Performance Analysis of Wind Turbine Monitoring Mechanism Using Integrated Classification and Optimization Techniques

Abstract

The advanced improvements in the techniques utilized in the field of energy generation using the wind mills has led to the remarkable minimization in its capital investments and the cost incurred in its operation. This has even enhanced the prominence of the winds farms worldwide and has raised the market share of the energy produced using the wind mills. Thus leading to the increase in the necessity for capable monitoring mechanisms that is cost effective to report the conditions of the wind turbines regularly. So that it would be helpful in early diagnosis of any fault that has occurred in the wind turbines. To have an accurate monitoring and minimized maintenance cost the paper integrates the Support Vector machine based Cuckoo Search Algorithm. The incorporation of the SVM with the CSO is validated in MATLAB under the gain-factor and the fixed value types of faults that are liable to occur in the wind turbines and the results acquired are compared with the other existing methods such as the SVM-PSO and K-NN. The results observed shows that the SVM based CSO is more accurate in predicting the fault models than the other existing models.