论文部分内容阅读
大地电磁粒子群算法反演属于非线性反演。在本文中,结合不同类型一维层状介质地电模型,介绍了大地电磁粒子群算法反演方法。在模型层数和参数增加、粒子维数增高时,针对标准粒子群算法处理高维问题容易陷入局部最优和收敛速度降低的缺陷,通过在搜索过程中引入交叉因子,增加了粒子多样性,克服了算法寻优过程中局部收敛的问题,提高了收敛速度。理论模型和实测数据试验结果表明,改进的粒子群算法在大地电磁反演中的寻优质量和效果较为理想。
Inverse inversion of MT electromagnetic particle swarm algorithm belongs to nonlinear inversion. In this paper, the geomagnetic particle swarm algorithm inversion method is introduced in combination with different types of one-dimensional layered media geoelectric model. When the number of model layers and parameters are increased and the dimension of particle is increased, the standard particle swarm optimization algorithm is easy to fall into the local optimum and the convergence speed is reduced. The crossover factor is introduced into the search process to increase the particle diversity. It overcomes the problem of local convergence in the algorithm optimization and improves the convergence speed. The theoretical model and experimental results show that the improved particle swarm optimization algorithm is superior to the optimal one in the magnetotelluric inversion.