Comparing two approaches to nonlinear time-delayed 1st-order plants control

The paper introduces two new approaches to control of simple but frequently considered nonlinear plants with a time-delayed dominant first-order dynamics. To guarantee a consistent performance over a broader operating range, in dealing with nonlinear plants, typically different gain scheduling approaches may be used as e.g. the exact linearization, velocity linearization, flatness-based model-free PID control, control based on generalized transfer functions to nonlinear systems, etc. By considering example of a time delayed one-tank liquid level control in a hydraulic plant the paper introduces a new alternatives to these traditional approaches based on a generalization of the exact linearization by using two types of linearized plant models. They allow to consider unmodelled dynamics and a nonlinear predictive disturbance-observer (NPDO) based integral action. The resulting two-degree-of-freedom model-based control shows new possibilities in designing transients with required monotonicity properties and different degrees of optimization.