Square root sigma point filtering for real-time, Nonlinear estimation

A square root version of the sigma point filter, which is suitable for on-line real-time usage, is presented. Covariance square roots are required for this scheme's base implementation. It is observed using simulations that this system can be used to track jumps in the data, which can occur as failures in the actual mission. Simulation results have also indicated that the system is useful for accurately developing online aerodynamic models, online detection of vehicle capabilities after failure or damage, and soft uncertainty bounds.