Acquirement of the Basic Value of Heat Deviation Parameters Based on the Generalized Regression Neural Network
Abstract
Acquirement of the basic value of heat deviation parameters is the key to application of heat deviation analysis. This paper put forward a method of acquirement of the basic value of heat deviation parameters based on generalized regression neural network (GRNN), which cannot fully understand the conditions of process mechanism, through the study of process data and modeling, excavating implicit relationship between the process parameters and approaching the process mechanism. Finally, take the main steam pressure for instance, the results show that the maximum predicting error of GRNN is 0.06%, and that of BPNN is 1.91%. It is visible that GRNN has a better predictive ability than BPNN, even if lack sample data, it indicates that GRNN prediction model is suitable for small samples. Besides, the design of GRNN model is simple and it has a fast convergence which improves the real-time processing and the prediction ability of reflecting the latest operation condition parameters.
Keywords
Heat Deviation Analysis, Generalized Regression Neural Network.
DOI
10.12783/dtetr/icmca2017/12336
10.12783/dtetr/icmca2017/12336
Refbacks
- There are currently no refbacks.