模糊神经网络和多传感器数据融合的系统设计——海底管道腐蚀监测系统

System Design Based on Fuzzy Neural Network Multi-sensor Data Fusion——Submarine Pipeline Corrosion Monitoring System

系统利用三组超声传感器和漏磁传感器获取原始信息,对多路原始信息进行小波变换和频谱分析,提取时域、频域的多种不同属性参数,并从中选择与输油管道腐蚀程度相关度大的特征参数。采用模糊神经网络多传感器数据融合方法,监测海底输油管道的腐蚀程度,可以在很大程度上提高对输油管道腐蚀程度的检测精度。实验结果表明了该方法的有效性和可行性。

System uses three ultrasonic sensors and flux leakage sensors to obtain the original information,the multiple original information under wavelet transform and spectrum analysis to extract a variety of parameters with different attributes in time domain and frequency domain,and to select the major characteristic parameters which have large degree correlation with oil pipeline corrosion.Using the fuzzy neural network multi-sensor data fusion method to monitor undersea pipeline corrosion,submarine pipeline corrosion inspection system can largely improve the detection accuracy of pipeline corrosion.Experimental results show the effectiveness and feasibility of the method.