A Spring Compensation Method for a Low-Cost Biped Robot Based on Whole Body Control

Bipedal Robot Gait Generation Using Bessel Interpolation

This paper introduces a novel approach to bipedal robot gait generation by proposing a higher-order form through the parameter equation of first-order Bessel interpolation. The trajectory planning for the bipedal robot, specifically for stepping up or down stairs, is established based on a three-dimensional interpolation equation. The experimental prototype, Roban, is utilized for the study, and the structural sketch of a single leg is presented. The inverse kinematics expression for the leg is derived using kinematic methods. Employing a position control method, the angle information is transmitted to the robot's joints, enabling the completion of both downstairs simulation experiments and physical experiments with the Roban prototype. The analysis of the experimental process reveals a noticeable phenomenon of hip and ankle joint tilting in the robot. This observation suggests that low-cost bipedal robots driven by servo motors exhibit low stiffness characteristics in their joints.

A Multi-Agent Reinforcement Learning Method for Omnidirectional Walking of Bipedal Robots

Achieving omnidirectional walking for bipedal robots is considered one of the most challenging tasks in robotics technology. Reinforcement learning (RL) methods have proved effective in bipedal walking tasks. However, most existing methods use state machines to switch between multiple policies and achieve omnidirectional gait, which results in shaking during the policy switching process for bipedal robots. To achieve a seamless transition between omnidirectional gait and transient motion for full-size bipedal robots, we propose a novel multi-agent RL method. Firstly, a multi-agent RL algorithm based on the actor-critic framework is designed, and policy entropy is introduced to improve exploration efficiency. By learning agents with parallel initial state distributions, we minimize reliance on gait planner effectiveness in the Robot Operating System (ROS). Additionally, we design a novel heterogeneous policy experience replay mechanism based on Euclidean distance. Secondly, considering the periodicity of bipedal robot walking, we develop a new periodic gait function. Including periodic objectives in the policy can accelerate the convergence speed of training periodic gait functions. Finally, to enhance the robustness of the policy, we construct a novel curriculum learning method by discretizing Gaussian distribution and incorporate it into the robot's training task. Our method is validated in a simulation environment, and the results show that our method can achieve multiple gaits through a policy network and achieve smooth transitions between different gaits.

Optimization Design and Performance Analysis of a Bionic Knee Joint Based on the Geared Five-Bar Mechanism

Animal joint motion is a combination of rotation and translational motion, which brings high stability, high energy utilization, and other advantages. At present, the hinge joint is widely used in the legged robot. The simple motion characteristic of the hinge joint rotating around the fixed axis limits the improvement of the robot's motion performance. In this paper, by imitating the knee joint of a kangaroo, we propose a new bionic geared five-bar knee joint mechanism to improve the energy utilization rate of the legged robot and reduce the required driving power. Firstly, based on image processing technology, the trajectory curve of the instantaneous center of rotation (ICR) of the kangaroo knee joint was quickly obtained. Then, the bionic knee joint was designed by the single-degree-of-freedom geared five-bar mechanism and the parameters for each part of the mechanism were optimized. Finally, based on the inverted pendulum model and the Newton-Euler recursive method, the dynamics model of the single leg of the robot in the landing stage was established, and the influence of the designed bionic knee joint and hinge joint on the robot's motion performance was compared and analyzed. The proposed bionic geared five-bar knee joint mechanism can more closely track the given trajectory of the total center of mass motion, has abundant motion characteristics, and can effectively reduce the power demand and energy consumption of the robot knee actuators under the high-speed running and jumping gait.