Abstract: To address the challenge that traditional hardware emulation techniques struggle to accurately reproduce the dynamic and smooth evolution characteristics of low-altitude channels, this paper proposed a Field Programmable Gate Array (FPGA)-based hardware emulation technique for smoothly evolving channels in low-altitude scenarios. First, continuous multi-path clustering in dynamic time-varying channels is realized by combining a Gaussian mixture model (GMM) clustering algorithm, whose initial parameters are optimized by a K-means++ algorithm based on the distances of multi-path components, with a cluster birth-death tracking algorithm. Then, a dynamic polyphase filter and a cluster amplitude linear interpolator, based on cluster movement and birth-death decisions, are utilized to perform smooth filtering on the input signal. An evolution control module is used to manipulate these components, thereby achieving a smooth evolution of the channel delay and power parameters while accounting for cluster birth and death. The test results indicatethat the power delay profile (PDP) generated by the proposed method exhibited a trend consistent with field measurement results. Furthermore, the correlation between clustering results at adjacent time instants is improved compared to the previous method, and the smooth evolution characteristic of the channel power-delay parameters is significantly enhanced. This work provides an effective means for the future optimization, verification, and evaluation of low-altitude communication systems.