Robust flight reconstruction for helicopter simulation and training

Abstract

The concept of using flight-recorded data for training via real-time simulator playback of flight maneuvers is introduced, and the requirements and technical methodology for simulator-based flight reconstruction are discussed. A robust Kalman filter design, which includes parametric uncertainties modeled as multiplicativ e white noise processes, is utilized as a state estimator flight simulator driver. Actual robust Kalman filter implementation utilizes a UH-60 helicopter nonlinear simulation (piecewise linear aerodynamics) for state propagation to facilitate easy transition between flight reconstruction and manual simulation. This robust technique is compared to the traditional linearized extended Kalman filter approach. Resulting flight data reconstruction system simulation, using computer-generated data and flight test data, demonstrates the flight payback insensitivity to helicopter model parameter changes. The study results indicate superior performance of the robust Kalman filter approach for the flight reconstruction for flight simulation (FRFFS) problem.