基于多类AdaBoost的液压系统故障诊断算法
作者：胡浩 李琛
来源：《中国科技纵横》2012年第04期

        摘要：主要研究了多类分类AdaBoost算法，及其在液压系统故障诊断中的应用。为了解决“一对一”算法和“一对余”算法的分类速度随着训练样本数或类别数的增多而变慢的问题，提出了基于决策树的AdaBoost算法。利用CART算法构造决策树建立AdaBoost分类器，并根据样本数据的分布情况，使得在决策树中每一个节点的最可分类别尽可能分开。通过对某型自行火炮液压系统故障进行分析，表明该算法的性能优于其他两个算法，具有更高的通用性，验证了该算法的有效性。

        关键词：决策树 AdaBoost 故障诊断 自行火炮 液压系统

        Multi-class AdaBoost Method for Hydraulic System Fault Diagnosis

        Abstract：Considering the limitations of the traditional methods, decision tree AdaBoost, which combines AdaBoost and decision tree, is proposed for multi-class classification. Decision tree is constructed by using CART algorithm, and AdaBoost classifiers are established. Based on the distribution of samples, the most separable classes could be separated at each node of decision tree. Compared with “one-versus-one” and “one-versus-rest”, experiments are conducted on hydraulic system of some self-propelled artillery to testify effectiveness. The results show that the proposed method has better performance and higher generalization ability than other two methods.

基于多类AdaBoost的液压系统故障诊断算法