Sliding-mode controller design and optimization for flexible manipulators based on estimation of distribution algorithms

To improve the robustness and position precision of the endpoint control of flexible-link manipulators (FLMs), a sliding-mode controller optimization method based on the estimation of distribution algorithm (EDA) is proposed. First, the dynamics of FLMs are decoupled into a slow subsystem and a fast one by using singular perturbation theory. Two sliding-mode controllers with uncertain parameters for the slow and fast subsystems are designed, respectively. Then, the controller optimization strategy based on the EDA is given, by which the sliding-mode controllers are optimized. Finally, a contrastive simulation on control of FLMs between the sliding-mode control based on the EDA and the traditional sliding-mode control is carried out. The results show that the former is superior to the latter on response speed, overshoot, control precision and vibration damping.