A Fuzzy Modeling Technique to Assist Submersible Inspection Robot for Internal Inspection of Transformers

Abstract

Transformers are preeminent units of any power transmission network which are prone to failures due to multiple interrelated factors. For recognition of obscure failure modes leading to severe degradation of the transformer, internal inspection is a necessity. Submersible inspection robots are state of the art devices which are utilized for routine internal inspection of an oil filled transformer for identification, segmentation and assessment of faults. In this research paper, by using data from assorted diagnostic tests, a fuzzified computational algorithm is developed for determining the condition in which performing internal inspection is an absolute necessity. Particular attention is given to establish a decision logic for minimizing the work envelope of internal inspection robot by identifying likely faulty subsystems from diagnostic test results. This algorithm will further reduce the maintenance time and workload of the robot for performing internal inspection of the transformer. This paper will be helpful for researchers and service engineers involved with the maintenance of transformers.