Title :
Neural network control of an unstable process
Author :
Feldkamp, L.A. ; Puskorius, G.V.
Author_Institution :
Res. Lab., Ford Motor Co., Dearborn, MI, USA
Abstract :
We explore the application of neural network methods to the development of controllers for unstable and highly nonlinear systems using the bioreactor benchmark problem as a concrete example. We demonstrate that the dynamic gradient method leads to controllers that are effective for the nominal plant and robust to reasonable changes of plant parameters
Keywords :
learning (artificial intelligence); nonlinear control systems; recurrent neural nets; stability; bioreactor benchmark problem; controllers; dynamic gradient method; highly nonlinear systems; neural network control; plant parameters; unstable process; Bioreactors; Concrete; Control systems; Gradient methods; Laboratories; Neural networks; Nonlinear systems; Robust control; Stability; Vehicle dynamics;
Conference_Titel :
Circuits and Systems, 1993., Proceedings of the 36th Midwest Symposium on
Conference_Location :
Detroit, MI
Print_ISBN :
0-7803-1760-2
DOI :
10.1109/MWSCAS.1993.343113
