Study On The Research Of Ultrasonic Based Wavelet

As a typical noncontact measurement, the ultrasonic positioning technology has many merits, such as simple structure, low cost, high accuracy of the whole positioning and so on, therefore, it is a wide spread applications in recent years especially in the fields of Vision Substitution or the ETAs (Electronic Travel Aid system). Be based upon the former ultrasonic ETAs architecture, which has been finished by our lab, this paper's purpose is to find a new way to overcome the common disadvantages in nowaday's ETAs, since all of them which be based on the ultrasonic can not recognize the all-terrain obstacles. The main points of this thesis are as follows:1. First, this paper summarizes the development and foundational principle of ultrasonic detection. Then it presents the working mechanization and characters of ultrasonic sensor. At the same time, it discusses a number of main technical parameters which affect the accuracy of the measurement. Moreover, after introducing the fundamental functions of the system, and be based on the former existing architecture, it proposes the whole structure of the experimental data acquisition system.2. Starting from the properties of the Mutant signal and the non-stationary signal, after researching the time-scale/frequency analysis method, this paper presents the detailed analysis of the signal processing. And these studies laid a theoretical foundation of the farther experimental analysis.3. Then the design principle and the structure of the experimental data acquisition system, which consists the microcomputer controll circuit, the transmit circuit of ultrasonic, the prepositive proportional amplifier circuit, TGC circuit, the filter circuit and so on, is detailedly described in this paper. The microcomputer program is compiled with the C language. The paper has also designed the circuit board of the system, collected the experimental data and verified the arithmetic.4. According to the wavelet theory, a new wavelet-based multi-resolution analysis of the echo method of treatment has been proposed. Through the extraction of echo information on in different layers, we can set up many kinds of models of the obstacles to realize the all-terrain obstacles recognation. Focuses on theories of multi-resolution wavelet analysis and mallat algorithm, this paper derived the characteristics of multi-resolution analysis from the research and equivalent it to the conjugate quadrature filter bank. Subsequently, after studying the multi-resolution signal decomposition and re-conformation, this paper has analyzed the ultrasonic eoch signal by the appropriate wavelet basis and verified the accuracy of algorithm.