Li, Xian; Xie, Jiangqiong; Wang, Zhiguo; Shen, Xiaojing Time-varying Gaussian process assume density filter algorithm. (Chinese. English summary) Zbl 1463.60056 J. Syst. Sci. Math. Sci. 40, No. 4, 578-586 (2020). Summary: Gaussian process is an effective data-driven modeling method, which has been applied to solve the state estimation problem of time-invariant dynamic systems. In this paper, in order to improve the adaptive ability of the Gaussian process dynamic system, the timer-varying dynamic system is considered, and the parameters are updated by the particle filter algorithm. The updated parameters are substituted into the Gaussian process assumed density filter algorithm to obtain the time-varying Gaussian process assumed density filter algorithm. A simulation example verifies the effectiveness of the proposed algorithms. MSC: 60G15 Gaussian processes 60G35 Signal detection and filtering (aspects of stochastic processes) Keywords:Gaussian process; dynamic system; particle filter algorithm; density filter PDFBibTeX XMLCite \textit{X. Li} et al., J. Syst. Sci. Math. Sci. 40, No. 4, 578--586 (2020; Zbl 1463.60056)