Stability Analysis and Quantitative Tuning of Modified Uncertainty and Disturbance Estimator-Based Control for a Class of Time-Delay Processes

Modified uncertainty and disturbance estimator (MUDE) based control shows great robustness against unexpected dynamics in the presence of time delay. Although effective in several applications, the stability is not theoretically guaranteed and the controller tuning is still challenging. The resulting control performance highly depends on the specialized expertise. To this end, this article analyzes the stability in terms of the scaled controller parameters, and proposes a unified quantitative tuning rule for a MUDE structure for a class of first-order time-delay process model (including open loop stable, integrating and unstable time delay processes). The stable regions of the MUDE parameters are theoretically derived. The influence mechanism of the scaled filter parameter and feedback gain on the disturbance rejection (DR) response is analyzed in detail, and then used to determine the feedback gain. Exhaustive robustness testings reveals that the delay margin is a critical index for controller tuning. A unified, robust and quantitative tuning rule is concluded as well as an explicit retuning guideline. Some comparison simulations and laboratory experimental are conducted to demonstrate the efficiency of the proposed tuning rule.
Note to Practitioners
—This paper provides a unified quantitative tuning method to apply the MUDE controller on a class of time-delay industrial processes (including open-loop stable, integrating and unstable time delay processes). Time delay widely exists in many industrial processes, which brings difficulties to controller design and parameter tuning. In terms of DR of time-delay systems, many advanced controllers are designed usually for a specific class of processes, such as open-loop stable processes. However, a unified approach (design and tuning, etc) is highly desired for ease of application. In this regard, MUDE provides a unified and effective control framework with many features expected by practitioners, such as simple structure and decoupled tracking & DR. Further, a unified, robust and quantitative tuning rule is provided in this paper, based on parameter scaling, stability analysis and robustness testing. Therefore, the engineers can simplify the time-consuming tuning procedure and have a comprehensive grasp of the performance and robustness of the time-delay control system. This paper aims to develop a ease-of-use controller with fixed control structure and quantitative tuning rule, just like PID controller.