首页 | 本学科首页   官方微博 | 高级检索  
     检索      

半马尔可夫过程在详细沉积断面分析中的应用
引用本文:金友渔.半马尔可夫过程在详细沉积断面分析中的应用[J].沉积学报,1990,8(1):133-142.
作者姓名:金友渔
作者单位:中国地质大学 武汉
摘    要:地层模型中的半马尔可夫过程分析同时考察岩性状态的转移和状态的延续时间--层厚分布,更全面地反映沉积过程,但在求解半马尔可夫过程转移概率时需要进行拉普拉斯逆变换,造成实际应用的困难.本文从另一思路导出了直接计算半马尔可夫过程转移概率的方法,并以某受潮汐影响的三角洲体系的详细沉积断面为例介绍半马尔可夫过程分析在定量地研究沉积作用过程及聚煤规律中的应用。

关 键 词:详细沉积断面    半马尔可夫过程    逐步回归分析    因子分析    剩余分析
收稿时间:1987-07-04

APPLICATION OF SEMI- MARKOV PROCESS IN ANALYSIS OF THE DETAILED DEPOSITION
Institution:Geological University of China, Wuhan
Abstract:Probability of transition of semi-Markov process means that when point A of section just is located upper bound of ith lithic stratum probability Pij (1) of point B is in jth lithic stratum, point B is upper the point A and distance between point B with A is l.This probability can be decomposed as two events probability, the two events are in what follows: (1) In point A, the ith lithic stratum translats to the jth lithic stratum with only passing one step and thickness of the jth lithic stratum is lager than 1; (2) In point O that is between piont A and B, the ith lithic stratum translats to the jth lithic stratum with passing n steps (n = 2, 3, …0, this is that n-1 stuatum layering are contained between point A and O apart l1 it translats at once to jth lithic stratum in point O and the jth lithic stratum thickness is larger than 1- l1. For the 1th event, if only a stratum layering thickness distribution is gave, the event probability can be computed, if thicknees .distribution of n s tratum layerings can be gave, the Zth event probability can be easy computed. Thus the author supposes that n stratum layering thicknesses submit independently same parameter - distrbution: Among others x is the stochastic variable of stratum layering thickness, v, a are the distribu- tion parameters and v 0, a 0. As a result, probability density function of n stratum layerrings thickness sum can be derived from characteristic function of n independent stochastic variables sum distribution: Proceed to the next step, directly comput formula of probability of transition of semi- Markov process can be derived. The author has wrote specific computer program to calculat it by means of the disital integration. With the directly comut metheod the author has studied semi- Markov process appeared by A unite part in a detailed deposition section, the section presents deltaic sedimentary system effected by tide in a coal measures.A unite is oldest unite ofsystem.It separates B lithofacies constitutes (mudstone, divaricationg channel, tide way, tidal flat, distal bar, limestone, coal, tidal sand ridge) .Statistical analysis of 1 practice survey columns of A unite in this section shows state transition among the lithofacies constitutes having obvious Markov property and shows Γ-distribution is considerably perfect for fitting the stritum thicknees of lithofacies constitute. With the directly comput method the author calculates transition probability matrix or semi-Markov process among the lithofacies constitutes Q-mode factor analysis for the transition probabillity matrix obtains factors, the factors reflect effect factors of deposition environment.the preceding 3 factors effect main coal gather action of marginal deltaic basin, low deltaic plain, roof of deltaic plain below the water, distal bar building action, mutual action of river with tide. With facies unite of every column as sample, the author carries on step regression analysis of coal bed total thickness with deposition environment effect factors in the facies unite. F- testing and regression residual analysis show the regression epuation to be obvious statisticlly.This illustrates that espression of deposition environment control to coal thickness is in what follows: (1) To formation of thick-bedded coal fen plain of marginal deltaic is the fa vourablest; (2) Marginal deltaic basin deltaic plain below the water and low deltaic plain of riv er and tide development only form thin coal bed.
Keywords:
本文献已被 CNKI 维普 等数据库收录!
点击此处可从《沉积学报》浏览原始摘要信息
点击此处可从《沉积学报》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号