Humanoid robots can walk stably on flat ground, regular slopes, and stairs. However, because of their rigid and flat soles, adapting to unknown rough terrains is limited, moreover, to maintain large scale four-point c...Humanoid robots can walk stably on flat ground, regular slopes, and stairs. However, because of their rigid and flat soles, adapting to unknown rough terrains is limited, moreover, to maintain large scale four-point contact for foot structures to keep balance is usually a key technical problem. In order to solve these problems, the control strategy and foot structures should be improved. In this paper, a novel flexible foot system is proposed. This system occupies 8 degrees of freedom （DOF）, and can obtain larger support region to keep in four-point contact with uneven terrains; Novel cable transmission technology is put forward to reduce complexity of traditional mechanism and control strategy, and variation of each DOF is mapped to cable displacement. Furthermore, kinematics of this new system and a global dynamic model based on contact-force feedback are analyzed. According to stability criterion and feedback sensor information, a method calculating the optimal attitude matrix of contact points and joint variables is introduced. Virtual prototyping models of a 30-DOF humanoid robot and rough terrain are established to simulate humanoid robot walking on uneven ground, and feasibility of this system adapted to uneven terrain and validity of its control strategy are verified. The proposed research enhances the capability of humanoid robots to adapt to large scale uneven ground, expands the application field of humanoid robots, and thus lays a foundation for studies of humanoid robots performing tasks in complex environments in place of humans.展开更多
A gait control method for a biped robot based on the deep Q-network (DQN) algorithm is proposed to enhance the stability of walking on uneven ground. This control strategy is an intelligent learning method of posture ...A gait control method for a biped robot based on the deep Q-network (DQN) algorithm is proposed to enhance the stability of walking on uneven ground. This control strategy is an intelligent learning method of posture adjustment. A robot is taken as an agent and trained to walk steadily on an uneven surface with obstacles, using a simple reward function based on forward progress. The reward-punishment (RP) mechanism of the DQN algorithm is established after obtaining the offline gait which was generated in advance foot trajectory planning. Instead of implementing a complex dynamic model, the proposed method enables the biped robot to learn to adjust its posture on the uneven ground and ensures walking stability. The performance and effectiveness of the proposed algorithm was validated in the V-REP simulation environment. The results demonstrate that the biped robot's lateral tile angle is less than 3° after implementing the proposed method and the walking stability is obviously improved.展开更多
基金supported by National Natural Science Foundation of China (Grant No. 50775008)PhD Programs Foundation of Ministry of Education of China (Grant No. 200800061019)Hubei Provincial Digital Manufacturing Key Laboratory Foundation of China (Grant No. SZ0602)
文摘Humanoid robots can walk stably on flat ground, regular slopes, and stairs. However, because of their rigid and flat soles, adapting to unknown rough terrains is limited, moreover, to maintain large scale four-point contact for foot structures to keep balance is usually a key technical problem. In order to solve these problems, the control strategy and foot structures should be improved. In this paper, a novel flexible foot system is proposed. This system occupies 8 degrees of freedom （DOF）, and can obtain larger support region to keep in four-point contact with uneven terrains; Novel cable transmission technology is put forward to reduce complexity of traditional mechanism and control strategy, and variation of each DOF is mapped to cable displacement. Furthermore, kinematics of this new system and a global dynamic model based on contact-force feedback are analyzed. According to stability criterion and feedback sensor information, a method calculating the optimal attitude matrix of contact points and joint variables is introduced. Virtual prototyping models of a 30-DOF humanoid robot and rough terrain are established to simulate humanoid robot walking on uneven ground, and feasibility of this system adapted to uneven terrain and validity of its control strategy are verified. The proposed research enhances the capability of humanoid robots to adapt to large scale uneven ground, expands the application field of humanoid robots, and thus lays a foundation for studies of humanoid robots performing tasks in complex environments in place of humans.
基金Supported by the National Ministries and Research Funds(3020020221111)
文摘A gait control method for a biped robot based on the deep Q-network (DQN) algorithm is proposed to enhance the stability of walking on uneven ground. This control strategy is an intelligent learning method of posture adjustment. A robot is taken as an agent and trained to walk steadily on an uneven surface with obstacles, using a simple reward function based on forward progress. The reward-punishment (RP) mechanism of the DQN algorithm is established after obtaining the offline gait which was generated in advance foot trajectory planning. Instead of implementing a complex dynamic model, the proposed method enables the biped robot to learn to adjust its posture on the uneven ground and ensures walking stability. The performance and effectiveness of the proposed algorithm was validated in the V-REP simulation environment. The results demonstrate that the biped robot's lateral tile angle is less than 3° after implementing the proposed method and the walking stability is obviously improved.
