Time varying parameter estimation with decomposition system

This paper presents an algorithm for on line parameter estimation with decomposition of MIMO linearized time varying system. The test data is the process and measurement noise. Generally speaking, the former online parameter estimation method for MIMO system needed either normalizing the system math model or a lot of calculation for inverse matrix in estimating processing. It is not only takes much operational time, but also easily diverges when the matrix is irreversible. The method that we present here hierarchically estimates parameters so that can avoid the inverse matrix and follow the time varying parameter step by step. That could raise the calculation precision and ensure the proceeding convergence. Using the simulation flight test data in processing and measurement noise, the method has been examined.