摘要
对大型圆钢锻件常见缺陷进行了超声识别技术研究, 利用感知器算法实现了典型缺陷自动识别, 并根据不同类型、位置和姿态的超声换能器, 对不同缺陷类型的敏感程度不同, 实现了三组换能器在决策级上的数据融合, 大大提高了缺陷识别可靠度。
The common flaws recognition in heavy ingots were researched using ultrasonic method.Flaw types auto-recognition was realized through perceptron algorithm and analysis.The result shows that the recognition results are improved by using data fusion technics in decision-order in which three groups different type,position and stance of transducers are used.
出处
《机床与液压》
北大核心
2005年第2期121-122,共2页
Machine Tool & Hydraulics
关键词
数据融合
缺陷识刖
超声
锻件
Data fusin
Flaw recognition
Ultrasonic
Heavy ingots
