Intelligent attitude and flapping angles estimation of flybarless helicopters under near-hover conditions

Inaccuracy of measurements, associated with most of the commercial-off-the-shelf (COTS) Inertial Measurement Unit (IMU), impede achieving an accurate attitude estimation during autonomous near-hover flight. Moreover, the unmeasured states of the Tip Path Plane (TPP) flapping angles of the main rotor make the estimation and control of unmanned helicopters more challenging. In this paper, an intelligent adaptive fuzzy data fusion algorithm is designed to obtain more accurate estimates of the attitude and flapping angles for a flybarless miniature helicopter. In this algorithm, the filter's measurement noise matrix is continuously adapted using the Innovative-based Adaptive Estimation (IAE) technique. This technique is based on evaluating the discrepancy between the actual and theoretical covariance of the filter's innovation sequence. A well-tuned multi-input, single output Fuzzy Inference System (FIS) takes the value of this evaluated discrepancy and its rate of change as inputs and provides the required adjustment value as an output based on a set of predefined fuzzy rules. Compared to the conventional Kalman Filter (KF) state estimation results, the proposed intelligent estimation results have demonstrated an obvious enhancement in estimating the attitude and the flapping angles. The estimated flapping angles have been also used to estimate the moments and forces of the helicopter rotors under near-hover assumptions. An actual near-hover flight was conducted to validate the performance of the proposed intelligent estimation method.