Susmita Kar

Data-Mining Model Based Intelligent Differential Microgrid Protection Scheme

This paper presents a data-mining-based intelligent differential protection scheme for the microgrid. The proposed scheme preprocesses the faulted current and voltage signals using discrete Fourier transform and estimates the most affected sensitive features at both ends of the respective feeder. Furthermore, differential features are computed from the corresponding features at both ends of the feeder and are used to build the decision tree-based data-mining model for registering the final relaying decision. The proposed scheme is extensively validated for fault situations in the standard IEC microgrid model with wide variations in operating parameters for radial and mesh topology in grid-connected and islanded modes of operation. The extensive test results indicate that the proposed intelligent differential relaying scheme can be highly reliable in providing an effective protection measure for safe and secured microgrid operation.

A Fuzzy Rule Base Approach for Intelligent Protection of Microgrids

Abstract The proposed work develops a decision tree-induced fuzzy rule base intelligent protection scheme for fault detection and classification in a microgrid with multiple distributed generation interfaces. The proposed protection scheme retrieves one cycle post-fault current signal samples of each phase from fault inception at bus ends of the respective feeder to derive some differential features. The retrieved current samples are pre-processed using S-transform to obtain a time–frequency contour. The statistical features, such as energy, mean, standard deviation, and entropy, are computed from the time–frequency contour, which is further used to calculate the differential features. The differential features are used to build the fault classification tree. From the decision tree classification boundaries, the fuzzy membership functions are drawn, and further, the corresponding fuzzy rule base is generated for the final relaying decision. The proposed scheme is developed on a MATLAB/SIMULINK (The MathWorks, Natick, Massachusetts, USA) platform, including wide variations in faulted conditions, and the extensive test results indicate that the proposed intelligent relaying scheme can reliably provide protection measures for microgrids with different modes of operation.

Intelligent anti-islanding protection scheme for distributed generations

The paper presents a passive islanding detection scheme for both synchronous and inverter based distributed generations interfaced to microgrid using Decision Tree. The scheme involves extraction of different features at target DG location and builds the Decision Tree for intelligent anti-islanding protection. Extensive simulations are carried out including wide variations in system loading and power mismatch during islanding, and a comprehensive Decision Tree is built for effective islanding detection for both synchronous and inverter based distributed generations in microgrid. The performance of the proposed data-mining based anti-islanding scheme is compared with existing Rate of Change of Frequency and Rate of change of Voltage relay and, provides improved performance at lower end power mismatch.

Combined S-transform and data-mining based intelligent micro-grid protection scheme

The paper presents a combined S-transform and decision tree based intelligent scheme for fault detection and classification in the micro-grid. The proposed method preprocesses the faulted current signals using S-transform to extract differential statistical features at the ends of the respective feeder, which are used to build decision tree based data-mining model for final relaying decision. One cycle post fault current samples of each phase from fault inception at bus-ends of the respective feeder are used to derive differential features. The differential features are used to train the three decision trees to provide fault detection, fault detection in different operating mode and fault class associated in the fault process. The algorithm is tested on simulated fault data with wide variations in operating parameters of the power system and the extensive test results indicate that the proposed intelligent relaying scheme can reliably provide protection measure for micro-grid with different modes of operation.

High impedance fault detection in microgrid using maximal overlapping discrete wavelet transform and decision tree

The paper presents a high impedance fault (HIF) detection scheme for microgrid using maximal overlapping discrete wavelet transform (MODWT) and decision tree (DT). The small, nonlinear, random and widely varying fault current of HIF makes the over-current relay insensitive. Thus, time-frequency information is required to distinguish HIF from no-fault event. The proposed work start with retrieving the fault current which is pre-process through (decompose using) MODWT to get the details and approximation coefficient. Further, some statistical features are estimated using the details and approximation coefficient. The features extracted are used to train decision tree (DT) for accurate classification of HIF from no-fault. The proposed scheme is tested for HIFs detection in microgrid for different operating conditions. The test results show that the proposed scheme provides an effective protection measure against HIFs for safe and secured operation of microgrid.

Data-mining based comprehensive primary and backup protection scheme for micro-grid

The paper presents a data-mining based comprehensive protection scheme for micro-grid using decision tree. The proposed scheme pre-processes the current and voltage signals at both ends of the faulted feeder using discrete Fourier transform (DFT) and compute the differential features, which are used to build two decision trees (DTs) for the final relaying decision. The differential features between immediate buses are considered for primary protection and differential features between far end buses are used to build data-mining model (DT) for secondary protection. The proposed scheme is tested for fault and no fault situations in the studied micro-grid with wide variations in system operating parameters and fault parameters including grid-connected and islanded modes of operation. The extensive test results show that the proposed intelligent relaying scheme can provide reliable protection measure considering both primary as well as backup protection for safe and secured micro-grid operation.

Overcurrent relay coordination for micro-grid with different operating conditions

The dynamic behavior of micro-grid imposes serious challenges to the design of the protection system. The fault current magnitude and direction changes significantly with variations in the operating conditions which may leads to failure of protection system and non-selective operation. To alleviate the problems, this paper proposes an effective relay coordination using Genetic Algorithm (GA) considering different topology and operating modes of the micro-grid. The proposed coordination enables the overcurrent relays to update their current setting according to the operating conditions to achieve the minimum possible operating times while maintaining the coordination criteria between the primary and backup relays. Extensive test results with both synchronous and inverter based DGs and operating conditions indicate the ability of the proposed relay coordination in devising an effective and reliable protection measure for microgrid.

Multiple features based anti-islanding protection relay for distributed generations

A multiple features based passive islanding detection technique using S-transform and decision tree is proposed in this paper. It consider synchronous as well as inverter based distributed generations interfaced to micro-grid. The technique involves extraction of different statistical features of frequency contour obtain by performing S-transform with the current signal retrieved at target distributed generation location and builds the decision tree for islanding and non-islanding classification. Extensive simulations are carried out for wide variations in system loading and power mis-match during islanding, and a decision tree is built for effective islanding detection for both synchronous and inverter based distributed generations in microgrid. Performance of the proposed anti-islanding technique is compared with existing rate of change of frequency and rate of change of voltage relay. It provides improved performance at lower end power mis-match.