高精度的林木应力波成像无损检测关键问题研究

Nondestructive testing is an important technology for monitoring the condition of forest resources and improving the utilization of wood. At present, stress wave imaging has become increasingly popular in wood nondestructive testing field.Traditionally, wood tomography based on straight ray tracing algorithm has not high accuracy enough, and sometimes it even deviates from the reality. This project aims at solving the existed problems ofstress wave imaging of wood, and focuses on studying the high-resolutioned definite difference simulation model of stress wave propagation in wood, non-linear two-dimensional wood tomography inversion and the three-dimensional wood tomography reconstruction. The proposed numerical simulation model will reflect the evolution of stress wave velocity, wavefront and energy precisely. The two-dimensional wood tomography inversing method based on anisotropy will be studied, and the problems of non-linear curve ray tracing and constrained tomography inversion equations will be solved.In addition, the mathematical model of three-dimensional tomography reconstruction will be presented. We will improve the traditional interpolation technique,and reduce the complexity of three-dimensional tomography reconstruction algorithms but also improve the accuracy of the generated three-dimensional tomography. Thus, some difficult problems in the traditional wood stress wave imaging technology will be solved, and the research results of this project can provide a good theoretical foundation for the development of low cost and practical stress wave imaging equipment with high resolution.

无损检测是森林资源健康监测和提高木材利用率的重要手段，应力波成像方法在活立木和木材内部缺陷的无损检测领域逐渐得到推广应用。传统的木材应力波成像方法以线性的直线路径跟踪算法为基础，检测精度不够高，有时甚至严重偏离实际。本项目拟重点研究应力波在林木内部传播的高精度有限差分模拟方法、二维应力波断层图像的非线性反演方法和三维应力波图像重构方法等三方面内容。研究目标在于使提出的数值模拟模型准确反映应力波的传播速度、波阵面和能量衰减等属性的变化规律；探索基于各向异性的林木二维断层图像反演方法，解决非线性的应力波曲线路径跟踪和图像反演过程中的约束求解问题；建立三维图像重构的数学模型，改进传统的内部插值方法，并在降低重构算法复杂度的同时提高三维图像的准确性。由此，解决传统林木应力波成像检测技术存在的几个关键问题，显著地提高检测精度，并为研发高精度、低成本和实用的林木应力波成像检测设备打下坚实的理论基础。

无损检测是森林资源健康监测和提高木材利用率的重要手段，应力波成像方法在活立木和木材内部缺陷的无损检测领域逐渐得到推广应用。传统的木材应力波成像方法以线性的直线路径跟踪算法为基础，检测精度不够高，有时甚至严重偏离实际。为了提高应力波成像检测精度，本项目重点研究了应力波在林木内部传播的传播速度模型、二维应力波断层图像的非线性反演方法和三维应力波图像重构方法等等关键科学问题。项目组建立了应力波在树木横截面的传播速度模型和径切面内的传播速度模型；提出了基于椭圆插值与速度补偿的应力波断层成像方法、基于速度误差校正的应力波断层成像方法、基于曲线路径跟踪的应力波断层成像方法；提出了一种基于多断层图像空间插值的三维成像方法和一种基于传感器随机立体部署直接三维成像方法。开发了相关的成像软件，并研发成功便携式林木应力波无损检测仪。项目组共发表学术论文18篇，其中被SCI收录8篇；申请国家发明专利6项，其中5项获授权。研究成果在江苏、浙江等地的古树名木保护工程中推广应用，产生了较好的社会和生态效益。