High-power industrial rolling mills rely heavily on the sustained operation of cycloconverters, a type of variable-frequency drive. This research proposes a methodology, which observes and diagnoses the operation of cycloconverters as either normal or abnormal by use of time-frequency signature analysis. Various features of the cycloconverter's input current in the time-frequency domain are identified and used to derive parameters that describe these two states in a quantitative manner. A reference model using the parameters is then developed, and comparisons in the time-frequency domain to real data are implemented. Based on these comparisons, a statistical decision boundary is delineated that is used to classify the health status of the cycloconverter.

中文翻译：

---

通过时间__频率特征通过谐波电流进行轧机周变流器状态评估


高功率工业轧机严重依赖循环变流器（一种变频驱动器）的持续运行。这项研究提出了一种方法，通过时频特征分析来观察和诊断循环变流器的运行正常或异常。循环变换器输入电流在时频域中的各种特征被识别并用于导出以定量方式描述这两种状态的参数。然后开发使用这些参数的参考模型，并在时频域中与实际数据进行比较。基于这些比较，描绘出统计决策边界，用于对循环变流器的健康状态进行分类。