摘要 为了进一步提高求取开采沉陷预计参数的精度,开展了基于变权PB组合预计模型的开采沉陷预计参数反演方法研究。系统地阐述了概率积分模型和Boltzmann函数预计模型预计原理,然后基于变权组合理论,构建了变权PB组合预计模型,该模型属于高度非线性模型,耦合烟花算法,提出了基于变权PB组合预计模型的开采沉陷预计参数反演方法。模拟试验结果表明,反演全部参数的平均相对误差约为4.0%,参数中误差约为3.9,拟合下沉和水平移动效果较好,构建方法能够精确可靠地反演开采沉陷预计参数。将本文构建方法、基于概率积分模型的求参方法和基于Boltzmann函数预计模型的求参方法应用到淮南顾桥矿某工作面的开采沉陷预计参数反演中,本文构建方法拟合程度远优于其他两种求参方法,3种方法拟合下沉和水平移动中误差分别约为82,107,100 mm,本文构建方法求参结果为:q=0.99,tanβ=1.90,b=0.42,θ=87.03°,S1=6.29 m,S2=-29.52 m,S3=49.78 m,S4=32.56 m。 In order to further improve the accuracy of the solved parameters,the inversion method for the prediction parameters of mining subsidence based on PB combination prediction model of variable weight was carried out.This paper first systematically elaborated the principle of the probability integral model and the Boltzmann function prediction model,and then the PB combination prediction model of variable weight was established based on the combination theory of variable weight.Taking into account the model belonging to a highly nonlinear model,the fireworks algorithm was introduced,and an inversion method for the prediction parameters of mining subsidence based on the PB combination prediction model of variable weight was proposed.The results of the simulation experiment show that the average relative error of all inversion parameters is about 4.0%,with 3.9 of the RMSE,and the effect of fitting subsidence and horizontal movement is better.The constructed method can accurately and reliably solve the prediction parameters of mining subsidence.The constructed method,the solved parameters method based on the probability integral model and the solved parameters method based on the Boltzmann function model have been applied to the inversion of the prediction parameters of mining subsidence in the working face of Huainan Guqiao Mine.The fitting degree of the constructed method is significantly better than that of the last two methods.The RMSE of the three methods in fitting subsidence and horizontal movement is about 82,107,and 100 mm,respectively.The result of the solved parameters by the constructed method is as follow:q=0.99,tanβ=1.90,b=0.42,θ=87.03°,S1=6.29 m,S2=-29.52 m,S3=49.78 m,S4=32.56 m。
机构地区 安徽科技学院资源与环境学院 安徽理工大学测绘学院
出处 《采矿与岩层控制工程学报》 北大核心 2021年第1期83-91,共9页 Journal of Mining and Strata Control Engineering
基金 安徽科技学院人才引进资助项目(ZHYJ201901) 安徽科技学院大学生创新创业资助项目(S201910879076)。
关键词 开采沉陷预计参数 概率积分模型 Boltzmann函数预计模型 组合预计模型 烟花算法 prediction parameters of mining subsidence probability integral model Boltzmann function prediction model combination prediction model fireworks algorithm
