Curvelet transform based statistical pattern recognition system for condition monitoring of power distribution line insulators

Abstract

The power distribution system is considered as the backbone of power system as it is responsible to deliver power to all the consumers. Due to an enormous increase in the power consumption, the damage in insulators at the electric poles is triggering interruption of power, and hence there is substantial loss occurring for the power sector. The power distribution system is protected from heavy transients by the use of insulators. So, monitoring system must be employed which regularly detects the condition of the insulators. Regular monitoring of the overhead power lines along with insulators, sending the images to the processing unit and application of image processing concepts to classify the insulator health condition is the proposed method and hence the determining breakage condition of the insulators. K-means clustering is used for segmenting the acquired image. Then, the insulators are extracted from the acquired image input, and curvelet transform-based features are obtained. These features are given to support vector machine for the determination of health of the insulator. Monitoring of the health of insulators can thus be done consistently, and this method of automatic classification reduces the human efforts too. Hence the efficiency of transmission is improved and continuous supply of power can be delivered.