研究生知识库

It’s significant to research and explore ore body with geophysical inversion methods. This dissertation is concerned with the geophysical inversion problem which is ill-posed and has non-unique solution. The goal of the thesis is to review and devleop new techniques for solving such problems. It’s universally known that a appropriate computational tool is considerabley significant and necessary for inversion. In adition, these problems are not easy to solve because of the multiplicity of local and global minima. The thesis can be divided into three parts. In the first part, field forward problems are reviewed simplly, Emphasis is given to Geometric Trellis Storing, which is to separate geometric trellis from forward formula that is invariable during the whole process of inversion , a...

[ More ]

It’s significant to research and explore ore body with geophysical inversion methods. This dissertation is concerned with the geophysical inversion problem which is ill-posed and has non-unique solution. The goal of the thesis is to review and devleop new techniques for solving such problems. It’s universally known that a appropriate computational tool is considerabley significant and necessary for inversion. In adition, these problems are not easy to solve because of the multiplicity of local and global minima. The thesis can be divided into three parts. In the first part, field forward problems are reviewed simplly, Emphasis is given to Geometric Trellis Storing, which is to separate geometric trellis from forward formula that is invariable during the whole process of inversion , and to store the trellis as a array which would make the following reiterative forward computation as a simple reading of array and multiply it with corresponding cell property. Therefore, the technique facilitates the inversion of gravity and magnetic anomalies faster. In the second part of this thesis, linear and nonlinear optimization algorithms are discussed, especially attention for genetic and differential evolution theory. The later has been shown to be a simple yet efficient optimization approach in solving a variety of benchmark problems as well as many real-world applications. In this paper, it is used for improving the traditional GA which may perform badly with the increasment of varible number. Simultaneously, projection-based exploration search method and adaptive directed mutation are also adopted. Based on these, a novel, simple and efficient real coded hybrid genetic algorithm is proposed and then employed to solve complex function optimization problems. Test results indicate that the proposed algorithm is fast, accurate, and reliable. Finally, improved GA is applied to geophysics inversion, then gravity and magnetic model tests demonstrate that this technique is preeminent, consiquently, it is available to apply the improved GA to joint inversion and field data inversion.