Research on a Quality Inspection Method for 3D Deliverables of Pumped Storage Power Stations Based on Point Cloud Data

With the increasing application of Building Information Modeling (BIM) and 3D collaborative design in pumped storage power stations, the quality and reliability of 3D deliverables have become crucial for ensuring engineering accuracy and lifecycle management. However, conventional manual inspection methods are time-consuming, subjective, and difficult to scale for complex hydropower projects. To address this challenge, this paper proposes a quality inspection method for 3D deliverables of pumped storage power stations based on point cloud data. The method integrates point cloud acquisition, preprocessing, and geometric feature extraction with model–point cloud alignment techniques to automatically evaluate dimensional accuracy, surface completeness, and spatial consistency of design models. A set of rule-based and statistical indicators is established to quantify deviations between the as-designed 3D model and the as-built point cloud data. Case studies conducted on a pumped storage power station project demonstrate that the proposed method achieves high precision in detecting geometric deviations, with inspection efficiency improved by more than 45% compared to traditional approaches. The results indicate that this method not only enhances the reliability of 3D deliverables but also provides technical support for engineering quantity extraction and digital handover in hydropower construction projects.