This paper is focused on a comparison between classic Feedback LinearizationControl and a proposed Active Disturbance Rejection Control method. The proposed technique does not require a precise mathematical description of the system, since it is based on the online estimation and rejection of the unmodeled elements of the dynamics. Robustness of the closed-loop control system (against external perturbation and parameters uncertainty) is discussed here. A model of a SCARA robot manipulator is used in the conducted case study as an exemplary plant. Conclusions are supported with results obtained with numerical simulations.
The classical sliding mode control (SMC) is a robust control scheme widely used for dealing with nonlinear systems uncertainties and disturbances. However, the conventional SMC major drawback in real applications is the chattering phenomenon problem, which involves extremely high control activity due to the switched control input. To overcome this handicap, a pratical design method that combines an adaptive neural network and sliding mode control principles is proposed in this paper. The controller design is divided into two phases. First, the chattering phenomenon is removed by replacing the sign function included in the switched control by a continuous smooth function; basing on Lyapounov stability theorem. Then, an adaptive linear neural network, that has the role of online estimate the equivalent control in the neighborhood of the sliding manifold, is developed when the controlled plant is poorly modeled. Simulation results show clearly the satisfactory chattering free tracking performance of proposed controller when it is applied for the joints angular positions control of a 6-DOF PUMA 560 robot arm.
This paper proposes a new robustification strategy of a primary H∞ controller based on a systematic selection of adequate fractional weights. Its robust performance (RP) margin is well enhanced with respect to unstructured multiplicative uncertainties presented in linear feedback systems. The proposed robustification strategy provides a robustified H∞ controller when following proposed hierarchical control is well respected. First, the primary
H
∞
controller is synthesized from solving a weighted-mixed sensitivity H∞ problem using initial integer weights. Thus, an initial RP margin is obtained. Second, an automatic selection of adjustable fractional weights is performed by the particle swarm optimization algorithm, in which some proposed tuning rules are accordingly well satisfied. Third, frequency response data of these weights are computed and then fitted by corresponding approximated integer weights using a frequency identification technique. Finally, these weights reformulate a new weighted-mixed sensitivity problem. The optimal solution to this problem updates the previous initial RP margin. These last three steps are repeated as the updated RP margin is diminished. Otherwise, the proposed hierarchical control is achieved by selecting the best adjustable fractional weights, providing, therefore, the best approximated integer weights and leading, therefore, to the robustified H∞ controller. In order to confirm the effectiveness of our proposed hierarchical control, primary and robustified H∞ controllers are applied on a permanent magnet synchronous motor where its actual behavior is modeled by an unstructured multiplicative uncertain model. The results obtained are compared in frequency domains using the singular value plots of their sensitivity functions. Otherwise, the same results are compared in time domains using the Powersim® software.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.