Some approaches have been presented for integration of control system design and tuning of a controlled object. The integration is more effective than the general separate designs because parameters of the controlled object can influence the properties of the control system.
In this paper, we propose an application of optimal control system design using the asymptotic characteristics to the simultaneous design of a control system and a controlled object.
The optimal control system approaches to the decoupled control system as the weighting factors for the control input variables in quadratic performance index tend to zero. In the decoupled control system, the responses between the reference signal and the output signal of the controlled object aren't influenced by parameters of the controlled object. We propose the following simultaneous design method using these properties.
(1) In the design of the decoupled control system, the parameters of the controlled object are decided in order to minimize the magnitude of the control input.
(2) In the design of the optimal control system, the weighting factors are adjusted for trade off between the response speed of the controlled output and the magnitude of the control input.
By tuning of the controlled object in the decoupled control system, the design procedure becomes simpler than one in the optimal control system because the response speed of the controlled output is fixed by the selection of the weighting factors and is not changed with the parameters of the controlled object.
View full abstract