Font Size: a A A

Research On Detection And Estimation Of Signals Interfered By Heavy Noises

Posted on:2007-12-19Degree:MasterType:Thesis
Country:ChinaCandidate:Y CengFull Text:PDF
GTID:2178360182470882Subject:Electronic Science and Technology
Abstract/Summary:PDF Full Text Request
The content of this thesis is in two aspects: theoretically, use the method of time-frequency analysis to research the problem on detection and pick-up of sound and vibration signals interfered by heavy noises. On the other aspect, develop a portable instrument based on the technique of Digital Signal Processing (DSP) and Advanced RISC Machines (ARM) to detect and analyze sound and vibration signals.First, Discuss the advantage and disadvantage of some methods of time-frequency analysis usually used in the field of signal processing on sound and vibration. Based on this, aim at the disadvantage on Wigner Distribution (WD) and Fractional Fourier Transform (FrFT) that can't restrain noise interference effectively when be used to detect Linear Frequency Modulated (LFM) signals interfered by heavy noises, we proposed a improved WD. Simulation results show that this method can restrain noise interference effectively and improve the ability of LFM signals detection. On the other hand, in multi-component cases, WD suffers from the cross-terms, which may impair some of the auto-terms. In order to separate original multi-component signals into components and reconstruct them separately, we can use the method of optimal filtering in fractional Fourier domain. But this method also has a problem that if signals have aliasing in the fractional Fourier optimal domain as well, the reconstructed signals will be distorted. In this thesis, a novel method of quasi Fourier transform is proposed in order to solve the problem of multi-component signal reconstruction, and have good results.Based on the research of signal processing on sound and vibration theoretically, we designed a portable instrument based on the technique of DSP and ARM to detect and analyze sound and vibration signals. This thesis focuses on designing the module of data gathering and processing, and the circuit design of the whole system. We have finished scheme design of data gathering and processing module, draw circuit schematic map, PCB design and hardware debugging. Recently, hardware works well, can realize the function of data gathering, storing and processing.
Keywords/Search Tags:Sound and Vibration Signals, Improved Wigner Distribution, Quasi Fourier transform, TMS320VC5502, Digital Signal Processing
PDF Full Text Request
Related items