Plant identification in the presence of disturbances and application to digital adaptive systems

THE NEED for an adaptive control system occurs when the parameters involved in the design of the system change in time. For small variations, conventional feedback configurations can sometimes reduce the sensitivity of the output to these changes to acceptable values, but large variations of the parameters require a more flexible system. The newer adaptive point of view in designing a control system can, in general, allow for changes in input signal characteristics, plant characteristics, and disturbance characteristics with a minimum amount of a priori knowledge. It appears that a general distinction has been made between adaptive controls1,2 and optimalizing (optimizing) controls3,4 although the characteristics are not always distinct. A recent note5 has suggested a possible classification according to whether there is an original feedback loop or not.