Affiliation: | 1. School of Electronic Information Engineering, Inner Mongolia University, Hohhot, China Contribution: Formal analysis, Methodology, Software, Writing - original draft, Writing - review & editing;2. School of Electronic Information Engineering, Inner Mongolia University, Hohhot, China Contribution: Formal analysis, Validation, Writing - review & editing;3. School of Mathematics and Statistics, Xidian University, Xi'an, China Contribution: Formal analysis, Validation, Writing - review & editing;4. School of Electronic Information Engineering, Inner Mongolia University, Hohhot, China |
Abstract: | This paper investigates the dynamic cooperative learning control for high-order output-feedback systems under the output constraints. To avoid complex computation, we use a system transformation strategy in control design. Only one neural network (NN) is employed to approximate the unknown synthetic function for each agent. Subsequently, a NN-based cooperative learning control mechanism is designed by introducing the barrier Lyapunov function (BLF). The proposed mechanism expands the NN approximation domain of the transformed systems, and the output of all subsystems remains constrained. Further, the NNs on identified uncertain system dynamics are used to construct the experience-based controllers to carry out same control tasks. Finally, the theoretical authenticity is demonstrated by a numerical example. |