Sensor alignment Kalman filters for inertial stabilization systems

To achieve a particular set of motion compensation requirements and transfer alignment requirements a device called a motion sensor subsystem (MSS) was developed. The MSS is a strapdown inertial navigation system coupled with two decentralized Kalman filters that perform transfer alignment and estimation of various SINS instrument errors and system angular misalignments. One Kalman filter transfer aligns the MSS to the master navigator and the other filter estimates and corrects the misalignments between the MSS instrument axes and the sensor axes. The design, simulation, implementation, and flight testing of the MSS are described. These filters are coupled together through a process called regeneration that transfers one filter's shared state estimate to the other. In addition, because the filters operate in a closed-loop system, the computational time delay incurred by the filters significantly affects loop stability. A method of providing loop stability is discussed. The results of the MSS flight test have shown that using a methodology for developing a large Kalman filter based on intensive system covariance analysis and Monte Carlo simulation yields a system design that meets requirements.<>