Neural networks for self-learning control systems

Derrick Nguyen, Bernard Widrow

International Journal of Control · 1991 · 92 citations · 8 references

Concepts

Abstract

ABSTRACT: Neural networks can be used to solve highly nonlinear control problems. This paper shows how a neural network can learn of its own accord to control a nonlinear dynamic system. An emulator, a multilay-ered neural network, learns to identify the system’s dynamic characteristics. The con-troller, another multilayered neural network, next learns to control the emulator. The self-trained controller is then used to control the actual dynamic system. The learning process continues as the emulator and controller im-prove and track the physical process. An ex-ample is given to illustrate these ideas. The “truck backer-upper,’ ’ a neural network controller steering a trailer truck while back-ing up to a loading dock, is demonstrated. The controller is able to guide the truck to the dock from almost any initial position. The technique explored here should be ap-plicable to a wide variety of nonlinear con-trol problems.

References

8