Ultrasonic Echo Parameter Study Estimated Ant Colony Algorithm

Ultrasonic detection plays an important role in the field of industrial nondestructive test and biomedical engineering. Ultrasonic detects the target through the transducer emitting acoustic pulse to the target and receiving the target reflection echo. The main purpose is to determine the physical characteristics of the reflecting target, such as geometric shape, size, etc. all kinds of information on the transmission path. If we want to obtain the practical information from the ultrasonic echoes, we must correctly understand and analyze the ultrasonic echo of target. So the accurate estimation of parameters vector of target’s echo is needed. The use of modeling method to estimate parameters of ultrasonic echoes is an effective method.In this paper, the ultrasonic echo is modeled as a Gaussian echo corrupted by Gaussian white noise. In the model, every Gaussian echo is a nonlinear function of a set of echo parameters:bandwidth factor, arrival time, center frequency, phase and amplitude, those are Gaussian echo model and asymmetric Gaussian chirplet model. According to the nonlinear model, the high accuracy parameters’estimation of ultrasonic echoes is acquired through numerical methods.The research of this paper is following in detail:1. The ultrasonic echoes are modeled as the Gaussian echo which corrupted by white Gaussian noise. Apply each Gaussian echo simulate a single ultrasonic echo and superpose the multiple Gaussian echo to simulate multiple ultrasonic echoes.The nonlinear model of ultrasonic echoes are setup, they are Gaussian echo model and asymmetric Gaussian chirplet model.2. The Gauss-Newton algorithm and ant colony algorithm are used to estimate the parameter vector of the single ultrasonic echo which is simulated by using Gaussian echo model and asymmetric Gaussian chirplet model.3. The affections of the relevant parameters in ant colony algorithm to the estimation performance of ultrasonic echoes’parameters are researched. In ant colony algorithm, its own parameters have a great impact on the performance of the algorithm. But there are no uniform and explicit set rules so far, therefore it must optimize the parameters according to the performance of algorithm in different applications. In this paper, in order to get the optima estimation performance of ultrasonic echo, we optimize the parameters of ant colony algorithm according to the stability, the consumption machine and the estimated errors.4. In order to improve the flaw of EM algorithm when it is used in the parameters estimation of multi-ultrasonic echoes, hybrid algorithms by optimizing the EM algorithm and ant colony algorithm are proposed to improve the accuracy of the parameters estimation of multi-ultrasonic echoes.5. The tank experiments verify the effectiveness of proposed algorithms. In the first place, we get the actual ultrasonic echoes by pool experiments, and then we estimate the parameters of actual ultrasonic using proposed algorithm. To verify the proposed algorithms are effective in the application of parameters estimation of ultrasonic echoes, the actual parameters’ values and the estimated values were compared carefully.The results of this research have certain significance to ultrasonic nondestructive testing technology in the application of in the field of industry and biomedical engmeering.