In this paper, a new method of online planning high smooth and time-optimal trajectory for robotic manipulators that applies an adaptive elite genetic algorithm with singularity avoidance (AEGA-SA) is presented. The strategy is designed as a combination of the time-optimal trajectory planning with quintic polynomial in Cartesian space. For improving optimization performance, elitist group and adaptive adjustment mechanisms are used based on genetic algorithm (GA) framework. In the meantime, GA is combined with singularity avoidance mechanism to avoid the singularities appearing in the trajectory, improves the recognition capability of optimum solution. Experimental results show that, the proposed approach is more effective and better performance than the original GA and its variants, with ensuring a both smooth and efficiency performance for the robotic manipulators.
1 October 2017 in Journal of the Franklin Institute
This paper addresses the fast trajectory tracking problem for discrete-time nonlinear systems, and whereby a finite-time model-free adaptive control (FMFAC) approach is innovatively devised. Here, the nonlinear systems’ dynamics are completely unknown and the widespread disturbances are considered. The nonparametric dynamic linearization technique and the adaptive extended disturbance observer are firstly deployed to achieve real-time linearization of the discrete-time nonlinear system, and an efficient finite-time performance index function is thereafter devised, and thereby contributing to a novel FMFAC control scheme. Theoretical analysis indicates the finite-time model-free tracking control within a stable region can be guaranteed, and the boundedness of the control input signal can be ensured rigorously, even though under the effect of various disturbances. The key points of the proposed FMFAC approach are model-free and possessing the fast regulation ability, and the corresponding effectiveness and superiority are validated via simulation studies.
18 October 2021 in Transactions of the Institute of Measurement and Control
In this paper, the robust trajectory tracking control problem of disturbed quadrotor unmanned aerial vehicles (UAVs) with disturbances, uncertainties and unmodeled dynamics is addressed, by devising a novel compound robust tracking control (CRTC) approach via data-driven cascade control technique. By deploying the data-driven philosophy, a data-based sliding-mode surface is proposed, and thereby contributing to strong adaptability to nonlinearity and model-unknown properties of the UAVs. By utilizing the backstepping technique, virtual control strategy and a novel cascaded compound robust PD control structure, the attitude and position subsystems are efficiently cohered such that a data-driven cascaded compound robust controller containing both PD control and sliding-mode control can be developed to conquer the lumped disturbances induced by uncertainties, disturbances and unmodeled dynamics. Eventually, the asymptotic convergence of the tracking errors with respect to both attitude and position subsystems can be guaranteed rigorously. Simulation studies on a prototype quadrotor UAV are performed to evaluate the efficacy and superiority of the devised CRTC method.
This paper is concerned with data-driven adaptive tracking control for unknown autonomous marine vehicles (AMVs) with uncertainties and disturbances. By deploying the data-driven technique and observer design, an equivalent data model of the AMV is firstly established. Based on the proposed data model, a novel data-driven adaptive tracking controller is designed, and the corresponding stability analysis for the closed-loop AMV system is presented theoretically. Finally, simulation studies are given to demonstrate the validity of the main results.