粒子群算法在概率积分法沉陷预计模型参数反演中的应用

Application of particle swarm optimization in parameter inversion of probabilistic integral subsidence prediction model

针对开采沉陷预计模型参数反演所存在的算法复杂、计算量大等缺陷,将粒子群算法引入到概率积分法开采沉陷预计模型参数反演中。研究粒子群算法反演概率积分法预计模型参数的基本原理、编码方法及适应度函数的构造方法,同时结合河北省某煤矿的实测数据,以下沉拟合值与实测值的中误差作为反演精度的评价标准对算法进行实例验证,对提高开采沉陷预计的精度有一定的参考实用价值。

In this paper,the particle swarm optimization(PSO)algorithm is introduced into the parameter estimation of the mining model of the probabilistic integration method,which is used to estimate the probability integral method.The basic principle of the parameter,the coding method and the construction method of the fitness function are combined.At the same time,the algorithm is validated by using the measured value of the settlement value and the measured value of the coal mine in Hebei Province as the evaluation criterion of the inversion accuracy.Improving the accuracy of mining subsidence will have a certain reference value.