In nature, some animals exhibit impressive navigation capability using ambient magnetic field. Particularly, certain kinds of sea turtles can associate geomagnetism to spatial representation for positioning in transoceanic migration across the seemingly clueless sea. In robotics, the previous works on magnetic navigation can position a robot using ambient magnetic field, but they focused on an exploitation of extensively-explored magnetic map, i.e. the magnetic field of every inch of the region of interest is recorded for mapping. In this work, we propose an algorithm that is based on our analysis on how sea turtles navigate at sea under magnetic disruption as investigated and reported by biologists [1], [2]. We propose a direct likelihood method that generates pseudo training data to improve the estimation accuracy of the Gaussian mixture models. The experimental evaluation demonstrates that our localization algorithm exhibits stable and accurate positioning results. This work contrasts with the previous works which focused on magnetic localization using extensive data collection. On the contrary, we address whether magnetic localization is still feasible under scarce data samples and how to overcome this challenge.

The nine degrees-of-freedom (DOF) inertial measurement units (IMU) are generally composed of three kinds of sensor: accelerometer, gyroscope and magnetometer. The calibration of these sensor suites not only requires turn-table or purpose-built fixture, but also entails a complex and laborious procedure in data sampling. In this paper, we propose a method to calibrate a 9-DOF IMU by using a set of casually sampled raw sensor measurement. Our sampling procedure allows the sensor suite to move by hand and only requires about six minutes of fast and slow arbitrary rotations with intermittent pauses. It requires neither the specially-designed fixture and equipment, nor the strict sequences of sampling steps. At the core of our method are the techniques of data filtering and a hierarchical scheme for calibration. All the raw sensor measurements are preprocessed by a series of band-pass filters before use. And our calibration scheme makes use of the gravity and the ambient magnetic field as references, and hierarchically calibrates the sensor model parameters towards the minimization of the mis-alignment, scaling and bias errors. Moreover, the calibration steps are formulated as a series of function optimization problems and are solved by an evolutionary algorithm. Finally, the performance of our method is experimentally evaluated. The results show that our method can effectively calibrate the sensor model parameters from one set of raw sensor measurement, and yield consistent calibration results.

The Kalman filter is widely used but the tedious and time-consuming tunings of the filter parameters must be unavoidably carried out before use, and frustratingly, after every reconfiguration on the sensors. In this paper, we formulated the measurement residual in a performance index, and utilised an evolutionary method to automatically and efficiently calibrate the parameters of the sigma-point Kalman filter. Without analytically resolving the nonlinear and multivariate process and measurement models in the filter, the proposed method implicitly solves for the filter parameters in a gradient-free manner through a series of strategies including the selection, crossover and shuffling mutation. Furthermore, to demonstrate the superior performance of the method, we applied this method to a highly nonlinear and coupled state estimation problem on an unmanned helicopter which experiences a GNSS outage. The empirical results showed that the proposed method not only automated and significantly accelerated the exhausting tweaking of the filter parameters, but also yielded a high quality tuning result that strikingly outperformed an earlier, painstakingly handcrafted calibration.

For years, the puzzling cross-coupled responses between the control axes on the hingeless helicopters have long been haunting researchers. Different from previous works that underestimated the gyroscopic moments as a minor off-axis cross-coupling and misinterpreted the precession phenomena on the rotors of the hingeless helicopters as a phase-lag which is physically meaningless, this paper proposes a new method to relate both on-axis and off-axis responses by the influential gyroscopic moments through the actuation mechanism of the hingeless helicopters. Therefore, by this new method the debatable cross-coupling due to actuation dynamics can be directly and analytically quantified. This method is based on the fact that when the angular momentum of the spinning rotor is disturbed by the incremental lift along the main blades due to the varying cyclic pitch angle controlled by the servo mechanisms, the off-axis moments are induced to counteract the changes of the angular momentum according to the principle of gyroscope, and hence these gyroscopic moments directly exhibit the on-axis responses. This new method yields a parametric framework to examine the previously unexplained cross-coupled responses on the hingeless helicopters, and it shows that in hingeless helicopters except the aerodynamics the intricate nonlinearities are also attributed to their unintuitive actuation mechanisms. Finally, simulations as well as experiments have been carried out to validate the proposed modeling method.

Most unmanned aerial robots use a Global Navigation Satellite System (GNSS), such as GPS, GLONASS, and Galileo, for their navigation. However, from time to time the GNSS fails to function due to geographical restrictions and deliberated jamming. This paper proposes an Unscented Kalman Filter-based GPS/IMU integration method in order to accurately estimate the position and velocity of an unmanned miniature helicopter even when the GNSS malfunctions completely. Different from previous GPS/IMU integration methods that cannot propagate noisy inertial measurements to the position and velocity estimations on the rapid vibratory Vertical Take-Off and Landing (VTOL) platforms during the GNSS outage, this method novelly prioritises the propagations of the states in the Unscented Kalman Filter (UKF) algorithm and leverages the time-varying GNSS dilution of precision in line with the adjustments of the measurement noise covariances. Moreover, this method models the stochastic process in the inertial sensors by the acceleration white noise bias in addition to the commonly used random walking process. Without considering the specific actuation models that vary from vehicle to vehicle, this method can particularly be applied to the quivering unmanned helicopters which equipped with two-stroke engines. It yields a rapid and precise compensation for the sensor errors in order to effectively facilitate the propagations of inertial measurements to the position and velocity estimations. Finally, the superior performance of the proposed method in terms of accuracy and endurance is empirically demonstrated using our fully instrumented JR Voyager GSR helicopter.

Different from previous works that require prior trim conditions on the helicopter, this paper proposes a hierarchical PD controller that is robust in controlling untrimmed and therefore critically unstable helicopters. This controller can yield asymptotic stability of the helicopter in horizontal motion control, which can be proven by the linear stability analysis. And this controller can flawlessly engage with traditional dual loop autopilot by using auto-varying references in an inner stabilizing loop. Moreover, to facilitate the controller design, this paper derives the dynamics of hingeless helicopters with an emphasis on gyroscopic effect. Finally, the stability and superior performance of the proposed controller are empirically demonstrated on an instrumented JR Voyager GSR helicopter.