Search

This paper builds on previous work which developed a robust control algorithm which used statistical confidence bounds. The key idea was to change the nominal design so as to reduce the overall variability from an a priori specified desired performance. However, due to the choice of cost function, the original design method did not necessarily guarantee stability. This paper addresses the latter problem in a mixed H/sub 2//H/sub /spl infin// framework.

There has been enormous world-wide interest in control engineering over the past 50 or more years. This is presumably a consequence of the importance of this technology to industry. Notwithstanding this interest, there is some evidence that support for control research is waning and that industry has become wary of the claimed benefits of control.

This paper examines some aspects of the roles of ultimate objectives and prior knowledge in system identification. It is shown that often an ultimate objective can be achieved without unique parameter or structure determination. New results based on this approach in adaptive prediction will be presented. Recent related results in adaptive control will also be examined to illustrate the role of identifciation when minimum variance control is the ultimate objective.

The choice of input signal in an identification experiment has a significant bearing on the final result. At one extreme, an input can result in violent system disturbances and the subsequent loss of product or damage to the system. At the other extreme, an input can be so conservative as to yield little or no information about the system's dynamic behaviour. The optimal input signal achieves a compromise between these extremes. This paper describes a design procedure for synthetising optimal input signals. A state-space description of the system is used, and the optimal input is chosen to minimise the variance of the parameter and state estimates. The method applies to both linear and nonlinear systems and the case of both white and coloured output noise has been considered. The optimal input signals are shown to result in a significant improvement in the identification accuracy.