Research On Three-Dimensional Imaging Technology Of Log Internal Defects Based On Stress Wave Regression Model

As the natural material,wood is an ideal raw material in household,construction,transportation and other fields.Defects such as voids,cracks and decay in wood seriously affect the physical and mechanical properties,reduce its quality and affect its use value.Nondestructive testing technology can detect and image internal defects from the outside,which is of great significance to improve the utilization rate of wood,predict the characteristics of plate and evaluate ancient wood structures and ancient trees.In the view of above,this paper takes the internal defects of logs as the research object,and adopts stress wave nondestructive testing technology to conduct high-precision 3D imaging research.As a kind of nondestructive testing technology,stress wave technology has the advantages of long propagation distance,strong anti-interference ability and convenient use,and has become one of the important means to detect internal defects of logs at home and abroad.However,when stress wave technology is used in log defect detection,some problems remain to be solved,such as single structure of propagation velocity model,difficulty in obtaining propagation velocity value at any angle in log 3D space,and low accuracy of 3D defect imaging.Therefore,this paper firstly analyzes the propagation velocity of stress wave in log space,establishes a regression model of stress wave propagation velocity and spatial angle in log space,and optimizes the regression model under the condition of considering the influence factors of moisture content.Finally,the research on high-precision 3D inversion imaging based on improved IDW and improved Kriging interpolation method is carried out,which provides the theoretical basis and technology for high-precision detection of internal defects of wood,ancient wood structures and ancient trees.The main research work of this paper is as follows.Through theoretical analysis,the theoretical model of stress wave propagation velocity in log is obtained.Non-defect logs of Betula platyphylla,Fraxinus mandshuric,Ulmus pumila and Larix gmelinii were taken as test samples,and the propagation velocity of stress waves in different direction angles in the longitudinal section of the samples and different included angles in the cross section were measured by Arbotom stress wave tester.Using regression analysis method,taking the theoretical model as the objective function,the experimental model of propagation velocity and direction angle in longitudinal section and the experimental model of propagation velocity and included angle in cross section were established respectively.On this basis,the experimental model of stress wave propagation velocity in larch log sample space is established.The results show that the correlation coefficient of each model is above 0.82,and the correlation coefficient of the regression model of propagation velocity in larch log space is as high as 0.98,which has a good goodness of fit and verifies the correctness of the regression theoretical model.Based on the theoretical model of stress wave propagation velocity,the influence of moisture content on the propagation velocity of stress waves under different direction angles and section angles of the larch samples was explored,and the regression optimization model of propagation velocity under different moisture content was constructed.A two-dimensional image superposition algorithm based on propagation velocity model was proposed to carry out tomography test on longitudinal section of logs with cavity defects.The results show that the correlation coefficients of the optimized model are 0.97 and 0.99 respectively,which has high goodness of fit.The optimized model plays a good guiding role in the detection of log internal defects,and the fitting degree of the reconstructed defect image is as high as 95.38%,which realizes the quantitative characterization of the defect area inside the log,and provides a theoretical and experimental basis for stress wave three-dimensional imaging technology.Based on the optimization model of stress wave propagation velocity,an improved strategy for 3D imaging of log internal defects based on inverse distance weighting algorithm(IDW)is proposed,which is improved from three aspects:the appropriate number of sampling points is determined,the sampling space is divided based on the azimuth search method,and the mutation function is introduced to calculate the weight coefficient.Three-dimensional inversion imaging experiment was carried out on the defective larch log samples,and the imaging results were compared with those of the classical IDW and cross-verified.The results show that the interpolation effect of the improved IDW is significantly better than that of the classical IDW.The 3D image is smoother,the deviation calculation index is lower,and the imaging accuracy is significantly improved.However,there are still some defects in the 3D imaging effect,and the overall interpolation accuracy still has room for further improvement.Aiming at the research of 3D imaging method of log internal defects,particle swarm optimization(PSO)is proposed to optimize the solution process of Kriging interpolation variogram model.This paper analyzes the problem that chaotic PSO lacks standard to measure the diversity of particle population,and proposes a greedy chaotic PSO based on particle dimension entropy,which describes the diversity of particle population by constructing a particle dimension entropy,adopts a greedy strategy to reduce the iteration time of chaotic PSO,and compares and verifies the optimality of the algorithm by using complex nonlinear function optimization experiments.Through the improved Kriging interpolation method,3D inversion imaging experiments were carried out on the defect samples.The results show that the spatial interpolation accuracy and stability of the improved Kriging interpolation method are obviously improved,and there is no "granular" burr in the imaging results,and the surface is smooth and continuous,which is better than the ordinary Kriging interpolation method and the improved inverse distance weighting algorithm,and realizes the high-precision 3D imaging of log internal defects.