基于D-S证据理论和RBF网络的航空发动机叶片损伤图像识别技术研究

针对航空发动机叶片损伤图像采集过程中存在的不确定性因素及单一RBF网络或D-S证据理论在叶片损伤图像识别中存在的不足,提出一种基于D-S证据理论和RBF网络相融合的决策级信息融合损伤图像识别算法。首先,用RBF网络对损伤图像进行初步识别;然后,将RBF网络识别输出结果作为D-S证据理论的基本可信度分配;最后,利用D-S联合规则进行合成,得出最终识别结果。通过单一优化RBF网络的图像识别结果和融合识别结果的对比分析,证明了该方法在航空发动机叶片损伤图像识别方面的优越性。

Research on Aero-engine Blades Damage Image Recognition Based on D-S Evidential Theory and RBF Network

for the uncertainty of aero-engine blades damage image in acquisition process and the shortage of single method of RBF network or D-S evidential theory in blades damage image recognition, a damage image recognition method based on decision-level information fusion was proposed, which was based on D-S evidential theory and RBF network. The RBF network was applied to the blades damage image recognition firstly, and then, the partial recognition results of RBF network was taken as the basic probability assignment, finally, the D-S evidence theory was applied to fuse different results from all the RBF network and got the finally results. The superiority of this method in aero-engine blades damage image recognition was proved by comparison of the single optimize RBF network image recognition with fusion recognition results.