Neurocontrol and Elastic Fuzzy Logic: Capabilities, Concepts,

In recent years, enormous progress has been made in neurocontrol-the use of neural nets as controllers. Designs from neurocontrol can also be used with a wide variety of nonneural system. For example, this paper will show how elastic fuzzy logic (EFL) nets make it possible to combine the capabili- ties of expert systems with the learning capabilities of neural nets at a high level. Still, ANN implementations have advantages in terms of hardware implementation, ease of use, generality, and links to the brain, which is still the only true intelligent controller available to us. Neurocontrol is useful in cloning experts, tracking trajectories or setpoints, and in optimization (e.g, approximate dynamic programming). There has been sub- stantial success in controlling robot arms (including the main arm of the space shuttle), chemical process control, continuous production of high-quality parts, and other aerospace applica- tions. This paper is a tutorial or roadmap of the basic designs and concepts, with reference both to applications and future research opportunities.

[1]  Paul J. Werbos,et al.  Applications of advances in nonlinear sensitivity analysis , 1982 .

[2]  Y. Le Cun,et al.  Comparing different neural network architectures for classifying handwritten digits , 1989, International 1989 Joint Conference on Neural Networks.

[3]  Y. Dote.  Fuzzy and neural network controller , 1990, [Proceedings] IECON '90: 16th Annual Conference of IEEE Industrial Electronics Society.

[4]  P. J. Werbos,et al.  Neurocontrol and neurobiology: new developments and connections , 1992, [Proceedings 1992] IJCNN International Joint Conference on Neural Networks.

[5]  Paul J. Werbos,et al.  Neurocontrol and fuzzy logic: Connections and designs , 1992, Int. J. Approx. Reason..

[6]  Eduardo D. Sontag.  Feedback Stabilization Using Two-Hidden-Layer Nets , 1991 .

[7]  P. Werbos,et al.  Beyond Regression : "New Tools for Prediction and Analysis in the Behavioral Sciences , 1974 .

[8]  Mark A. Atkins,et al.  Neural joint control for Space Shuttle Remote Manipulator System , 1992 .

[9]  Paul J. Werbos,et al.  Maximizing long-term gas industry profits in two minutes in Lotus using neural network methods , 1989, IEEE Trans. Syst. Man Cybern..