Search Results

This paper shows the methods, techniques and codes developed as Artificial Intelligence tools for turboprop maintenance. Diagnostics is the main area of maintenance considered while Expert Systems and Back Propagation Neural Networks are the Artificial Intelligence subjects discussed. The codes allow consideration of the turboprop with different configurations. The Knowledge Base of Expert Systems and the training patterns of Neural Networks are obtained from the matrices of influence obtained through massive use of engine simulation programs. The paper shows examples of both Expert System and Neural Network utilization.

The methods and the techniques set-up for evaluating the control laws for the digital Main Fuel Control Unit (MFCU) of a small thrust turbojet engine are shown. A suitable parameter defining the surge margin is considered. The paper deals with the results of a study carried out in order to evaluate the influence of engine operating parameters on the surge margin. For this aim the steady state engine simulation programs, based on the state vector technique, have proved to be very effective. Finally the codes and the obtained results have been used for calculating some suitable control laws for a digital MFCU.