On Iterative Feedback Tuning and Disturbance Rejection using Simple Noise Models

The objective of this contribution is to present a simple methodfor tuning a controller for disturbance rejection by means ofiterative feedback tuning. The controller parameters arefound by minimizing a specified cost function using gradientsearch techniques. The gradients are estimated using data fromiterative experiments. This is a rather difficult problem if noexternal input excitation is allowed. The basic idea here is to use a simple fixed continuoustime noise model as a design variable to specify the desirableclosed loop properties. We will study three closely related controller design methods based on thenoise model, namely minimum variance control, weighted minimumvariance control and a loop shaping design. We propose an IFTcost function based on a output signal weighting tailored to the fixed noise model. The approach is illustrated on a simple example.