Ground subsidence induced by mining activities poses significant risks to ground structures. The Jiaozuo and Juye mining areas in eastern China are known for their extensive coal reserves, with deep-buried thick alluvial layers and thin bedrock. The thin bedrock and weak alluvial layers lead to easy deformation of the rock strata, resulting in ground subsidence. The current subsidence prediction models often neglect the impact of rock mass displacement processes and alluvial layer consolidation on ground subsidence, despite their significant influence on outcomes. Integrating these factors into subsidence prediction models is essential for improving forecast accuracy in mining regions, not only in China but also in other countries where mining is conducted. In this study, based on the Zhaogu No. 2 coal mine, the unmanned surface vehicle was used to observe ground subsidence, which reached 5.448 m. The influence of alluvial layer consolidation during the observation period contributed to a subsidence of up to 0.3 m. Consequently, incorporating the consolidation process of the alluvial layer can substantially enhance forecast accuracy. Additionally, this study analyzes the characteristics of ground subsidence under rock mass displacement processes, particularly focusing on the breaking of thin bedrock and the caving of alluvial layers. It proposes a new prediction model that integrates alluvial layer consolidation. In practical applications, this method reduces the average error by 7.6% and the maximum error by 20% compared with the probability integral method, providing a valuable tool for ground structure protection. Future improvements to the forecasting model could consider time factors.