论文部分内容阅读
针对齿轮箱加速过程中振动信号的非平稳特性和传统阶比分析法存在的不足,笔者提出了一种阶比分析、自回归模型和支持向量机(SVM)相结合的故障诊断方法。利用计算阶比跟踪,将时域非平稳信号转化为角域准平稳信号,并对角域信号建立自回归模型(Auto Regresstive,AR),取得模型的参数向量组成及特征向量矩阵,以此作为输入建立支持向量机分类器,判断齿轮的故障类型。试验结果表明,该方法可有效提取故障特征,在变转速工况小样本条件下也能保证足够的诊断速度与精度。
Aiming at the non-stationary characteristics of vibration signals and the shortages of the traditional order analysis method, the author presents a fault diagnosis method combining order analysis, autoregressive model and support vector machine (SVM). By using calculation order tracking, the time-domain non-stationary signals are transformed into angular stationary signals and the autoregressive model (Auto Regresstive, AR) is established for the angular domain signals to obtain the parameter vector components and eigenvector matrix of the model. Enter the establishment of support vector machine classifier to determine the type of gear failure. The experimental results show that this method can effectively extract fault features and ensure sufficient speed and accuracy of diagnosis under small sample conditions of variable speed conditions.