Abstract
A posteriori probability analysis method was proposed based on a parallel optimization algorithm, and applied to quantitative analysis of matched inversion results. The frequently appeared the uncertain of convergence criterion was resolved. The problems of harmony between sample number and calculation time, as well as between sample number and sample quality in higher order statistics calculation were achieved. Examples based on synthetic data show that posteriori probability analysis for matched field inversion can be conducted effectively by the proposed parallel optimization algorithm.
| Original language | English |
|---|---|
| Pages (from-to) | 2873-28779 |
| Number of pages | 25907 |
| Journal | Xitong Fangzhen Xuebao / Journal of System Simulation |
| Volume | 17 |
| Issue number | 12 |
| State | Published - Dec 2005 |
Keywords
- A posteriori probability distributions
- Inversion
- Matched field process
- Optimization