摘要
针对单一的智能模型在发动机磨损模式识别中的局限性,提出了一种基于数据融合技术的多模型磨损模式识别方法。它利用模糊优选模型、神经网络模型和灰色关联度模型等3种单一智能模型的识别结果作为信息源,经D-S证据理论对其进行融合得到最终识别结果。实际计算表明,该模型具有良好的通用性、适应性和容错性,比单一的智能模型具有更好的识别效果。
In order to overcome the disadvantages of single model for wear pattern recognition of engine, a multi-intelligent model based on the data fusion technology was proposed. The results of three single model ,fuzzy optimum model, neural network model and grey correlation degree model were used as the different data sources, an improved result was obtained by combination them used the effective data fusion technique D-S evidence theory. According to the application in the wear pattern recognition, the new method can get a more precise result in comparison with the single method.
出处
《润滑与密封》
CAS
CSCD
北大核心
2007年第6期60-63,共4页
Lubrication Engineering
关键词
数据融合技术
D-S证据理论
内燃机
磨损模式识别
data fusion technology
D-S evidence theory
internal combustion engines
wear pattern recognition
