Adaptive Fusion of Inertial Navigation System and Tracking Radar Data

Abstract

Against the range-dependent accuracy of the tracking radar measurements including range, elevation and bearing angles, a new hybrid adaptive Kalman filter is proposed to enhance the performance of the radar aided strapdown inertial navigation system (INS/Radar). This filter involves the concept of residual-based adaptive estimation and adaptive fading Kalman filter and tunes dynamically the filter parameters including the fading factors and the measurement and process noises scaling factors based on the ratio of the actual residual covariance to the theoretical one. In fact, due to the unknown and fast-varying statistical parameters of the radar measurement noises and their nonlinear characteristics, applying a conventional Kalman filter to INS/Radar data fusion yields a low performance navigation and in-flight alignment. The Monte Carlo simulation results of the integrated navigation system on an interceptor missile trajectory indicate the new algorithm has an effective performance in face of nonlinearities and uncertainties of the tracking radar measurements. these results allow to know whether the fine in-flight alignment and high performance navigation can be possible for the long-range air defense missile using the low-cost INS/Radar system without aiding global navigation satellite system signals or not.