The fluctuation and uncertainty of acoustic measurement in shallow water wave-guide

摘要

Acoustic measurement signals in shallow water are often randomly influenced by environmental factors, such as source-receiver geometry swing, random medium conditions as internal wave and turbulent current, the disturbance of ocean environment noise etc. The signal sound intensity and the phase fluctuations affected the uncertainty of the measured result, which will affect negatively the validity of the data and degree of confidence. At the same time, the multi-path channel in the ocean (not clear) lead into signal distortion and interference, which caused the departure from the free field measurement results, and diminish the correlation for underwater acoustic sensors. Quantitative Analysis for different factors producing the uncertainty would enhance the degree of confidence about optimized experiment schemes. The monitoring parameters for underwater environments in preceding and processing of the experiment can evaluate and calibrate the uncertainty of the metrical data. This paper? uses the normal mode model, statistical theory such as Monte Carlo method, and the MATLAB tools, to analyze the underwater acoustic uncertainty, which can be applied for precise measurement in shallow water. The actual prior data is used in the analysis, involving temperature chain and sound speed profile, the simulations showed that source-receiver depth, subbottom and surface parameters, random internal waves, and tides variously influenced the measurement results for different frequency the probability density function (PDF) and deviation for sound propagation transmit-loss are specified as a result of the quantification analysis, which caused by environment statistical distribution. The various amplitude of ocean environment noise is spliced on the temporal target signals, on the purpose of simulating the infection by different signal noise ratio (SNR) conditions, and the result showed that deviation between actual and polluted signal is less than 1 dB per 1/3 octave bandwidth, while SNR greater than 6 dB; this conclusion can be used for evaluating the vessel noise testing data. An acoustic propagation experiment data is used to validate these fluctuations and uncertainties. The degree of the confidence for measurement result can be evaluated and optimized by using this analysis method.