Evolutionary design of a multi-ANN control structure for a submersible catamaran

The development of an automatic design environment for obtaining artificial neural network based control system structures for complex underwater vehicles is presented. The design procedure is carried out by means of evolutionary techniques from a set of specifications using as a fitness evaluator a new type of hydrodynamic simulator based on added mass constants. The model's equations of motion employed were obtained by combining theoretical formulae with experimental data. The resulting design environment, called ADEUS, was used to obtain the neural network based controllers for a submarine catamaran. Results of the simulations obtained using this technique, of the application of the automatic design procedure and of the operation of the controllers thus obtained are presented.