基于遗传算法的概率积分法预计参数反演  

Research on Parameters Inversion in Probability Integral Method by Genetic Algorithm

作　　者：查剑锋[1,2] 冯文凯[2] 朱晓峻[1] 

机构地区：[1]中国矿业大学国土环境与灾害监测国家测绘局重点实验室,江苏省资源环境信息工程重点实验室,江苏徐州221116 [2]成都理工大学地质灾害防治与地质环境保护国家重点实验室,四川成都610059

出　　处：《采矿与安全工程学报》2011年第4期655-659,共5页Journal of Mining & Safety Engineering

基　　金：国家自然科学基金重点项目(50834004);国土环境与灾害监测国家测绘局重点实验室开放基金项目(LEDM2009B01);地质灾害防治与地质环境保护国家重点实验室开放基金项目(SKLGP2010K002);中国矿业大学青年基金项目(09091252);国家自然科学青年基金项目(41104011)

摘　　要：地表移动观测站实测数据反演求取概率积分法参数是这一方法应用过程中的关键问题,现有算法存在求参易发散,难以获取最优解的缺陷.提出采用遗传算法反演概率积分法参数,研究了该算法反演结果的准确性和可靠性.研究结果表明,遗传算法反演概率积分法预计参数精度高,参数相对误差小于1.5%,对观测站中的观测值随机误差、粗差、观测点缺失的问题具有较强的抗干扰能力,较最小二乘法、模矢法在参数反演准确性和可靠性方面有明显优越性.Parameters inversion of probability integral method using surface movement observation data is one of the key issues in its application. The existing algorithms have some defects, such as divergence in parameter calculation, difficulty in optimum acquisition, et al. In this paper, we propose to use genetic algorithm to inverse the parameters of probability integral method, and have studied the accuracy and reliability of this algorithm. The results show that parameters of probability integral method can be inversed precisely by genetic algorithm, and the relative errors of inversion parameters are less than 1.5%. Moreover, this method has strong anti-interference ability to the influences of random errors in observed values, gross errors, losses of observation points, et al. The genetic algorithm has significant advantages in the accuracy and reliability of parameters inversion, compared to the methods of least square and pattern search.

关 键 词：概率积分法 遗传算法 参数反演 模矢法 

分 类 号：TD17[矿业工程—矿山地质测量]