Sensors · 2017 · 237 citations · 17 references
EngineeringSmart CityGlobal Navigation Satellite SystemPositioning SystemReal Urban ScenariosLocalizationSocial SciencesGlobal Positioning SystemCalibrationSystems EngineeringPositioningTransportation EngineeringAutomatic NavigationInertial SensorsComputer EngineeringUrban PlanningGeodetic NetworkSatellite Navigation SystemsInertial Navigation SystemsAerospace EngineeringCivil EngineeringTight Gnss/ins IntegrationsTerse Mathematical ModelTight IntegrationGlobal Satellite Navigation SystemsTransportation Systems
GNSS remains the primary positioning method, yet its accuracy degrades in many environments, and while tight‑coupled GNSS/INS systems offer benefits, they are usually limited to high‑grade IMUs. This study evaluates the performance gains of a tightly‑coupled GNSS/INS integration that employs low‑cost sensors and a mass‑market GNSS receiver. The authors develop concise tight‑coupling algorithms, then test them in real urban scenarios and compare the results against commercial modules operating in standalone or loosely‑coupled modes. The experiments demonstrate that the low‑cost tight‑coupled system achieves practical performance improvements over standalone and loosely‑coupled alternatives in urban environments.
Global Navigation Satellite Systems (GNSSs) remain the principal mean of positioning in many applications and systems, but in several types of environment, the performance of standalone receivers is degraded. Although many works show the benefits of the integration between GNSS and Inertial Navigation Systems (INSs), tightly-coupled architectures are mainly implemented in professional devices and are based on high-grade Inertial Measurement Units (IMUs). This paper investigates the performance improvements enabled by the tight integration, using low-cost sensors and a mass-market GNSS receiver. Performance is assessed through a series of tests carried out in real urban scenarios and is compared against commercial modules, operating in standalone mode or featuring loosely-coupled integrations. The paper describes the developed tight-integration algorithms with a terse mathematical model and assesses their efficacy from a practical perspective.
17