Title :
Modeling of non-stationary process by modular separation of stability and plasticity
Author_Institution :
Lab. of Comput. Eng., Helsinki Univ. of Technol., Espoo, Finland
Abstract :
This paper presents a method for modeling a non-stationary process by a combination of fast learning and slowly learning modules, where the fast learning modules transform the input and output data for stable kernel module, which models a situation normalized to be stationary. The proposed method is applied to modeling a non-stationary chemical process
Keywords :
backpropagation; chemical industry; learning (artificial intelligence); multilayer perceptrons; process control; stability; backpropagation; chemical process; kernel model; learning modules; modular separation; multilayer perceptrons; nonstationary process modelling; stability; Adaptive control; Chemical processes; Data engineering; Intelligent robots; Kernel; Laboratories; Multilayer perceptrons; Programmable control; Radial basis function networks; Stability;
Conference_Titel :
Neural Networks Proceedings, 1998. IEEE World Congress on Computational Intelligence. The 1998 IEEE International Joint Conference on
Conference_Location :
Anchorage, AK
Print_ISBN :
0-7803-4859-1
DOI :
10.1109/IJCNN.1998.682262
