Ship Transformer Gases in Oil Protection Using Neural Network Simulation

Power transformer is one of the major apparatus in the ship power system. Power transformer breakdown or damage interrupts ship power distribution system. Hence, to manage the life of transformers and to reduce failures some measures are being adopted. According to the moist environment of the sea dissolved Gas Analysis (DGA) is a reliable and commonly practiced technique for the detection of incipient fault condition within power transformer. This paper presents the application of Artificial Neural Network (ANN) for detecting the incipient faults in power transformer by using dissolved gas analysis technique. Using historical transformer failure data, ANN model was developed to classify seven types of transformer condition based on the percentage of three hydrocarbon gases. In this paper, to simulate the neural network performance, we used two methods, genetic and RPROP algorithms.

Keywords:

Quayside Cranes , Container Loading / Discharging , Productivity , Cycle-Time Analysis