Das dynamische Perzeptronmodell zur experimentellen Modellbildung nichtlinearer Prozesse

Zusammenfassung. Die Modellbildung technischer Prozesse ist eine zentrale Aufgabe der Regelungstechnik. Solche Prozeßmodelle werden für ein breites Anwendungsspektrum wie den Reglerentwurf, die modellbasierte Fehlerdiagnose und die Echtzeitsimulation benötigt. Der vorliegende Beitrag vermittelt im ersten Teil einen kurzen Überblick über die Anwendung neuronaler Netze zur Identifikation dynamischer nichtlinearer Systeme. Der zweite Abschnitt stellt das neue Modell eines dynamischen Perzeptrons vor und untersucht seine Fähigkeit zur Identifikation dynamischer Systeme. Im dritten Abschnitt wird eine Anwendung des dynamischen neuronalen Netzes zur Identifikation des Ladeprozesses in einem Dieselmotor mit Abgasturbolader beschrieben.Abstract. Modelling of technical processes represents a primary task within the field of control engineering. Such process models are required for a wide spectrum of applications as diverse as controller design, model-based fault diagnosis and real-time simulation. This contribution provides within the introductory section a brief overview of neural network types currently applied to the identification of dynamic nonlinear systems. Within the second section the so-called dynamic perceptron is presented and its system identification capability is investigated. The final section describes the application of the dynamic neural network to the identification of the charging process within a turbocharged diesel engine.

[1]  Rolf Isermann,et al.  Identifikation dynamischer Systeme , 1988 .

[2]  Giovanni Soda,et al.  Local Feedback Multilayered Networks , 1992, Neural Computation.

[3]  Rolf Isermann,et al.  Adaptive control systems , 1991 .

[4]  Martin Brown,et al.  Neurofuzzy adaptive modelling and control , 1994 .

[5]  M.M. Gupta,et al.  A Muti-Functional Dynamic Neural Processor for Control Applications , 1993, 1993 American Control Conference.

[6]  Lennart Ljung,et al.  Nonlinear Black Box Modeling in System Identification , 1995 .

[7]  I. J. Leontaritis,et al.  Input-output parametric models for non-linear systems Part II: stochastic non-linear systems , 1985 .

[8]  Mihiar Ayoubi.  Nonlinear system identification based on neural networks with locally distributed dynamics and application to technical processes , 1996 .

[9]  Jeffrey L. Elman,et al.  Finding Structure in Time , 1990, Cogn. Sci..

[10]  P. Frasconi,et al.  Local Feedback Multi-Layered Networks , 1992 .

[11]  Ronald J. Williams,et al.  A Learning Algorithm for Continually Running Fully Recurrent Neural Networks , 1989, Neural Computation.

[12]  Michael I. Jordan.  Serial Order: A Parallel Distributed Processing Approach , 1997 .

[13]  M. Thoma,et al.  Neurocontrol: Learning Control Systems Inspired by Neuronal Architectures and Human Problem Solving Strategies , 1992 .

[14]  Stephen A. Billings,et al.  Properties of neural networks with applications to modelling non-linear dynamical systems , 1992 .

[15]  Johan A. K. Suykens,et al.  Nonlinear system identification using neural networks , 1996 .

[16]  M. Ayoubi.  Nonlinear dynamic systems identification with dynamic neural networks for fault diagnosis in technical processes , 1994, Proceedings of IEEE International Conference on Systems, Man and Cybernetics.

[17]  Mihiar Ayoubi.  Nonlinear system identification based on neural networks with locally distributed dynamics and application to technical process , 1996 .

[18]  K.S. Narendra,et al.  Intelligent control using neural networks , 1992, IEEE Control Systems.