Optimization Inversion of the Relative Weight Ratioλ in the Joint Inversion Models
DU Zhixing1,2, OU Jikun1, JIN Fengxiang3, HAN Baomin1, CHAI Yanju1(1. Institute of Geodesy and Geophysics, Chinese Academy of Sciences, Wuhan 430077, China; 2. Wuhan University GPS center, Wuhan 430079, China; 3. Institute of Geoscience and Technology, Shandong University of Science and Technology, Taian 271019, China)
How to choose the relative weight ratio λ is an important problem deserved to be studied in inversion models when using more than two physical measurements including the geodetic displacements to implement the joint inversion. After expounding the traditional methods of calculating the value of the relative weight ratio λ, the idea of computing the λ with it and other inversion parameters as the parameters to be inversed is put forward, and the inversion model is derived in detail. In addition, the method of resolving this problem employing the direct inversion is also clarified. At last, an example that simulated two schemes with six different cases is illustrated. The results show that: ① the iteration of inversion is convergent and the material parameters are quite close to their simulation values; ② the value of the relative weight ration λ is influenced by the distribution of the observable physical measurements; ③ the value of relative weight ratio λ tends to be stable with the iteration closing to convergence, which means that this value is much reliable than the others.