Research On De-Noising Method Of Ship Fire-Fighting Pipeline Leakage Signal

In recent years,pipeline leak location technology based on acoustic attenuation characteristics has been gradually developed,and good results have been achieved in scenarios with high signal-to-noise ratios.However,in the fire fighting pipeline system of ships,due to the influence of environmental noise,the signal-to-noise ratio of the signal is low,and it is difficult to accurately locate the leakage without de-noising.In order to solve the above problems,this paper builds a pipeline test platform to simulate the leakage conditions of ship fire pipelines,establishes a fire pipeline leakage location model based on the sound wave attenuation characteristics,and analyzes the statistical characteristics and frequency characteristics of leakage sound signals.Empirical Mode Decomposition(EMD)denoising algorithm for spectral analysis and multi-scale optimization denoising algorithm based on Fast Fourier Transform(FFT).This paper mainly completes the following research work:(1)A pipeline test platform was built.In order to simulate and collect the leakage sound signal of ship fire pipeline,a test platform is built according to the actual shape of the pipeline,and the corresponding hardware connection and software programming are completed.By analyzing the types of simulated noise and its frequency domain distribution characteristics,the leakage acoustic signals are divided into three working conditions:quiet,single noise and multi-noise.Acquire the leakage sound signal under simulated noise interference.After timefrequency analysis,the signals of the three working conditions are obviously different,which can be regarded as the analog signal of the fire fighting pipeline of the ship.(2)A leak localization model based on acoustic attenuation characteristics is established.Based on the characteristics of sound wave attenuation,a leak location model of ship fire fighting pipelines under quiet conditions is established,and the sound attenuation coefficient and sensor correction coefficient are obtained by signal fitting in quiet conditions.The quiet working condition signal is substituted into the positioning model to verify the accuracy of the positioning model.(3)An EMD denoising algorithm based on spectrum analysis is proposed.In order to remove the noise under the single noise condition,the empirical mode decomposition method is used to divide the noisy signal into different modes.By analyzing the statistical characteristics and frequency characteristics of the modes,the noise-dominated mode components are removed and the signal is reconstructed.Substitute the signals before and after denoising into the leak location model to obtain the location result curve and the location error curve.The curve analysis results show that the average location error of leak location decreases from 11.4%to 7.88%after processing by the denoising algorithm,indicating that the algorithm can effectively remove the noise under single noise conditions.(4)A multi-scale optimal denoising algorithm based on FFT is proposed.In order to remove the noise under multi-noise conditions,different analysis scales are used to divide the frequency domain signal into different frequency bands.If the proportion of the noise signal in a certain frequency band reaches a certain threshold,it will be replaced with the signal of this frequency band under the quiet condition.By analyzing the positioning results under different analysis scales,the optimal scale parameters under multi-noise conditions are obtained.Substitute the signals before and after denoising into the leak location model to obtain the location result curve and the location error curve.The curve analysis results show that the average error of leak location decreases from 21.9%to 9.8%after being processed by the optimal scale denoising algorithm.It shows that the algorithm can effectively remove the noise under multi-noise conditions.