To address challenges related to topic ambiguity and inaccurate alignment problems in the existing deep document clustering methods when processing dynamic textual data with topics varying with time， a Deep Evolutionary Topic Clustering Model （DETCM） was proposed. In DETCM， information of topics varying with time was captured in dynamic text， and historical topic information was integrated with document features of the current time slice， thereby discovering event topic trajectories and generating dynamic document topic representations. In specific， to solve topic ambiguity problem of topics varying with time， a topic fusion learning module based on a hybrid encoder was designed， in which topic information from preceding time slices was utilized to enhance topic discrimination and feature extraction of the current time slice. Furthermore， a topic inheritance module across different time slices was designed to achieve topic match alignment across different time slices， so that topic information on historical slices was effectively transferred and incorporated into cluster assignment process of the current time slice. Experimental results based on the real-world arXiv evolving textual document dataset demonstrate that compared with the Deep Evolutionary Document Clustering model with Instance-level Mutual Attention Enhancement （DEDC-IMAE）， DETCM achieves an average improvement of 3.08% （-0.37% to 5.43%） in Normalized Mutual Information （NMI） across all time slices， verifying the superior capability of DETCM in tracking topic evolution under dynamic scenarios， enabling more accurate capture of temporal variation features in topics and leading to better clustering performance.