Evidential clustering based on the theory of belief functions has become one of the topics of machine learning due to its ability to characterize the uncertainty and imprecision between clusters. However, there are still problems such as ambiguity of basic concepts, high computational complexity, and inability to detect imbalanced and arbitrary clusters effectively. This thesis is devoted to addressing the above problems and proposes corresponding solutions, which include four parts. First, we comprehensively survey existing evidential clustering algorithms and give related concepts and definitions, detailing why evidential clustering can characterize the uncertainty and imprecision between clusters. Second, we propose a dynamic evidential ...