부산대학교 도서관

In this article, the problem of state estimation is investigated for stochastic descriptor linear systems with inaccurate process and measurement noise covariance matrices. Under the assumption that the noise and its covariance matrix, respectively, obey the Gaussian distribution and the inverse Wishart distribution, a novel adaptive moving horizon estimator is designed for descriptor systems based on the variational Bayesian inference method. By iteratively updating their approximate posterior probability distributions, the states of descriptor systems are estimated when noise covariance matrices are inaccurate. Finally, the feasibility and effectiveness of the proposed method are verified by experiments performed on the prototype of a dynamic point-the-bit rotary steerable drilling tool.