Fault identification and diagnosis of grid linked DFIG system based on a kernel PCA ESMO technique

Abstract

Fault Identification and Diagnosis (FID) is an imperative operation in newlinewind turbines to ensure the protection of the system, transient stability and newlinesystem quality. FID approaches are mostly used to avoid the shutdown of newlineelectrical energy production and protect the system from mal functions. In newlinethis world, the power production based upon Wind Energy Conversion newlineSystem (WECS) fed Doubly-Fed Induction Generators (DFIG) occupy a newlineleading position due to the robustness, this generators provides high wind newlineenergy, increase the reliability, effective monitoring and efficiency newlineimprovement. Although, DFIG based Wind Turbines (DFIG-WT) are more newlineprone to instability in parameters, such as voltage, current and power. newlineAdditionally, the fault rate is increasing by the grouping of power converters. newlineHence, it necessitates supplementary protection to guarantee nominal newlineoperation. Therefore, in this thesis, proposed an efficient hybrid FID strategy newlineto enhance the performance of DFIG based WECS. The major objective of newlinethe proposed method is enhance the fault ride through capability of the DFIG newlineand also improves the power quality of the system. The proposed technique is newlinethe hybrid wrapper of Kernel principal component analysis (Kernel PCA) and newlineenhanced Spider Monkey Optimization (ESMO) approach; hence it is called newlineKernel PCA-ESMO technique. The Spider Monkey Optimization (SMO) newlineapproach foraging behaviour is improved by the Galactic Swarm newlineOptimization (GSO) approach; hence it is named as enhanced SMO. The newlineproposed approach is operated in two stages like identification stage and newlinediagnosis stage. The Kernel PCA approach is used to identify the failure newlinestages of DFIG at grid connected system. newline newline