Improvement Of Artificial Bee Colonyand Its Application In Rayleigh Wave Inversion

Rayleigh wave exploration has the advantages of high resolution,strong antiinterference ability,low impact by the site,economical,rapidity and non-destructive.It has been widely used in many fields,and Rayleigh wave dispersion curve inversion is the core.Inversion of the Rayleigh wave dispersion curve can obtain the shallow surface layer structure information and the current inversion methods are mainly divided into local linearization inversion and nonlinear global optimization inversion.In this paper,a new global optimization algorithm,the artificial bee colony algorithm,is appropriately improved and used to invert the Rayleigh wave dispersion curve to obtain the subsurface shear wave velocity and the formation thickness.The traditional bee colony algorithm has strong local search ability and can find extreme points in the region.However,because there is no global optimal solution guiding behavior in the algorithm,and the global search strategy is relatively single,the algorithm may be trapped in local extremum,affecting the accuracy of the inversion.Referring to the optimal solution guidance strategy in the particle swarm optimization algorithm and the cross operation in the genetic algorithm,this paper makes an improvement on the bee colony algorithm and can effectively overcoming the above problems.This thesis focuses on the “improvement of artificial bee colony and its application in Rayleigh wave inversion”.Firstly,the basic principle and implementation process of the bee colony algorithm are introduced.Based on its defects,three improved strategies are proposed,and the optimal improved strategy is screened based on the efficiency of the operation and the characteristics of the dispersion curve inversion.A number of theoretical model trials are performed using improved algorithms to invert the basic dispersion curve.Compared with the inversion results of the traditional artificial bee colony algorithm,on the one hand,the validity and applicability of introducing the bee colony algorithm into the inversion of the fundamental wave dispersion curve are tested,on the other hand,the improved effect is also tested.Then,the paper conducts in-depth analysis and research.Using the improved bee colony algorithm to carry out noisecontaminated inversion,multi-mode dispersion joint inversion and frequency band missing inversion in different models,and test the improved algorithm from different directions.Finally,two typical examples of landfills in Italy and Wyoming in the United States of were inverted,and the practicality of the improved algorithm were tested.Main contributions of this thesis are as follows:1.The artificial bee colony algorithm was successfully applied to the Rayleigh wave dispersion curve inversion,and the effectiveness and feasibility of the algorithm in the inversion and the advantages compared with the current inversion algorithm were tested.2.The improved algorithm successfully draws on the particle swarm optimization algorithm and the genetic algorithm,and the bee colony algorithm is good at exploring and neglecting the defects of development.The flaws in the bee colony algorithm that are good at exploring and neglecting development have been effectively improved.3.Successfully performed multiple fundamental wave dispersion curve inversions based on the improved algorithm.Compared with the traditional bee colony algorithm inversion results,the applicability and efficiency of the improved algorithm inversion of the fundamental wave dispersion curve are tested.4.Successfully performed multiple noise-contaminated inversion,multi-mode dispersion joint inversion and frequency band missing inversion in different models,the feasibility of the improved algorithm to invert complex data is tested.5.Execution of a typical case study of the landfills in Italy and the U.S.Wyoming area verified the practicability of the improved algorithm.The original contributions in this thesis are as follows:1.The traditional bee colony algorithm lacks of the global optimal solution participation process and the global exploration strategy is relatively single.This paper introduces the global optimal solution guidance strategy in the particle swarm optimization algorithm and the crossover operation in the genetic algorithm to make targeted improvements,effectively avoiding the algorithm falling into the local extremum.2.Introducing the improved artificial bee colony algorithm into dispersion curve inversion.More than inversion of the fundamental wave dispersion curve is carried out,and the corresponding inversion studies are carried out in the multi-mode,noisecontaining and band-missing situations.The performance of the improved algorithm in Rayleigh wave inversion was systematically tested.