Resilient Event-Triggered Terminal Sliding Mode Control Design for a Robot Manipulator

A novel non-singular terminal sliding mode controller (NTSMC) has been developed for the purpose of tracking and stabilizing tasks in uncertain electro-hydraulic robot manipulators. It is supposed that the controller communicates with the robot through a network that is vulnerable to cyber-attacks. To reduce the communication burden on the network layer and achieve resiliency against cyber-attacks, an event-based strategy is employed. For this purpose, the event-triggering rule is derived so that the Zeno-free behavior is guaranteed. Then, based on the cyber-attack characteristics, i.e., frequency and duration of the attacks, the resilient behavior of the proposed scheme in the presence of denial of service attacks, unmodelled dynamics, and external disturbance are analyzed. Moreover, to capture the nonlinear nature of the robot an experimentally validated analytical model of an uncertain 7-DoF manipulator with a hydraulic model of the joints and actuators, namely, Brokk-Hydrolek, is employed. Finally, the merits of the proposed methodology in terms of resiliency, robustness, and preservation of the communication resources are validated, and the results are compared to the state-of-the-art approaches based on the
$\rho$
index criterion
Note to Practitioners
—The aim of this study is to address the problem of network-based control of robotic manipulators. These systems, relying on the network layer for data collection and control commands, are highly vulnerable to catastrophic cyber-attacks. Furthermore, they should comply with network restrictions, such as limited bandwidth, to achieve the desired performance. Therefore, due to the collaborative behavior of robot manipulators in industries, it is vital for engineers and practitioners to be assured of achieving desired performance in the presence of these threats and limitations. As a first step to deal with these issues, a 7-DoF robotic manipulator model is mathematically formulated and experimentally validated. Then, a controller design procedure that guarantees the desired performance in spite of model uncertainties, denial-of-service cyber-attacks, and network restrictions is derived. Additionally, a clear relation between cyber-attack characteristics and designed parameters is defined while resilient behavior is maintained. Note that the proposed approach can be applied to a wide range of network-based nonlinear dynamic systems.