摘要
通过对航空发动机叶片损伤图像进行识别,可以快速准确地发现叶片损伤状况,有利于对故障进行及时有效的预测。本文对损伤图像进行分割,提取损伤图像特征参数,采用改进GA算法优化RBF网络参数的方法建立航空发动机叶片损伤图像识别模型,对损伤图像特征参数样本进行仿真实验,识别正确率为93.33%,同时与单一RBF网络模型识别结果进行对比分析,结果表明该方法更加优越有效。
Through the pattern recognition of aero-engine blades damage image, the blades damage condition can be fast and accurate detection and it is effective to predict malfunction problems. The damage image is segmen- ted and characteristic parameter is extracted. The adaptive genetic algorithm is used to optimize the basis neural network parameters, and dynamic adaptive GA-RBF recognition model is established. The method is applied in simulation. The classification accuracy rate is 93.33%. The method is compared with the radial basis neural net- work, the result show this method is more effective than the basis neural network.
出处
《科学技术与工程》
北大核心
2013年第28期8534-8538,共5页
Science Technology and Engineering
基金
航空科学基金(2008ZG54024)资助
关键词
图像分割
特征图像提取
GA算法
RBF网络
image segmentation characteristic image extraction genetic algorithm radial basis neu-ral networks
