ABSTRACT
This paper presents a Fuzzy Inference System (FIS)-based method to detect a fault, identify the faulty section, and recognize the faulty pole in the Voltage Source Converter (VSC)-based MultiTerminal High Voltage Direct Current (MT-HVDC) system. The method uses only rectifier end measurements of voltage and current signals. To achieve complete protection of the MT-HVDC system, three separate frameworks of FIS modules have been developed. The FIS-1 detects the presence of fault in either AC or DC segments. The FIS-2 identifies the fault section and eventually, the FIS-3 recognizes the faulty pole. The proposed method provides a rapid fault detection and it does not require any communication media as it is based on the rectifier-end measurements only. The efficacy of the implemented FIS-based method is evaluated in an MT-HVDC system simulated in MATLAB environment.
INTRODUCTION
A key aspect answerable for getting tremendous care in the growth of traditional technology is growing demand for the electricity and hence growing the power transmission capacity of the transmission network. As a consequence of rising need for the electricity and the inclusion of renewable energy resources in power networks, multi-terminal HVDC (MT-HVDC) systems have become more appealing in recent years. In the literature survey, transmission network to transmit large amount of power over far distances has been addressed based on the Current Source Converter (CSC). However, as opposed to the MT-HVDC system with voltage source converter (VSC), the ability to expand the MT-HVDC system with CSC becomes comparatively more troublesome. Thereby, the enhancement to the MT-HVDC system of the established two-terminal transmission networks is not compelling for large-scale construction. Compared with CSC-based HVDC systems, VSC-based HVDC systems improve the flexibility of power transmission network. In addition, less operational cost and redundant switching states will be provided in the VSC-based HVDC system. Moreover, the VSC-based HVDC systems have overcome few of the drawbacks associated with Line Commutated Converter (LCC-based HVDC); for instance, the power electronics switches used in the VSC provide both turn-on and turn-off capability at desired instant as compared to the thyristors in the CSC which can provide only turn-on capability and turns off as per natural line commutation. Since VSC-based HVDC system offers a steady voltage, frequency, and phase angle for wind farms, it is a common technology for connecting integrated systems such as large-scale wind farms. Further, the VSC-based HVDC can easily start and incorporate the offshore wind farm into the grid, which is simpler than using LCC-based HVDC. In the VSC, changing the direction of the DC current achieves the power flow reversal, whereas the DC voltage polarity in the LCC has to be inverted at all interconnected stations. The use of VSC-based HVDC system enables the implementation of multi-terminal HVDC (MT-HVDC) grids. One of the main benefits is the suitable approach to integrating VSC over various DC terminals for the MT-HVDC system. The recent growth in research makes it easy to incorporate the numerous VSC-based MT-HVDC systems into distributed energy resources. Generally, compared with the CSC-based HVDC systems, the VSC-based MT-HVDC network is more suitable for various application mentioned. By prevailing literature survey, various fault detection, location and classification methods have been discussed. These are the decisive actions to obtain a reliable and correct operation. An artificial neural network (ANN)- based fault detection method is proposed by applying harmonics of voltage waveform in rectifier side in the HVDC system. A novel method for detection of a fault and fault distance in the HVDC system using ANN is presented. Here, the ANN is used as a pattern recognizer to correlate post-fault DC voltage signal to fault distance. A fault location method in the HVDC system based on ANN is also described. A fault detection method based on the wavelet in the MT-HVDC system for combination of underground and overhead transmission lines are introduced. Such methodology preferred for finding the location of fault raises the calculative burden due to numerous constrictions, such as selection of mother wavelets, noisy conditions, and need for various sensors. Three modules implemented based on fuzzy inference system (FIS) detect the faults, discriminate the fault section, and identifies the faulty pole for two-terminal HVDC systems. The method has not been tested for operation in a MT-HVDC system. The primary advantage of the MT-HVDC system relative to the HVDC transmission system is that there are many transmission lines between converter stations to shape mesh networks and maximize network stability, which increases the efficiency of the power supply, decreases the number of converter stations, and lowers maintenance and operational costs. This FIS-based method senses a fault and identifies a defective segment within a short time, which clarifies the supremacy of the FIS-based method over other methods. A fault detection based on FIS in MT-HVDC system is proposed, but more consideration and fault classification tasks are not reported. A fault identification scheme in HVDC line based on the convolutional neural network (CNN), fast Fourier transform (FFT), and gramian angular field (GAF) is proposed. The scheme suffers from high computation complexity using three tools. An Intrinsic Time Decomposition-based scheme is designed to protect the a HVDC system. Finding the threshold value is a challenging issue in the proposed scheme. A fault identification scheme by using wavelet transform modulus maximum is proposed in a HVDC line. The scheme needs high sampling frequency to detect a fault.

Block Diagram