粒子群优化算法在确定含水层参数中的应用  

The Application of Particle Swarm Optimization Algorithms to Estimation of Aquifer Parameters from Data of Pumping Tests

作　　者：郭建青[1] 李彦[2] 王洪胜[1] 周宏飞[2] 

机构地区：[1]长安大学环工学院,西安710054 [2]中国科学院新疆生态与地理研究所,乌鲁木齐830011

出　　处：《中国农村水利水电》2008年第4期4-7,共4页China Rural Water and Hydropower

基　　金：中国科学院知识创新工程重要方向性项目(KACX3-SW-326)

摘　　要：以泰斯公式为例,将粒子群优化算法应用于求解分析抽水试验数据,确定含水层参数的函数优化问题。就粒子数目和待估导水系数初值范围等因素对算法收敛性的影响,进行了数值实验。结果表明:①粒子群算法能够有效地应用于求解分析抽水试验数据,确定含水层参数计算问题;②粒子数目的大小对运算时间和算法收敛性有一定的影响,在实际运算时,宜设置较大的粒子数;③待估导水系数的选取初值大小对收敛速度有一定的影响,但不会影响最终计算结果。与其他方法相比较,粒子群优化算法具有原理简单、易于编程和实现等优点。With the application of particle swarm optimization algorithms （PSO）, the function optimization problem of analyzing pum ping test data in aquifer to estimate such parameters as transmissivity and storage coefficient was solved. With the different number of particles and the initial guessed values of transmissivity, the numerical experiments were conducted to explore the effect of these factors on the convergence of PSO algorithm. The results show that PSO algorithm may be effectively used to solve the function optimization problem of analyzing pumping test data to estimate transmissivity and storage coefficient, that the convergence of PSO algorithm and the computation time are influenced by the number of particles, and the fewer iterations are needed in computation with the larger number of particles and that the ranges of initial estimated values of transmissivity may also bring some effect on the convergence of PSO algorithm and the computation time, the larger the range is, the more iterations and longer computation time are needed to guarantee the convergence of PSO algorithm. Compared with other methods,PSO method is easy to understand and its procedure of computation is easy to realize.

关 键 词：粒子群优化算法 抽水试验数据 含水层参数 算法控制参数 算法收敛性 

分 类 号：TV211.12[水利工程—水文学及水资源]