Input analysis, process data steps

The steps in developing such a database are (1) collection of machine and process data and (2) database setup. Input requirements of the software are machine and process specifications, analysis parameters, data filters, alert/alarm limits, and a variety of other parameters used to automate the data-acquisition process. 

As discussed and illustrated in the introduction, data analysis can be conveniently viewed in terms of two categories of numeric-numeric manipulation, input and input-output, both of which transform numeric data into more valuable forms of numeric data. Input manipulations map from input data without knowledge of the output variables, generally to transform the input data to a more convenient representation that has unnecessary information removed while retaining the essential information. As presented in Section IV, input-output manipulations relate input variables to numeric output variables for the purpose of predictive modeling and may include an implicit or explicit input transformation step for reducing input dimensionality. When applied to data interpretation, the primary emphasis of input and input-output manipulation is on feature extraction, driving extracted features from the process data toward useful numeric information on plant behaviors. 

The above analysis shows that, as long as the disturbances are fast relative to the process dynamics, an accurate model can in fact be constructed from step response data. However, many process 2ire affected by slow, drifting disturbances that effectively mask the true process response. For these types of disturbances, the proposed Ls erre approach may produce process models with significant errors. In this case, other types of input signals, such as a random binary input signal or a periodic input signal, should be used to enable the effect of the disturbances on the process output to be separated from the process response due to the input variable. 

---