Band pass filter and AFVmean‐based unintentional islanding detection

Abstract

Several international standards have suggested to detect the unintentional island formation to safeguard the working personnel and protect the electrical equipment in the system. An active islanding detection technique, which employs the q-axis current disturbance injection, has been utilised and an analysing technique is proposed in this work based on a combination of band pass filter and mean of absolute frequency variation (AFVmean) to detect and differentiate the island formation from other transient events. The proposed analysing technique (PAT) detects island formation within 180 ms, which is less than the prescribed time of various international standards, applicable to dynamic loading conditions without the need of adaptive threshold and communication, and offers zero non-detection zone (NDZ). The disturbance injection makes significant variation in the frequency of voltage signal on post-islanding condition even for zero-power mismatch condition, therefore, the NDZ is zero. The PAT is corroborated for various islanding and non-islanding events and the simulations are performed in MATLAB/Simulink tool.