Research On Structural Member Recognition Based On Non - Contact Acoustic Resonance Detection Technology

With the development of pattern recognition techniques, using the machine automatically complete recognition of the military weapons’equipment and medical drug has been more and more attention all over the world. Foreign scholars have carried out chemical weapons and ammunition classification, medical tablets quickly distinguish, nuclear material identification research work, but the domestic research field is less. According to the problem of structures’fast and accurate recognition, this paper develops a non-contact acoustic resonance detection technology. The main contents of this paper include the following aspects:First, this paper analyzes the development status of acoustic resonance detection technique about domestic and foreign. It uses acoustic resonance detection technology to identify different structures from their intrinsic frequency. According to the problem of the contact acoustic resonance cannot be applied to irregular structure identification, this paper the non-contact acoustic resonance detection technology based on acoustic excitation and laser vibration measurement. Based on the detailed analysis of acoustic resonance detection principle and laser vibration principle, this paper compares the advantages and disadvantages of non-contact acoustic resonance detection technology and the contact acoustic resonance detection technique.Second, this paper introduces the working principle of non-contact acoustic resonance detection, the type and function of the hardware equipment. It builds the experimental environment of non-contact acoustic resonance detection method. It develops the system software based on VC++to realize data transmission and storage. It uses MATLAB to develop the data processing software, including data mapping module, correlation analysis module, signal feature extraction module and pattern recognition module.Last, this paper carries out the non-contact acoustic resonance detection experiment for iron, copper, aluminum, tungsten four kinds of material structure. It analyzes and processes the experimental data to verify the feasibility of the technique is applied to the structure identification by drawing ratio and correlation. It uses finite element analysis software ANSYS to simulate and compute. There are obvious differences between natural frequency of Fe, Cu, Al three kinds of material for structural parts. The calculating results are consistent with the experimental results.