Acta Scientifica Malaysia (ASM)

Water Pollution Model is a nonlinear system which present the random bias. The most common method is to use augmented state Cubature Kalman Filter, but the computational requirement of augmented state Kalman filter may become excessive. It is easily overflow and fail when running on digital computer. In this paper, two-stage Cubature Kalman filter is proposed to solve this problem. The estimate of two-stage Cubature Kalman filter can be expressed as the output of the advanced bias free filter and bias filter. Contrast augmented state Cubature Kalman filter, twostage Cubature Kalman filter is equivalent to the augmented state Cubature Kalman Filter in terms of computational accuracy, but computation is much smaller than augmented state Cubature Kalman Filter. The simulation results prove the validity of the two-stage Cubature Kalman filter in Water Pollution Model and prove the equivalence of the two algorithms.