Algorithm Based Error Estimation & Navigation Correction

ABSTRACT:The guidance and navigation capabilities of the majority of the strategic systems have been and still are largely based on autonomous Inertial Navigation Systems (INS) that process measurements from on-board high-precision Inertial Measurement Units (IMUs). Because these systems do not rely on any external aiding sensors, the errors can accumulate over time. In order to improve the accuracy and responsiveness of such systems without requiring costly hardware changes, we desire algorithmic-based solutions that require only modifications to the navigation system software. Under the Phase I effort, STR developed a nonlinear filtering approach, based on the homotopy-based (flow-based) particle filter, and showed that it has the potential to greatly improve navigation accuracy. This effort will build upon these concepts to develop and demonstrate a prototype nonlinear filtering navigation system. We will evaluate the algorithms using realistic trajectories and IMU error models, and perform sensitivity studies to understand the impact of different errors, and tradeoffs in computation and accuracy. We will also perform a feasibility study on transitioning the capability to flight capable hardware for different missile systems. BENEFIT:The proposed effort will provide strategic strike vehicles with effective means for accurate navigation and guidance, using only the on-board IMU, thereby reducing the susceptibility to various kinds of interference and countermeasures, and enabling operational capability under GPS denied environments. The techniques developed by this proposed effort will also provide commercial non-GPS navigation applications.