Particle Swarm Inversion Of Rayleigh Wave Dispersion Curve Based On The Dispersion Function

The technology of Rayleigh wave detection had been more and more popularduring these years. Contrasting with the traditional detecting technologies, Rayleighwave detection has many advantages, for instance, easy control, high accuracy andtaking less time. Moreover, we don't need to break the place where we want todetect. So, Rayleigh wave detection has been paid more attention by researchers,and been used more and more widely. The inversion of Rayleigh wave is the core ofthe technology of Rayleigh wave detection. But there are some disadvantages thathinder the spread of Rayleigh wave detection in the traditional method of Rayleighwave inversion. There are two main disadvantages: first, the objective functionwhich is used in the traditional method increases the calculated amount when itsolve complex inversion problems; second, the dependence of the initial model istoo strong. In this paper, we improve the traditional inversion method, and the mainresearches and achievements are as follows:(1) Through the theoretical analysis, the reasons why the disadvantagesgenerated have been found. According to this, we use Rayleigh wave dispersionfunction as objective function and particle swarm optimization(PSO) algorithm tofind the globally optimal solution. Dispersion function is used to decrease thecalculated amount. PSO is almost not influenced by initial model.(2) A theoretical model has been built, then invert it by using multi-modes ofRayleigh wave dispersion curves with different methods. By contrasting the resultsof inversion, the improved inversion method is proved that it can solve the inversionproblem quite well. Then, by testing the anti-noise capacity of the improvedinversion method, we find that the improved method has very good anti-noisecapacity, and this also can testify that the improved method truly have advantages.(3) By inverting the data of numerical simulation and actual measurement, theresults of different method can be contrasted. Then it is found that a better fittin gdegree of dispersion curves is made by the improved method, and when deal withthe inversion of high-mode dispersion curves, the advantages are even more obvious.So it is testified that when dealing with the complex inversion problems, asignificant improvement is made by the improved method, so that it could be used ina wider scope, and let the technology of Rayleigh wave detection be more widelyspread.