REDUCED ORDER BASED LADRC DESIGN AND ITS APPLICATIONS

Abstract

The work undertaken in this dissertation primarily deals with the formulation and analysis of a model independent linearized active disturbance rejection-based control (LADRC) technique that has a potential to replace the ubiquitous proportional integral derivative (PID) controller in industrial applications. The LADRC controller comprises of an extended state observer which estimates the cumulative effect of endogenous disturbances and exogenous disturbances via a fictitious auxiliary state and an estimated state feedback-based controller which is designed to ensure effective rejection of the generalized disturbance and a swift tracking of the reference signal. The existing LADRC approach is modified via the proposition of an improved technique, wherein the concept of model order reduction is incorporated for the simplification of complex higher order systems, leading to the ease of tuning, simplification of analysis and reduction in the number of controller and observer parameters. The corresponding generalized formulae are derived. The efficacy of the improved LADRC scheme is then demonstrated for the angular position control of a DC servo system. Further, the effectiveness of the proposed improved LADRC technique is verified via the comparison of response with the existing techniques in literature namely, PID control algorithm and internal model control (IMC) approach. Finally, the LADRC technique is corroborated via its implementation for the position control of the experimental setup of permanent magnet DC servo motor. The system identification of the hardware setup is undertaken and the corresponding mathematical model is developed. An extensive robustness analysis involving the effect of parametric uncertainties, disturbances and nonlinearities is also undertaken. The simulation results provide a validation to the strength of the proposed technique.