DOCTORAL THESIS PROPOSAL Biped Locomotion: Augmenting an Intuitive Control Algorithm with Learning
نویسنده
چکیده
Foot placement is a key determinant for the stabilization of walking speed and lateral motion of a biped. However, there is no closed form expression for the foot placement parameters in term of the walking speed or other gait parameters. A simple and intuitive control algorithm (called “Turkey Walking”) based on Virtual Model Control (VMC) was successfully applied to planar bipedal walking. However, it has deficiencies that pertain to the foot placement problem. I propose augmenting the algorithm with reinforcement learning (RL) algorithms to overcome the deficiencies. No dynamic model is required for the overall control architecture. The RL algorithms are used to learn the leg swing parameters, whereas the “Turkey Walking” (TW) algorithm generates the desired joints’ torque based on the gaits’ parameters. The control architecture is first tested on planar bipeds. After that, it is extended to a 3D biped. The research emphasizes on establishing the detailed structure of the architecture and illustrating the generality of the approach by simulation analysis.
منابع مشابه
Learning Control of Bipedal Dynamic Walking Robots with Neural Networks
Stability and robustness are two important performance requirements for a dynamic walking robot. Learning and adaptation can improve stability and robustness. This thesis explores such an adaptation capability through the use of neural networks. Three neural network models (BP, CMAC and RBF networks) are studied. The RBF network is chosen as best, despite its weakness at covering high dimension...
متن کاملVirtual Model Control: An Intuitive Approach for Bipedal Locomotion
Virtual model control is a motion control framework that uses virtual components to create virtual forces generated when the virtual components interact with a robot system. An algorithm derived based on the virtual model control framework is applied to a physical planar bipedal robot. It uses a simple set of virtual components that allows the robot to walk successfully over level terrain. This...
متن کاملLearning from Demonstration and Adaptation of Biped Locomotion with Dynamical Movement Primitives
In this paper, we report on our research for learning biped locomotion from human demonstration. Our ultimate goal is to establish a design principle of a controller in order to achieve natural human-like locomotion. We suggest dynamical movement primitives as a CPG of a biped robot, an approach we have previously proposed for learning and encoding complex human movements. Demonstrated trajecto...
متن کاملReinforcement Learning for a CPG-driven Biped Robot
Animal’s rhythmic movements such as locomotion are considered to be controlled by neural circuits called central pattern generators (CPGs). This article presents a reinforcement learning (RL) method for a CPG controller, which is inspired by the control mechanism of animals. Because the CPG controller is an instance of recurrent neural networks, a naive application of RL involves difficulties. ...
متن کاملLearning CPG-based Biped Locomotion with a Policy Gradient Method: Application to a Humanoid Robot
In this paper we describe a learning framework for a central pattern generator (CPG)-based biped locomotion controller using a policy gradient method. Our goals in this study are to achieve CPG-based biped walking with a 3D hardware humanoid and to develop an efficient learning algorithm with CPG by reducing the dimensionality of the state space used for learning. We demonstrate that an appropr...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1999