Condition monitoring of 11 kV distribution system insulators incorporating complex imagery using combined DOST-SVM approach

The increasing demand for the uninterrupted supply of power is equally increasing the necessity to monitor the health of the distribution network. The failure of the network implicates loss of supply and hence incurs huge loss to power distribution utilities. This paper envisages an algorithm to monitor the condition of overhead distribution line insulators with complex background, keeping in view pragmatic considerations. Complex background necessitates an arduous effort to monitor the condition of insulators and requires a powerful technique to monitor and identify the condition of the insulator. Support vector machines (SVM) and Adaptive neuro-fuzzy inference system (ANFIS) are used to estimate the condition of the insulator with the aid of features extracted from Discrete Orthogonal STransform (DOST). The algorithm is developed in MATLAB environment using the image processing toolbox. Superiority of the SVM over ANFIS has been shown from the results for complex background to validate the efficacy of proposed technique.