A navigation accuracy compensation algorithm for low-cost unmanned surface vehicles based on models and event triggers

Integrated navigation is often used for navigation of unmanned surface vehicles (USVs). For a class of costsensitive USVs, it is desirable to compose integrated navigation by low-cost inertial navigation systems (INS) and global positioning systems (GPS). In the cases of GPS signal failure or interference, only INS onboard can be used for navigation, which often leads to the navigation accuracy deteriorating and may fail to meet application requirements. Therefore, a navigation accuracy compensation algorithm for low-cost USVs is proposed. An eventtriggering mechanism is introduced, and the thrust models of the USVs are adopted to design the event trigger. According to different events, various weights can be assigned to the integrator based on INS, and measurementbased linear velocity VINS and heading angle psi INScan be obtained. Moreover, based on dynamic models, the model-based linear velocity Vm and heading angle psi m also can be calculated. Meanwhile, the event trigger is combined with the Particle Filter (PF) algorithm to accurately estimate the linear velocity psi PF for navigation. Finally, field experiments are conducted and the differential GPS (DGPS) is used as the benchmark for comparison. The results show that our algorithm can improve navigation performance for the low-cost USVs. VPF and heading angle