Islanding Detection in Grid-Connected Distributed Photovoltaic Generation Using Artificial Neural Network

In response to the problem of increased load demand, efforts have been made to decentralize the power utility through the use of distributed generation (DG). Despite the advantages of DG integration, un-intentional islanding remains a big challenge and has to be addressed in the integration of DG to the power system. Islanding condition occurs when the DG continues to power a part of the grid system even after the connection to the rest of the system has been lost, either intentionally or un-intentionally. The unintentional islanding mode of operation is not desirable as it poses a threat to the line workers’ safety and power quality issues. There are many methods which may be used to detect the islanding situation. Passive methods such as under/over voltage and under/over frequency work well when there is an imbalance of power between the loads and the DG present in the power island. However, these methods has larger Non Detection Zone (NDZ) and fail to detect the islanding condition if there is a balance of power supplied and consumed in the island. Remote technique of islanding detection is reliable but is not economical in small network area. Active technique of islanding detection distorts the power quality of the system as it introduces external signal in the system. This paper uses the Wavelet Transform (WT) to extract the features of voltage signal at PCC (Point of Common Coupling) and these features have been used to train Artificial Neural Network (ANN). The ANN model trained by these WT features, which understands the pattern of input feature vector, have been used to classify the islanding and non-islanding events. In this proposed method, NDZ has been efficiently eliminated which is created due to difference between active and reactive power during islanding condition. No power quality problem exists in this method as there is no disturbance injection. Hence, this proposed method is better than conventional passive and active methods.

Download Full-text

A Novel Islanding Detection Technique for a Resilient Photovoltaic-Based Distributed Power Generation System Using a Tunable-Q Wavelet Transform and an Artificial Neural Network

Energies ◽

10.3390/en13164238 ◽

2020 ◽

Vol 13 (16) ◽

pp. 4238 ◽

Cited By ~ 3

Author(s):

S. Ananda Kumar ◽

M. S. P. Subathra ◽

Nallapaneni Manoj Kumar ◽

Maria Malvoni ◽

N. J. Sairamya ◽

...

Keyword(s):

Neural Network ◽

Power Generation ◽

Artificial Neural Network ◽

Wavelet Transform ◽

Detection Technique ◽

Second Step ◽

Islanding Detection ◽

Distributed Power ◽

Distributed Power Generation ◽

Artificial Neural

Finding an appropriate technique to detect an islanding issue is one of the major challenges associated with the design of a resilient grid-linked photovoltaic-based distributed power generation (PV-DPG) system. In general, the technique used for islanding detection must be able to sense the disruptions from the electric grid and quickly disconnect PV-DPG from the grid. The quick disconnection of PV-DPG mostly avoids power quality problems, damage to power assets, voltage stability issues, and frequency instability. In this paper, a new islanding detection technique that is based on tunable Q-factor wavelet transform (TQWT) and an artificial neural network (ANN) is proposed for PV-DPG. The proposed approach consists of two steps: in the first step, the vital detection parameters are computed by performing simulations considering all possible switching transients, islanding events, and faults from the grid side. Then, the decomposition of obtained signals is done using TQWT on different levels. Using the obtained coefficients, at each level, features such as range, minimum, mean, standard deviation, maximum, energy, and log energy entropy are computed. The optimal feature set was selected as the input for the second step. The classification of the non-islanding and islanding states for PV-DPG is made using the ANN classifier in the second step, which achieved an accuracy of 98%. The results representing the efficiency of the proposed approach in noisy and non-noisy environments are also explained. Overall, it is understood that the proposed islanding detection technique would provide suitable insights to detect an islanding issue.

Download Full-text