Improved rotor track and balance performance using an expert system

A major source of vibration in the helicopter is derived from the main rotor dynamics and imbalance. Excessive vibration can cause premature wear to the aircraft and rotating components. Additionally, vibration causes pilot fatigue that in itself can be a safety issue. While great improvements have been made in the rotor track and balance (RTB) techniques, it still requires a skilled maintainer. Systems such as the Goodrich IVHMU can give good solutions. However, it requires a high degree of user expertise in selecting which RTB control adjustments should be active for the current helicopter measured vibration. Selection of RTB control adjustments or solution strategies is based on the type of imbalance/vibrations that are being generated by the main rotor. A system which recognizes the current vibration state can be trained to use the best controls and solution strategy. Such an expert system would improve the quality of the RTB solution and in general reduce the overall workload of the maintainer. We present a framework for a Bayesian classifier expert system. The classifier is based on a multiple hypothesis testing and demonstrates the ability to deliver an improved RTB adjustment over a generalized solution strategy.