Online measurement of vibration displacement of mine hoisting wire rope based on 3D point cloud

Aiming at the problem of accurate online measurement of spatial vibration displacement of mine hoisting wire rope, an online measurement method based on a 3D point cloud is proposed. For the problems of background interference and outliers in the wire rope point cloud segmentation, the wire rope point cloud segmentation method combining the random sampling consistency algorithm (RANSAC) and density-based clustering algorithm (DBSCAN) is adopted to achieve the effective separation of the wire rope point cloud and background. To address the loss of geometric features in the point cloud downsampling process of the traditional voxel grid method, an adaptive voxel grid algorithm is proposed, which firstly determines the optimal initial voxel size by using Particle Swarm Optimisation (PSO), and then dynamically adjusts the voxel size according to the local density of the wire rope point cloud to achieve efficient streamlining of the point cloud of the wire rope. For the problem of wire rope vibration displacement calculation, a dynamic measurement method based on segmental least squares fitting is proposed, which uses the least squares method to linearly fit the point clouds of each segment and introduces error weights to calculate the coordinates of the centroid. Then it uses the relative coordinate method to calculate the displacement between the dynamic centroid and the static reference point, to achieve an accurate measurement of the spatial vibration displacement of the wire rope. To verify the effectiveness of the proposed method, experiments are conducted in the laboratory and mine hoisting environment. The experimental results show that the point cloud segmentation method combining RANSAC and DBSCAN can effectively extract the wire rope point cloud from the complex background, and the precision and recall are above 93.9% and 94.3%, respectively; compared with the traditional voxel grid algorithm, the streamlined wire rope point cloud of the adaptive voxel grid algorithm is smoother and more continuous, and the maximal and average errors are reduced by 68.7% and 74.7%, respectively, at the streamlining rate of 94.5%. In the laboratory environment, the average absolute error of the vibration displacement measurement method proposed in this paper is less than 0.62 mm in both X and Z directions, and the maximum error is within 1.2 mm; in the actual mine hoisting environment, the average absolute error is less than 0.87 mm in both X and Z directions, and the maximum error is within 2.5 mm. It takes 20 ms to complete a displacement measurement, which can meet the accuracy and real-time requirements of mine hoisting wire rope vibration displacement measurement.