Fault detection, identification and localization in medium-voltage networks using fuzzy-logic

Secondary substations are usually equipped with short-circuit detectors only indicating that an over-current has occurred for a given line. In case of a fault, further information about type and location is required to recover an assured grid operation as fast as possible. Conventional algorithms for fault detection used in high voltage grids are not well suited because of the high uncertainties regarding current and voltage measurement in secondary substations. A fuzzy-based approach is well suited in this case. This paper presents a fuzzy logic for detecting, identifying and locating faults occurring within typical medium voltage grids. The concept of this method is to define multiple fuzzy rules for each type of fault. The rules consist of two credibility classes. This approach makes it easier to adapt expert knowledge, e.g. the impact of the line topology and neutral point treatment can be taken into account by appropriate rules. For proving the applicability of the proposed fuzzy logic, simulations in MATLAB/Simulink have been done. Further benefits and the results of the simulations are shown.