基于多重预设广义频率响应函数模型的故障诊断方法

为了避免在基于非线性谱分析的故障诊断中估计被分析系统当前的广义频率响应函数（GFRF），提出了一种基于多重预设GFRF模型的故障诊断方法。该方法使用不同的预设GFRF模型来表征系统的各种工作模态，并根据各预设模型与系统当前状态的匹配程度来判定系统故障，应用这一方法对实际的车辆减震弹簧的故障进行了诊断，测试结果表明，该方法对于工程应用是可行的，且该方法能避免在线估计GFRF，具有在线计量小、便于用微处理器实现的优点。

Fault Diagnosis Method Based on Multiple Preset Generalized Frequency Response Functions Model

To avoid estmating current generalized frequency response functions (GFRF) of the analyzed system in the fault diagnosis based on nonlinear spectral analysis, a new diagnosis method based on the multiple preset GFRF model is presented. In this method, various preset GFRF models are used to describe the corresponding working mode of the system. And the matching extent of each preset model and the current state of the system can diagnose the faults. This method has been used to diagnose the fault of a actual vehicle damping spring. Test results indicate that this presented method is efficient for engineering application. Furthermore, this method can avoid estimating the current GFRF of the system. Its online computation requirement is small, and its realization by using microprocessors is convenient.