Nonlinear and intelligent control based controller design for nonlinear systems

This Ph.D. dissertation presents a number of new fruitful control strategies for regulating linear and nonlinear systems. To appeal to a broader audience of control professionals and researchers, these strategies are introduced at varying levels of complexity and difficulty. For regulating linear systems, which can be modeled by integrating or unstable process transfer functions with time delays, new analytical adjustment formulas based on the centroid of the stability region are proposed to solve the difficulties of setting the gains of the PI-PD controller. The proposed methods relative to the techniques published in the literature are straightforward and time-saving. Also, a slightly more intricate scheme based on a dynamic sliding mode controller and PID disturbance observer is developed for controlling linear integrating and unstable industrial processes when a more robust controller is required to operate in a hostile industrial environment associated with time-varying turbulences and parameter perturbations. The proposed control offers reduced chattering levels with a strong performance against sharp disturbances and model fluctuations. Moreover, a new adaptive dynamic nonsingular terminal sliding mode control and a new PID fractional disturbance observer are introduced for controlling industrial processes in the form of second-order linear processes with time delays. In comparison to the aforementioned technique, the proposed controller with a somewhat more sophisticated structure delivers smaller tracking errors, chattering rates, and overshoots. However, in systems with strong nonlinearity, linear systems may not allow the controller to appropriately compensate for nonlinear forces. As a result, this dissertation proposes various new nonlinear control mechanisms for complex nonlinear systems. In this context, a novel adaptive backstepping control technique, that is convenient for execution in real-time and robust against disturbances and parameter perturbations, is proposed to regulate nonlinear triangular systems. In addition, an innovative adaptive dead-beat sliding mode control for governing nonlinear uncertain nontriangular systems is built, which offers fewer chattering levels and less tracking errors than techniques taken from the literature. Furthermore, the controller's topology is updated utilizing neural networks and fuzzy logic systems to forecast unmodeled nonlinear components in extremely sophisticated unknown nonlinear systems. To demonstrate the practicality and superiority of the proposed methodologies, various simulated examples, simulated applications in industrial settings, and several real-time tests are employed. The real-time setups that have been used for testing the proposed techniques include a twin rotor multi-input multi-output system, a cart inverted pendulum, and a 7.5-kW induction motor.