A multi-feature-based passive islanding detection scheme for synchronous-machine-based distributed generation

摘要

In this article, a new passive islanding detection scheme is presented for synchronous based distributed generation units (SBDGs) considering most vital islanding detection factors such as non-detection zone (NDZ), event detection time, high reliability in clustering events, cost of implementation, and independency of SBDG parameters. To do so, three-main concepts are considered: (a) the use of the electrical parameters from the SBDG unit which are independent from inertia, (b) the combination of various passive parameters, and (c) the utilization of the d-q framework of SBDG that can present its true behavior during the normal and abnormal events. Therefore, five parameters including q-component of damper winding current, d- as well as q-components of stator winding currents, reactive power and positive sequence of voltage are selected owing to the fast abruptly changes imposed on their pattern of variations during the disturbances. In this regard, to accurately detect the occurred events, three features based on the selected parameters are considered, which are (dI(dm)/dt), (dI(d)/dI(q)) and (dV(ps)/dQ). To demonstrate the effectiveness of the proposed islanding detection scheme (IDS), exhustive simulation studies are performed in MATLAB environment. The results illustrate that by means of the defined parameters, high reliable perfomance of the proposed approach is satisfied through recognizing all the islanding and grid-connected disturbance (GCD) conditions, including zero power imbalance, within minimum time compared to the existing methods available in literature. It is also revealed that the SBDG parameters do not have any adverse effect on the suggested IDS performance.