| Abstract: |
| This paper presents a design of robust intelligent protection technique using adaptive neuro-fuzzy inference system
(ANFIS) approach to detect and classify the fault types during various faults occurrence in large-scale grid-connected
wind farm. Also, it is designed to determine the fault location and isolate the wind turbine generators located in the
faulted zone during fault occurrence and reconnect them after fault clearance. The studied wind farm has a total rating
capacity of 120 MW, where it consists of 60 doubly fed induction generator (DFIG) wind turbines each has a capacity of
2 MW. Moreover, the wind farm generators are positioned in 6 rows, where each row consists of 10 generators. The
impacts of fault type, fault location, fault duration, cascaded faults, permanent fault and external grid fault on the
behaviours of the generated active and reactive power are investigated. Also, the impacts of internal and external faults
in cases of different transition resistances are investigated. The simulation results indicate that, the proposed ANFIS
protection technique has the ability to detect, classify and determine the fault location, then isolate the faulted zones
during fault occurrence and reconnect them after fault clearance. Furthermore, the wind turbines generators which are
located in un-faulted zones can stay to deliver their generated active power to the grid during fault period. |
| Key words: ANFIS, Fault location, DFIG, Wind farm, Active power, Reactive power |
| DOI：10.1186/s41601-018-0090-4 |
|
| Fund: |
|
