Effective techniques for the identification and accommodation of disturbances

The successful control of dynamic systems such as space stations, or launch vehicles, requires a controller design methodology that acknowledges and addresses the disruptive effects caused by external and internal disturbances that inevitably act on such systems. These disturbances, technically defined as uncontrollable inputs, typically vary with time in an uncertain manner and usually cannot be directly measured in real time. A relatively new non-statistical technique for modeling, and (on-line) identification, of those complex uncertain disturbances that are not as erratic and capricious as random noise is described. This technique applies to multi-input cases and to many of the practical disturbances associated with the control of space stations, or launch vehicles. Then, a collection of smart controller design techniques that allow controlled dynamic systems, with possible multi-input controls, to accommodate (cope with) such disturbances with extraordinary effectiveness are associated. These new smart controllers are designed by non-statistical techniques and typically turn out to be unconventional forms of dynamic linear controllers (compensators) with constant coefficients. The simplicity and reliability of linear, constant coefficient controllers is well-known in the aerospace field.