基于加权PCR证据融合的多传感器Lamb波损伤成像

为提高基于Lamb波损伤定位成像的对比度和可靠性,提出了一种基于证据理论多传感器信息融合的损伤成像方法。基于连续小波变换,确定了传感路径的时间延迟,进而采用椭圆定位法进行损伤成像。以信号时频带能量变化率为指标,确定了传感路径的权重值,提出了一种基于加权比例冲突再分配(Proportional conflict redistribution,PCR)证据融合算法,并以此对多组传感路径的定位成像进行融合。试验结果表明,该方法具有较好的成像对比度,且能降低因部分传感路径时间延迟判断失误对成像结果造成的影响。

Damage Imaging by Lamb Waves Based on Weighted PCR Evidence Fusion for Multi-sensor

The damage imaging method based on evidence theory data fusion is proposed in order to improve the contrast ratio and reliability of damage detection image. Time delay of sensing path is determined by continuous wavelet transform. Damage location image is carried out by using the ellipse location method.The weight value of the sensing path is determined by the frequency band energy rate of the signal. A weighted proportional conflict redistribution (PCR) evidence fusion algorithm is proposed, which is used to fuse the multi sensor path. The test results show that the algorithm has better image contrast ratio, and can reduce the influence of imaging result by the incorrect time delay of some sensing paths.