用于土壤分层电阻率模型反演的人工蜂群结合混沌搜索算子及混沌算法

Artificial Colony Algorithm Combined with Chaotic Search Operator and Chaotic Pool Used for Resistivity Model Inversion of Layered Soil

土壤结构参数的测量与反演是接地设计中的重要环节。为此基于求解土壤分层模型电位分布的高阶复镜像法以及土壤模型结构反演的原理,开发了基于直接搜索法的土壤分层结构反演算法。进而在人工蜂群算法的基础上,加入混沌搜索算子和混沌池,提出了1种用于土壤分层结构反演的混沌池人工蜂群算法,增强了反演算法的局部搜索能力和全局搜索能力。计算表明,较用于反演土壤分层结构的遗传算法,提出的算法拥有更小的反演误差,最大可以将反演的均方根误差缩小到遗传算法的1/10。通过对比几种基于直接搜索的算法,证明了该算法在收敛速度以及寻优能力上的优势,比遗传算法更加适合作为土壤结构反演的算法。

Measurement and inversion of soil structure parameters are important parts of grounding design. Using a high-order complex image method for solving the potential distribution of layer soil models and the inversion principle of soil model structure, we discussed the direct search method for soil layered structure inversion. Then we proposed a chaotic pool artificial bee colony algorithm for soil structure parameter inversion, in which the chaotic search operator and chaotic pool were combined with the artificial colony algorithm to enhance the inversion algorithm＇s ability in global search and local search. Simulative calculations show that, compared with the genetic algorithm, the proposed algorithm in soil structure inversion has smaller inversion error, specifically the root mean square error is reduced by 9/10. Further comparison between the proposed algorithm and some common algorithms based on direct research indicates that the proposed method is advantageous in both convergence speed and optimal capacity, and it suits soil structure inversion better than genetic algorithms.