State of the art and survey of learning control applications

Adaptive control systems which improve their own performance, using information contained in their past input, output and environmental feedback signals, undergo a learning process. Consequently, a learning or self-organizing control system is defined as one which demonstrates both adaptive and memory capabilities. This paper attempts to describe the present position (as of August 1972) of learning control applications, and to give a brief account of the types of control systems that have arisen through the use of such techniques.