engleski

Neural network selection of a maximum efficiency ship screw propeller

The choice of an optimum ship screw propeller is one of the persistent problems in naval architecture. Even though the selection can be performed using the propeller series diagrams or regression polynomials, the idea of the paper is to apply the advantages of the neural networks to this relatively simple problem as an introduction to more complex ship design problems. The network presented in the paper was created and trained to provide the characteristics of the maximum efficiency propeller. The neural network was created using Fast Artificial Neural Network Library. The structure is formed by a multilayer perceptron network with a single hidden layer. To train the network, data regarding the blade number, advance speed, delivered power, rate of revolution, diameter, pitch ratio, expanded area ratio as well as thrust and efficiency were set as inputs and outputs. The testing of the network proved its efficiency, which makes it not only a reliable tool for the preliminary screw propeller selection, but also a step towards the further development of neural networks applicable in various aspects of ship design process.

ship screw propeller; maximum efficiency; neural network

nije evidentirano