Urban real-estate markets represent typical grey systems featuring incomplete information, multi-factor coupling and dynamic uncertainty, where conventional statistical approaches fail to capture medium-long-term nonlinear evolution. This paper establishes an integrated forecasting-evaluation framework using updated grey-system theory. Based on Tai’an’s 2016-2025 real-estate time-series data, an improved dynamically-weighted GM(1,1) with polynomial residual correction forecasts market scale, price trends and supply-demand structure for 2026-2035. Dynamic grey relational analysis and grey clustering quantitatively rank influencing factors and assess market quality, while revealing latent grey risks. Empirically, the improved model cuts MAPE by 6.8% and RMSE by 3.5% versus standard GM (1,1), enabling reliable 10-year projections. Residents’ disposable income, urban-construction investment and population-structure changes are dominant strong-correlation factors (correlation > 0.85). Over 2026-2035, Tai’an’s market will undergo steady-scale adjustment, rational price swings, improved supply-demand matching and accelerated industrial transition. Core grey risks stem from land-resource limits, inadequate supporting amenities and real-estate financial volatility. Multi-layer suggestions targeting governments, enterprises and urban planning are put forward. This work offers theoretical foundations and decision-making references for sound real-estate development in small-and-medium-sized cities exemplified by Tai’an.