Modelling and implementation of a hydraulic actuated quadruped robot

In this thesis, a novel robot leg was designed to use in a quadruped robot with an innovative approach. The application of optimization techniques for the design itself is one of the innovative aspects of the study. In the design phase, the mechanism and the topology of the robot leg was optimized to reduce the load of the hydraulic power unit, and consequently, the energy efficiency, which is one of the major problems related to mobile robots, was tried to be improved. Furthermore, to design various controllers to control position of the designed robot leg, and to benchmark performances of the designed controllers, simulations were performed. In addition, optimization-based approaches were used to shape the parameters of the designed controllers according to the determined metrics. In this context, the mathematical model and the physical model of the designed robot leg were obtained to use in simulations. Then, to verify the validity and reliability of the controllers obtained with the simulations, control experiments were performed on the real system after manufacturing the leg model prototype. The mentioned optimizations were realized by using The Bees Algorithm which is a novel search algorithm based on the swarm-intelligence metaheuristic and Genetic Algorithm which is widely used in the literature and at the same time is another metaheuristic algorithm. Finally, a quadruped robot concept with self-balancing and high mobility was proposed based on the already obtained model of the designed and manufactured robotic leg. Furthermore, a trajectory design was made for the robot leg during the walking of a quadruped robot. In this context, a Central Pattern Generator was designed in order to provide the trajectory determination of the robot leg during quadruped walking. Gait patterns of a quadruped robot were analyzed by using the already obtained model of the designed and manufactured robot leg.