signal strength increases significantly with the increase of δ.Besides,the phase difference indicator θcr between the sending and receiving ends of the faulted section remains consistently and significantly positive,therefore the faulted section can be identified accurately in both the simulation tests and simulated experiments.Sensitivity analysis confirms that the θcr of the faulted section remains more than 10° larger than that of the healthy sections under the combined effects of high fault resistance,long cable lines,and measurement errors caused by asynchronism.Furthermore,comparative analysis verifies that the proposed method can accurately locate the fault under conditions of a fault inception angle of 5° and a fault resistance of 2 kΩ,while the existing methods based on transient reactive power or current increment exhibit abnormal location results. According to the analysis above,the following conclusions can be drawn:(1)The proposed section location method generates distinct and controllable signal characteristics,which facilitates the detection and extraction of fault signals under the interference of factors such as fault resistance.(2)The variation in the zero-sequence phase-frequency characteristics of the section can be obtained to determine the faulted section by utilizing the generated current signal,eliminating the need for voltage measurement data.(3)The proposed method identifies the faulted section based on changes in the zero-sequence phase-frequency characteristics,a criterion that is independent of the grounding fault type and is minimally affected by fault location and resistance.(4)The signal strength affects the accuracy and synchronization of the signals detected at each terminal,which in turn impacts the precise acquisition of the difference in zero-sequence phase-frequency characteristics between the sending and receiving ends of the section.The signal strength can be increased by controlling δ.