摘要
构建基于分位算法的极限参数模型,在样本数据逐渐积累的情况下,通过调整修正参数,对汽车变速器进行迅速、可靠的故障判别。而基于标准差的故障诊断方法对样本数据的规模及统计分布具有严格的约束条件。两种方法实验比较结果证明了基于分位算法的故障诊断方法具有更强的自适应能力和更高的泛化能力。
The limit parameter model based on fractile algorithm is built first.When the sampling data is accumulated,the faults of the automotive gearbox can be diagnosed quickly by adjusting the parameter.No strict constraint conditions for the data scale and distribution should be met for the method while the one based on the standard deviation needs the constraints.The comparison test results show that the method based on fractile algorithm is more adjustable and more generalized.
出处
《长春工业大学学报》
CAS
2012年第1期11-15,共5页
Journal of Changchun University of Technology
基金
国家自然科学基金资助项目(60904047)
关键词
分位算法
故障诊断
极限参数
标准差
fractile algorithm
fault diagnosis
limiting parameter
standard deviation.
