Control of the CEDRA Brachiation Robot Using Combination of Controlled Lagrangians Method and Particle Swarm Optimization Algorithm

This paper studies the control of a brachiating robot imitating the locomotion of a long armed ape. The robot has two revolute joints, but only one of them is actuated. In this paper, after deriving dynamic model of the robot, the Controlled Lagrangians (CL) method is used to design a controller for point to point locomotion. The CL method involves satisfying a number of equations called matching conditions. The matching conditions are derived using the extended λ-method in the form of a set of partial differential equations (PDEs). Solving the PDEs, a class of controllers is found that satisfies the matching conditions. The fittest controller in the class of controllers is then chosen by particle swarm optimization algorithm. Performance of the developed controller is investigated by numerical simulations. Finally, experiments are performed to validate theoretical results.