Optimal estimation and sensor selection for autonomous landing of a helicopter on a ship deck

[摘要] ENGLISH ABSTRACT: This thesis presents a complete state estimation framework for landing an unmannedhelicopter on a ship deck. In order to design and simulate an optimal state estimator,realistic sensor models are required. Selected inertial, absolute and relative sensorsare modeled based on extensive data analysis. The short-listed relative sensors includemonocular vision, stereo vision and laser-based sensors.A state estimation framework is developed to fuse available helicopter estimates, shipestimates and relative measurements. The estimation structure is shown to be bothoptimal, as it minimises variance on the estimates, and flexible, as it allows for varyingdegrees of ship deck instrumentation. Deck instrumentation permitted rangesfrom a fully instrumented deck, equipped with an inertial measurement unit and differentialGPS, to a completely uninstrumented ship deck. Optimal estimates of allhelicopter, relative and ship states necessary for the autonomous landing on the shipdeck are provided by the estimator. Active gyro bias estimation is incorporated intothe helicopter's attitude estimator. In addition, the process and measurement noisecovariance matrices are derived from sensor noise analysis, rather than conventionaltuning methods.A full performance analysis of the estimator is then conducted. The optimal relativesensor combination is determined through Monte Carlo simulation. Results showthat the choice of sensors is primarily dependent on the desired hover height duringthe ship motion prediction stage. For a low hover height, monocular vision issufficient. For greater altitudes, a combination of monocular vision and a scanninglaser beam greatly improves relative and ship state estimation. A communicationlink between helicopter and ship is not required for landing, but is advised for addedaccuracy.The estimator is implemented on a microprocessor running real-time Linux. Thesuccessful performance of the system is demonstrated through hardware-in-the-loopand actual flight testing.

[发布机构] Stellenbosch University