Islanding Detection for Inverter-Based Distributed Generation Using Unsupervised Anomaly Detection

Abstract

Islanding detection with the rising grid supporting inverter-based distributed generation is becoming more critical protection due to its high droop gains and overall decreased system inertia leading to rapid changes in the electrical parameters. Traditional methods for islanding detection in this regard are susceptible to significant problems such as non-detection zone, false-positive detection, and inefficient mode of validation. Therefore, to attenuate these problems, this paper proposes a hybrid islanding detection technique based on unsupervised anomaly detection using autoencoders. This technique uses the rate of change of frequency as primary and phase angles of the voltage and current as secondary detection parameters, demonstrating improved performance, reliability, and robustness due to its shared advantage of both active frequency drift and autoencoder. Furthermore, a dialectic model of offline and online validation schemes is also proposed to ensure the reliability of detection. Extensive simulations and validations have been carried out on multiple networks to generate data sets used to train, test, and validate the technique and compute its statistical significance, thereby confirming its effectiveness. The optimal islanding detection time for the base cases was recorded as 20 milliseconds with an F1-score of 0.991, dependability index of 0.998, security index of 0.995, with total harmonic distortion of 4.56% and zero non-detection zones, which complies with IEC 61000-3-2 and IEEE standard 1547's requirement of detection within two seconds after islanding.

Islanding Detection Methods for Microgrids: A Comprehensive Review

“…The optimal gain K f is directly proportional to the quality factor of the load; however, when the quality factor is more than five, K f will be too large to create a false detection and can even cause system instability. The authors in [88] proposed a Q g −f droop curve method to keep K f to a safe value and monitor the change of frequency for islanding detection. During normal operation, the reactive power is controlled by the grid; however, during islanding operation, since the DG units operate at unity power factor and produce no reactive power, this creates a frequency difference between the actual and the rated system frequency.…”

“…This works by combining the rate of change of frequency and voltage methods to detect islanding [88]. The rate of change of voltage and the rate of change of frequency at the PCC are monitored as a first step to detect if one of them exceeds the predefined threshold to indicate islanding might have occurred.…”

1D recursive median filter based passive islanding detection strategy for grid‐connected distributed generations network

“…The existing approaches confronted the islanding detection problem in modern grid-tied networks in a comprehensive manner. However, the existing research has some limitations: (i) Noisy measurements are not considered in some of the existing islanding detection schemes [15,36]; (ii) some of the islanding detection strategies have a very high computational burden [22][23][24][25][26][27][28][29][30][31][32][33]; (iii) threshold setting in some schemes is difficult to implement [15,21]; and (iv) NDZ is large in few passive schemes, especially during low power miss-match moreover, some islanding detection indexes get failed [17,20].…”

“…The index was developed based on the rate of change of frequency (ROCOF) of the DG. The authors in [15] presented an unsupervised anomaly detection to detect islanding events in the distribution network. The ROCOF and the phase angle of the voltage/current were used as the first and second detection parameter respectively.…”

Microgrid Protection Challenges and Mitigation Approaches–A Comprehensive Review

“…If the fault occurs on the utility grid side which is outside of the PCC and then DER remains supplied to the fault along with the utility grid [96]. During external fault (islanding) conditions, islanding detection should occur and the microgrid should be isolated from the utility grid area through PCC because fault current flowing in the reverse direction from the microgrid DER towards the fault point which is located at the utility grid area to protect the microgrid equipment [97], [98].…”