The Research Of Vibration Source Localization Algorithm Based On Acceleration Sensors

The vibration source localization is a technology that ascertain target position by usingvibration wave information, which is an important method to ensure the national economicconstruction and people’s life and property security. Due to the complexity of mediumstructure characteristics, nonuniformity and other factors, make the localization accuracy ofthe traditional method is low and can’t meet the requirements of engineering practice.According to the problem, this dissertation proposes the vibration source relative localizationmethod based on time-frequency analysis and processing of the vibration signal and optimizedNewton algorithm. This method can be applied to explosion bombed position localization,blasting vibration, the petroleum exploration,coal mine and the pipeline safety monitoringengineering field and so on, and it has the important research significance and practical value.The dissertation proposes the vibration source localization system scheme design basedon the research development current situation of vibration signal analysis and processing andvibration source localization methods, meanwhile introduces the index and function of thedata acquisition module and vibration source localization module, using the MEMS three axisacceleration sensors to collect vibration signal. According to the instantaneous non-stationarycharacteristics of the vibration waves, the HHT method is used to analysis signalcharacteristics and the wavelet threshold value method is used to reduce the noise. Thevibration wave first arrival time is picked up to get the time delay between base stations byusing power ratio method according to the pretreatment vibration signal. The vibration sourcelocalization devides into two steps—primary localization and emendatory localization the vibration source parameters, locating the vibration source by using the iterative algorithm andthe optimized Newton algorithm respectively, the algorithms are optimizated while thevibration source is localized accurately. Lastly the error sources that influenced thelocalization accuracy and the localization accuracy evaluation index are also researched in thedissertation.Simulation and experiment results show that the relative localization method can realizevibration source localization. The vibration source localization accuracy can be improvedeffectively via vibration signal pretreatment and the optimized Newton localization algorithm,localization results meet the requirements of the engineering practice basically.