介绍用自回归滑动平均 (ARMA)模型形式表示复杂的电厂热工过程的方法 ,其采用了样本自协方差函数和自相关函数估计模型参数。这种方法可用于多输入多输出过程众多参数离线辨识 ,属灰色系统辨识方法的一种。此法对多输入的对象辨识准确可靠 ,稍加修改 ,即可用于在线辨识。 This paper introduces a method to represent the complex thermal power plant process in the form of autoregressive moving average (ARMA) model, which uses the sample autocorrelation function and the autocorrelation function to estimate the model parameters. This method can be used for many parameters of multi-input and multi-output process offline identification, which is a kind of gray system identification method. This method of multi-input object recognition accurate and reliable, a little modified, you can use for online identification.