<p>The Moon, as the only natural satellite of the Earth, holds significant importance for understanding the history of its formation and evolution through the accurate determination of its surface mineral composition. It also provides a scientific basis for the reasonable development and utilization of lunar resources in the future. This study proposes a new method for lunar mineral abundance inversion. First, the data sources and preprocessing methods are introduced, followed by the establishment of the mineral abundance inversion model. The scientific validity of the new inversion method is verified through experimental implementation. The study shows that the Lunar Mineral Mapper, as a hyperspectral imaging instrument, can reflect the mineral distribution characteristics on the lunar surface. The Apollo mission samples offer reliable ground-truth reference data for quantitative analysis of mineral abundance. The mineralogical analyses of these samples are highly accurate and authoritative. By fusing the mineral composition data from Apollo samples with the remote sensing data of the Lunar Mineral Mapper, the problem of insufficient accuracy in remote sensing data can be effectively mitigated. Sample constraints improve the accuracy and stability of the inversion model. The conclusion is that the new method of mineral abundance inversion based on the fusion of Lunar Mineral Mapper data and Apollo samples is not only a key approach to overcoming the limitations of traditional single data source inversion but also a necessary means to improve the precision of quantitative mineral analysis on the Moon.</p>