Research On Relative Positioning Method Of Multiple Unmanned Flying Vehicles

Network and formation are main methods to realize collaborative operation of unmanned flying vehicles, which depends on measurement of relative motion state and persistence of relative configuration, so high-precision relative location is basement of collaborative operation. Based on this, this paper deeply investigates into relative positioning methods under two different types of movement condition.First, in relative maneuver situation, method of network adding (acquisition of relative position) and network maintain (persistence of relative configuration) of multi-UVs are proposed. Research emphasis is stability. By using combination of simple motion models, stability of network adding is enhanced by many steps such as main network nodes selection, manuver detection, simple multi-model and results fusion. In network maintaining, acceleration filter is separated from coordinate-velocity filter. The two-level filter can track maneuver in time and H∞filter makes the algorithm more robust. Simulation results show that algorithm of network adding and maintain is stable and effective, and no divergence occurs in maneuver moment.Then, in non-relative maneuver situation, based on measurement of distance, high-precision relative positioning method of formation satellites which requires high accuracy in relative configuration is proposed. Research emphasis is to obtain accurate relative position. We first choose GPS location result, single difference of GPS pseudorange and relative distance between satellites as the foundation measurement of initial location algorithm. Subsequently, in the support of high accuracy measurement of relative distance between satellites, location refinement algorithm significantly enhance the accuracy of relative allocation and the stability of filtering by strong tracking filter (STF) technology which effectively limits the model error arising from J2 perturbation. Simulation results show that by intial location and location refinement process, the accuracy of relative positioning can achieve centimeter-level.