Radial Basis Neural Network Based Islanding Detection in Distributed Generation

In this paper, a passive neuro-wavelet based islanding detection technique for grid-connected inverter-based distributed generation was developed. The weight parameters of the neural network were optimized by intelligent water drop (IWD) to improve the capability of the proposed technique in the proposed problem. The proposed method utilizes and combines wavelet analysis and artificial neural network (ANN) to detect islanding. Connecting distributed generator to the distribution network has many benefits such as increasing the capacity of the grid and enhancing the power quality. However, it gives rise to many problems. This is mainly due to the fact that distribution networks are designed without any generation units at that level. Hence, integrating distributed generators into the existing distribution network is not problem-free. Unintentional islanding is one of the encountered problems. Discrete wavelet transform (DWT) is capable of decomposing the signals into different frequency bands. It can be utilized in extracting discriminative features from the acquired voltage signals. In passive schemes with a large non-detection zone (NDZ), concern has been raised on active method due to its degrading power quality effect. The main emphasis of the proposed scheme is to reduce the NDZ to as close as possible and to keep the output power quality unchanged. The simulation results from Matlab/Simulink shows that the proposed method has a small non-detection zone, and is capable of detecting islanding accurately within the minimum standard time.

Islanding detection for inverter-based DG coupled with using an adaptive neuro-fuzzy inference system

Background: The growth of renewable energy sources is increasing in the world to meet the energy consumption demand. The major problem after the integration of renewable sources is islanding. The islanding is not safe for equipment and customers. As per Distributed Generation (DG) interconnection standards, it should be detected within 2 seconds. Objective: This paper presents the review of various islanding detection methods for increasing the stability of islanded DG. This will help future researchers for selecting the best islanding detection method with zero NDZ. Methods & Results: The islanding detection methods are classified as local and remote techniques. The local techniques are again classified as active, passive and hybrid methods. Each method is presented with their islanding detection time, power quality issues, Non Detection Zone (NDZ), advantages and disadvantages. Conclusion: The fuzzy based artificial intelligence with Particle Swarm Optimization (PSO) passive methods have been reduced the NDZ to zero and increase the stability of DG without degrading the power quality as active and passive methods.

In the area of renewable energy technologies, islanding of distributed generation system needs to be prevented due to safety reasons and to ensure quality of power supplied to the customers. Several islanding detection methods based on passive and active schemes have been proposed in the literature. Passive methods have a large non detection zone (NDZ), While active schemes degrade power quality. This paper proposed a new passive islanding detection method which combines the Wavelet-transform and neural network techniques. This method can reduce the NDZ to zero without any perturbation that deteriorates the power quality. The simulation results show that the proposed islanding detection method is effective and robust in all kind of conditions.

An active islanding detection technique for an inverter-based distributed generator (DG) is presented. Two control strategies are studied: one in which the DG only supplies active power and one in which the DG supplies both active and reactive power to improve the power factor. The nondetection zone (NDZ) of the over/under voltage and frequency protection (OVP/UVP and OFP/UFP) method for each control strategy is determined. For the presented interface control, it is observed that the NDZ of the OVP/UVP and OFP/UFP is larger when the DG interface is designed to improve the power factor. In order to decrease the NDZ, an islanding detection method is designed that is based on creating a mismatch between the load and DG reactive power, thus forcing the frequency to deviate outside the OFP/UFP limits. It is shown that the proposed islanding detection technique is robust and has a negligible NDZ.

Distributed generators (DGs) require detection of islanding condition due to personnel safety hazards and equipment damages. Anti-islanding methods are divided into passive and active methods. The passive methods are based on measurement of the natural effects of islanding. The active methods use the effects by intentional transients or harmonics. When the power generated by the DG matches power consumption by load, passive methods fail due to small natural effect of islanding. Therefore, the passive methods have a non-detection zone (NDZ). The active methods can reduce the NDZ size. But these methods reduce the power quality of the grid. In this paper, a novel anti-islanding method (AIM) is proposed. The single phase DG using the proposed AIM injects output current included a little harmonic current into the grid and monitors the harmonic component of voltage at the point of common coupling (PCC) by using Goertzel algorithm. The Goertzel algorithm is a kind of discrete Fourier transform (DFT). It extracts magnitude and phase of the desired frequency from the input signal, with minimum computation. The proposed islanding detection algorithm does not only have NDZ but also bad effect on the power quality of the grid due to injecting harmonic components qualified the interconnection standard. The proposed islanding detection method (IDM) is verified by the PSIM simulations and the experimental results.

The increased requirement for clean, cost effective and uninterrupted electricity has led to the incorporation of microgrid including distributed generation (DG) units. One of the major issue with DG interconnection is islanding detection. This paper presents the comparative assessment report of islanding detection techniques on the basis of their non detection zone (NDZ) and detection time. A comprehensive table comparing the different available islanding detection methods is also presented.

This paper provides an analytical survey of the islanding detection techniques for the distributed generation systems. Islanding phenomena on takes place when the power supply from the main utility is intermittent due to numerous reasons, but the distributed generation keeps supplying power into the distribution networks. Islanding can be dangerous to workers and electrical equipment, even the power grid. Therefore, islanding detection is the priority among priorities. Islanding detection is a precondition of two working modes switching for distributed generation. In this paper, a detailed description followed by the classification of the islanding detection techniques has been made based on features, such as detection time, size of non-detection zone, power quality disturbances, system cost and operation under multiple distributed generation units. The merits and demerits of islanding detection methods are analyzed and summarized in this paper. In this manner, the greatest detection outcome can be achieved by exploiting the merits of each method. This paper is aimed to serve as a convenient reference and guidance value for deciding the islanding detection method for future islanding users in distributed generation system.

Many techniques used and still in usage for solving the problem of islanding detection are intrinsically passive, active, or hybrid of both. Each one of them has its own benefits and drawbacks. In this paper, we propose a method, which takes the advantage of a machine learning (ML)-based algorithm, namely, support vector machine (SVM), in order to produce the results more efficiently. The results of the simulations based on the model and experimentally measured parameters of a real-life practical photovoltaic (PV) plant give much better output than the traditional reported methods. During the tests and simulations, an additional problem, namely, grid fault, emerged, posing new challenges for the proposed method. Occurrences of islanding and grid fault are grouped together with the same kernel dimension and no custom hyperplane bordering. Discrimination between islanding and grid fault events is an essential dilemma, which is handled by the proposed SVM-based algorithm to achieve more precision in islanding detection and simultaneously detect the grid faults authentically. Nondetection zones (NDZs) and detection time (DT) are tested using two dimensions, namely, the generated active energy from PV plant (0%–110% of $P_{n}$ ) and distribution network voltage levels (±10% of $U_{n}$ ). Simulations based on the model and parameters of a real-life practical PV power plant are performed in MATLAB/Simulink environment, and several tests are executed for several scenarios. Finally, comparisons with previously reported techniques prove the effectiveness, authenticity, selectivity, accuracy, and precision of the proposed islanding and grid fault detection strategy with allowable impact on power quality according to UL1741 and its superiority over other methods.

The capability of islanding detection has become a requirement for grid connection of inverters utilized in distributed generation (DG) systems. The parallel RLC resonant circuit comprised, as a local load, in the standard test of islanding detection schemes offers lower impedance for higher frequency harmonics. With the presence of such a parallel RLC circuit, the full power balance between the distributed generation and local load will never be satisfied when the voltage at the point of common coupling is slightly distorted. In this paper, the performance of grid-connected inverter during islanding detection routine test is analyzed with sinusoidal and marginally distorted grid voltage. The results presented in this paper show that grid background harmonics may have significant impacts on inverter control and islanding detection algorithms.

An Islanding Fault Detection Method with CFDF-SVM Based RPV Approach under Pseudo Islanding Phenomenon

Rapid grid integration of inverter-interfaced distributed generators (DGs) and low-voltage ride-through capability of the smart inverter, as per IEEE 1547 standard, has made unintentional islanding detection (UID) cumbersome. Also, the feeder reconfiguration capability in meshed microgrids with multiple DGs has made the existing UID techniques ineffective. The methods have limitations in UID for small power mismatch conditions characterizing nondetection zone (NDZ). In hybrid methods for UID, a disturbance injection is continuous or for an extended period, which degrades the power quality (PQ) and creates stability issues. Hence, a decentralized closed-loop hybrid islanding detection method (HIDM) based on the indices extracted from the subcycle phaselet (passive) technique with limited-time (only for two cycles) <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex-math notation="LaTeX">$q$</tex-math></inline-formula> -axis dc offset voltage injection (active) is proposed for UID of DGs. The proposed HIDM mitigates the NDZ due to the closed-loop scheme. Subcycle estimations from the phaselet technique result in quick UID and fault detection and classification process. Furthermore, the PQ issue is reduced due to the limited-time voltage injection (based only on the command from phaselet technique) in the proposed HIDM. Simulation results from the real-time digital simulator on a meshed microgrid and performance comparison with other existing methods prove the efficacy of the proposed HIDM.

One of the facing problems in distributed generation is being unintentionally islanded which the detection of this phenomenon is of great importance. Common methods are active and passive islanding detection. These two methods are based on changing in parameters such as frequency, voltage and current harmonics. If, after the islanding occur, this changes are revealing and visible, islanding phenomenon is demonstrated which is detected by passive methods. If the generated power of inverter is equal to local loads consumption and assuming of the generated reactive power of grid connected inverter equal to zero, it should be detected by active methods but these two methods have some challenges such as reduction in power quality or wide Non Detection Zone (NDZ). In this paper, in proposed method, change of a new parameter will be studied for islanding detection diagnosis. With the difference that the studied system is a three phase system and unbalanced load is studied which is more consistent to reality. Because of small changes in parameters, we have to find the right approach toward active methods. Generally there are two views about the islanding performance. At first view, this phenomenon is seems negatively but in fact this is an opportunity. The first view suggests that all generators in the island should bring out of grid within a maximum of two seconds. In this view the concept of anti-islanding protection is proposed to tackle the task of creating such islands. The second view expresses the concept of micro grid. In this case as far as possible, power required by Inner Island loads were provided by distributed generators. The studied system was similar to standard IEEE-1547 and UL-1741. With the difference that we used three phase and local loads are assumed unbalanced. Conventional active detection methods are not able to identify the island with the aim to help keep it stable. They usually try to bring out the electrical parameters such as voltage and frequency from nominal limits. So in micro grids the island should be diagnosed using passive or remote techniques. Note that detection time of 2 seconds which is referred to on standards is not good and the detection time must be lower than this value. However, the proposed method that is one of the active detection methods is capable for use in micro grids. Simulation results using PSIM software on standard network shows the effectiveness of the proposed method. Keywords: Islanding, Unbalanced Loads, dq-frame, Phase Locked Loop (PLL), Non Detection Zone (NDZ), Micro-grid.

ABSTRACTImproved efficiency, enhanced integration with distributed generations (DGs), and the proliferation of DC loads in power systems have driven a surge in the adoption of DC microgrids (DCMGs). However, unintentional islanding poses a significant threat to these microgrids, necessitating robust islanding detection techniques to safeguard personnel, equipment and system stability. This paper proposes two novel passive islanding detection methods for DCMGs, explicitly addressing the non‐detection zone (NDZ). The first proposed method involves collecting voltage samples at the DG terminals and computing their empirical mode decomposition (EMD). Subsequently, intrinsic mode functions (IMFs) are extracted, and the energy in each window is calculated to establish a diagnosis. To effectively distinguish islanding from other disturbances and faults, various scenarios are investigated. The second method is predicated on ensemble empirical mode decomposition (EEMD). Following the collection of voltage samples from DG terminals, white noise is introduced to these samples. The method subsequently extracts IMFs through averaging and quantifies the energy content within each window. The proposed methods are validated through simulations in Simulink/MATLAB. The obtained results demonstrate the methods’ effectiveness in achieving fast and accurate islanding detection.

A large number of distributed power sources are injected to the distribution network, which can cause island operation in planned or unplanned situations. The islanded operation caused by the fault will cause the inverter to go off the grid in serious fault conditions such as short circuit between phases and disconnection. After the island is formed, if the common connection point before the island exchanges more power, a sudden change in the voltage within the island occurs. The frequency and phase angle changes of the voltage in the island are closely related to the characteristics of the inverter phase-locked loop, and the frequency mutation generally does not cause sudden changes in frequency due to a certain inertia of the phase-locked loop. Due to sudden changes in load power angle, input fault ride-through, etc., sudden changes in the phase angle of voltage in islands may occur. In order to solve the possible problems of power quality deterioration in islands, passive, active, and remote islanding detection schemes can be adopted, and the three schemes have their own advantages and disadvantages. In actual operation, it is necessary to consider the cooperation relationship between island detection, protection, safe automatic control and fault ride through.