Analytical Study On Real-time Algorithm Of Sound /Vibration Signal In Embedded System

This dissertation describes the research on embedded sound/vibration signal processing algorithms, and the software design of noise/vibration signal analysis instrument. Firstly, studies on and designs the software of bottom layer based on the overall framwork, which consists of the drivers of embedded system and the drivers of DSP; secondly, studies on the algorithms of sound/vibration signal processing and designs the software of these algorithms. By much tests, the performance of the instrument can meet the specifications.At first, this dissertation introduces the target and significance of this task, and describes of the sound/vibration signal detecting and analysis technology's developing trends. The DSP (Digital Signal processing) algorithms are analyzed, and these sound/vibration signal analysis is used in embedded system. Next, describes the work of the author have done in this task. After the overall schem of system software and hardware, the following chapters describe the drivers programming, bottom layer programming of DSP, the algorithms of signal processing, and present the design process of sound/vibration signal analysis instrument.Theoretically, first, analyzes and discusses the noise/vibration signal processing algorithms, for example: frequency-weight, time-weight, FFT (Fast Fourier Transform), octave analysis and one-third octave analysis, and the complexity of these algorithms. To reduce the complexity of octave analysis and one-third octave analysis, the multirate filters analyse method is proposed. This method can reduces much complexity. The dissertation also studied the effective method for designing narrow transition-band FIR (Finite Impulse Response) filter, and proposes a new method for the design of the narrow transition-band FIR filter. This method can reduce much complexity in the same specification and improve the performance of the system.The highlight of the hardware system is the dual-core architecture, which profits from the DSP'effective processing on signal and the plentifule interfaces of ARM. The whole system, realizes all function of the sound/vibration signal deceting using software. And the precision of the system can be adjusted by software conveniently. The GUI (Graphic User Interface) is based on Qtopia, which increases comfort to users.