Extension of Friedland's separate-bias estimation to randomly time-varying bias of nonlinear systems

By extending B. Friedland's (1969) separate-bias estimation algorithm for linear systems to nonlinear systems and combining the result with the suboptimal fading extended Kalman filter proposed by D.H. Zhou (1990) and by D.H. Zhou et al., a pseudo-separate-bias estimation algorithm for randomly time-varying bias of a class of nonlinear time-varying stochastic systems is obtained. A simulation example is presented to illustrate the effectiveness of the algorithm. >