Abstract: In recent years, the coal mining industry in China has been experiencing rapid growth, resulting in an increasing adoption of intelligent technologies. Among these advancements, precise positioning and navigation technology for open-pit coal mining environments have become crucial. However, Simultaneous Localization and Mapping (SLAM), a key technology for unmanned driving, is currently facing significant challenges in open-pit coal mines. Limited environmental feature points and environmental degradation have necessitated SLAM to rely solely on sparse feature points for localization and mapping, thus increasing its complexity. Furthermore, sensor jitter caused by slopes and uneven roads can lead to motion distortion during robot operation. To address these challenges, a novel solution is proposed in this paper. Firstly, the external parameters of the sensors are being recalibrated. Secondly, the integration of inertial guidance and LiDAR is being utilized to improve data consistency and accuracy. This approach aims to enhance the performance of SLAM in open-pit coal mines, improving localization accuracy and mapping effectiveness. Building upon this foundation, our approach is leveraging full feature point matching to directly down sample and extract the point cloud from the LiDAR data. To enrich the preprocessed laser point cloud data, Iterative Closest Point (ICP) matching is being incorporated at the front-end of the algorithm, facilitating the extraction of the key-frame point cloud X. Subsequently, this data is being integrated with inertial guidance information to correct aberrations in the point cloud, leading to the formation of the refined point cloud P. ICP matching is once again being employed to align X and P. Furthermore, a factor graph is being incorporated into our back-end to enhance loopback detection, strengthening constraints and further improving localization accuracy and mapping effectiveness in open-pit coal mine environments. Experimental results demonstrate the high localization precision and undistorted map building capabilities of our proposed algorithm. Notably, the sidewall texture remains clear, exhibiting a certain degree of robustness, effectively enhancing both robustness and accuracy in open-pit coal mine settings.