Research On The Data Processing Technology Of Point Cloud Applied To The Inner Wall Reconstruction Of Oil Tank

The large-scale single-layer steel oil storage tanks widely used at present are generally corroded inside the wall,which is very likely to cause leakage of internal oil and gas and pollute the surrounding soil.The demand for double-layer transformation of the oil storage tank is urgent.Considering safety and efficiency,how to realize the non-excavation reconstruction of deep underground oil storage tanks is a big problem.In view of the extensive use of multi-axis manipulator in the industrial field,this paper uses six-axis manipulator to realize the automated transformation of tank inner wall.The base of the manipulator is prized at the entrance of the oil tank.By replacing the end modification tool of the manipulator,the modification processes of sand blasting,rust removing,resin spraying and rolling on the inner wall of the oil tank are completed,thus realizing the double-deck modification of the oil tank.And the laser scanning equipment is used to obtain the geometric features of the inner surface of the tank,so as to ensure that the end modification tool of the manipulator can form a uniform thick and thin modification layer on the inner wall of the tank.In this paper,the point cloud data processing technology applied to the tank inner wall reconstruction is studied.The main contents include an adaptive hybrid point cloud topology,a denoising algorithm for eliminating outliers from point clouds of oil tanks,and a method for determining the trajectory of the end of a manipulator based on point cloud data.In order to realize automatic and efficient tank inner wall modification,the hardware platform of the manipulator and the development of human-computer interactive control system are built in this paper.The functions of the software system include point cloud data processing,end path planning,manipulator status monitoring and data communication.Generally speaking,the main achievements of this paper are as follows:(1)The hardware equipment of the manipulator is selected and the hardware platform of the manipulator system is built according to the requirement of the modification of the inner wall of the oil tank.(2)Traditional point cloud topology construction methods,such as raster method,octree method and K-D tree method,have the disadvantages of large space overhead and long time-consuming neighborhood search.This paper proposes a new method to establish hybrid point cloud topology based on raster method and octree method.This method can adaptively establish the topological relationship between the point cloud data of oil tank according to the transformation process.Because the traditional statistical filtering method has the disadvantage of long time-consuming and poor denoising effect in eliminating outliers in oil tank point cloud,this paper improves the statistical filtering method.When eliminating outliers from oil tank point cloud,the denoising effect and denoising efficiency are greatly improved.(3)The information needed by the manipulator in the reconstructing operation includes the path coordinates of the trajectory of the reconstructing tool at the end of the manipulator on the wall of the oil tank and the attitude angle at the end of the corresponding position.In this paper,the moving least squares method is used to fit the local point cloud.The curve is used as the trajectory of the end of the manipulator on the inner wall of the tank.The coordinates of the end path point can be obtained by discretizing the fitting curve.The attitude angle information is determined by the normal vector of the point cloud corresponding to the path point.This paper improves the traditional principal component analysis method.After using the adaptive K-neighborhood search method to get the neighborhood points of the path points,the covariance matrix is weighted bilaterally by the distance and curvature weights,which makes the points closer to the path points and the points with greater curvature change contribute more to the normal vectors of the path points,and the calculated normal vectors can better reflect the path.Local characteristics at the radius point;(4)Lab VIEW is used to develop the operation interface of human-computer interaction system,which integrates the functions of point cloud data processing,path planning at the end of the manipulator,condition monitoring of the manipulator and data communication.Based on the experimental verification of the manipulator system,the results show that using the path information obtained from the point cloud data processing based on the algorithm in this paper,the manipulator has a good effect on the modification of the inner wall of the tank and meets the actual requirements of the modification.