FLIGHT PATH RECONSTRUCTION FOR AN UNMANNED AERIAL VEHICLE USING LOW-COST SENSORS

摘要

This paper shows how low-cost, off-the-shelfsensors can be used to reconstruct the position,velocity, and attitude of an aircraft. An extendedRauch-Tung-Striebel Smoother is used offline toextract state information from recorded raw datasignals. Some of the inherently large errors oflow-cost sensors are determined by means of aMaximum-Likelihood scheme.The applicability of this approach is shownusing simulated as well as real data from a lowcost flying test-bed, equipped with sensors obtainedfrom the RC model market. In both cases,results are very promising. The proposed algorithmthus constitutes a possibility of obtaininghigh accuracy estimates of the true state trajectoriesfor analysis and modeling.