MEMS Accelerometer Feasibility Study on Earthquake Measuring
|School||Southwest University of Science and Technology|
|Course||Safety Technology and Engineering|
|Keywords||earthquake measuring ADXL327 PIC STA/LTA wavelet transform polarization analysis|
This paper reviewed the application situation and development trend of seismometer technology, compared the traditional seismometers and MEMS digital acquisitions’main performance parameters, summarized their respective advantages and disadvantages. And then, comparative analysis the performance parameters of the different three-axis MEMS accelerometers,finally,we choice the high performance/price ratio ADXL327 accelerometer to manufacture 3-component MEMS acceleration sensors to detect earthquake shook (vibration).We use the low cost, low power consumption,multi-function PIC microcontroller as MEMS digital detector’s“heart”, it converse the two three components MEMS acceleration sensors analog signal into digital signal to eliminate traditional detectors analog signal long-term transmission interference; we use MCP6S26 programmable gain amplifier control gain and select input channel to reduce the tube of micro controller feet, cost and power consumption demand. To reduce the power consumption, we adopt STA/LTA, external key and level of changes triggered interruption to detect micro seismic shook (vibration).when MEMS digital detectors in no vibration, it enter into hibernation mode, these measures greatly reduce the power consumption of the MEMS digital acquisition.We use the MEMS accelerometer and PIC microcontroller as the core device develop digital detector, set earthquake shook (vibration) detection unit and data acquisition at an organic whole, realized the MEMS digital acquisition militarization, light-duty, low power consumption, low cost, wide band, big dynamic range to adapt to all kinds of mining roof and floor, mine pressure monitoring application needs, it will be strong to ascend micro seismic measuring mine geology hazard prediction ability.Accurately pick up the micro seismic events have important meaning for source orientation and explain focal mechanism in earthquake measuring. We contrast research three kinds of earthquake measuring P- arrived automatic loot algorithm, STA/LTA wins high micro seismic signal noise ratio effectively and impressive. We compared different short-time window length influence of STA/LTA accuracy, the result show that the greater short-time window length, the greater decrease of the sensitivity. Therefore, the trigger threshold should be set increase with the decrease of the short-time window length.Based on wavelet transform multi scale analysis thought, We use the wavelet transform coefficients instead of the original signals, applications the kinematic features of seismic phase information and polarization including in wavelet coefficients puts forward that united the wavelet transform and the polarization analysis automatic loot P-wave method. The results show that this algorithm pick up the low SNR micro seismic signal still satisfactory. But when sliding window to calculate covariance matrix and positioning function, computation is larger and cost time long.STA/LTA algorithm can’t or can’t be accurately automatic loot weak or low SNR micro seismic events, the multi-scale wavelet denoising micro seismic signals and reconstruct the micro seismic signals can improve the SNR, therefore, we propose that Wavelet de-noising combined classic STA/LTA pick up P wave. This method has realized full automatic loot micro seismic P-wave arrived and improved the loot efficiency and precision.