Adaptive Sliding Mode Control of Constrained Quadruped Robot

This paper presents a novel approach for controlling a quadruped robot in a constrained environment by proposing a robust sliding mode control law based neural network. The use of this technique for the identification and control of complex systems has become increasingly popular due to their ability to make generalizations. The proposed control structure consists of switching controller and equivalent controller, where the latter is estimated using a neural network. The control law designed can drive a constrained quadruped robot of 12 degrees of freedom to track desired trajectories accurately. To demonstrate the efficacy of the proposed control law, we have tested it in various conditions and scenarios. The results obtained indicate that the proposed control law not only performs well but also exhibits remarkable robustness. Overall, this research provides a promising method for controlling quadruped robots in constrained environments, which can be utilized in various real-world applications.