摘要：Since the landm.rk work of R. E. Kalm.n in the 1960s, considerable efforts have been devoted to tim. series state space m.dels for a large variety of dynam.c estim.tion problem.. In particular, param.tric filters that seek analytical estim.tes based on a closed-form.m.rkov-Bayes recursion, e.g., recursion from.a Gaussian or Gaussian m.xture （Gm. prior to a Gaussian/Gm.posterior （term.d ＇Gaussian conjugacy＇ in this paper）, form.the backbone for a general tim. series filter design. Due to challenges arising from.nonlinearity, m.ltim.dality （including target m.neuver）, intractable uncertainties （such as unknown inputs and/or non-Gaussian noises） and constraints （including circular quantities）, etc., new theories, algorithm., and technologies have been developed continuously to m.intain such a conjugacy, or to approxim.te it as close as possible. They had contributed in large part to the prospective developm.nts of tim. series param.tric filters in the last six decades. In this paper, we review the state of the art in distinctive categories and highlight som. insights that m.y otherwise be easily overlooked. In particular, specific attention is paid to nonlinear system. with an inform.tive observation, m.ltim.dal system. including Gaussian m.xture posterior and m.neuvers, and intractable unknown inputs and constraints, to fill som. gaps in existing reviews and surveys. In addition, we provide som. new thoughts on alternatives to the first-order m.rkov transition m.del and on filter evaluation with regard to com.uting com.lexity.