Correlation-based direct tuning of multivariable controllers with stability constraints based on spectral analysis

This paper proposes the model-free controller tuning method for MIMO systems ensuring closed-loop stability by imposing a set of constraints for discretely defined frequency points. This constraints are defined as a set of convex constraints on maximum singular values of a specific transfer function matrix estimated by spectral analysis. For linearly parameterized controllers, the optimal parameters can be obtained by convex optimization using the acquired input/output data. The resulting controller is guaranteed to stabilize the plant as the data length tends to infinity. The effectiveness of the proposed method is confirmed by experiments for a tension-and-speed control apparatus which is a two-input two-output system.