Accurate classification of the vertical distribution of tropospheric aerosols is critical for reducing uncertainties in climate effect assessments. To address the challenge of aerosol classification uncertainties inherent in traditional lidar retrievals under complex mixed scenarios, this study leverages the unique locational advantage of the Atmospheric Boundary Layer Eco-Environment Shanghuang Observatory (ABLES) to develop an advanced synergistic retrieval algorithm based on a multi-wavelength Mie-Raman-fluorescence lidar system. The proposed scheme establishes a seven-parameter synergistic constraint, integrating fluorescence capacity, particle depolarization ratios (<em>PDR</em>), backscatter-related Ångström exponents (<em>BÅE</em>), and lidar ratios (<em>LR</em>). By combining Monte Carlo simulations with least squares minimization, the algorithm achieves a quantitative decomposition of scattering contribution fractions for smoke, urban, pollen, and dust. A key advantage is the robust physical constraint system, which ensures classification relies on intrinsic microphysical properties rather than signal intensity alone, thereby avoiding biases from backscatter anomalies. Multi-platform cross-validation confirms the high reliability of the algorithm across a wide dynamic range, with the coefficient of determination between near-surface retrieval results and in situ monitoring data exceeding 0.6. Furthermore, sensitivity analysis indicates that the multi-parameter scheme effectively captures the differential microphysical responses of aerosols across seasons and altitudes. This physically decouples meteorologically driven optical enhancement from actual mass concentration fluctuations, providing strong technical support for high-precision, high-spatiotemporal-resolution aerosol typing and mass retrieval at high-altitude background stations.