System Integration,Parameter Identification,and Locomotion Control Of A High-Dynamic Bipedal Robot

Compared to tracked and wheeled mobile robots,legged robots possess unique footground contact characteristics that make them particularly well-suited for traversing unstructured and rugged terrain.Bipedal robots,with their human-like walking style and body structure,have shown great potential in both industrial and service industries,and have garnered the attention of numerous researchers in recent years.In the motion of a bipedal robot,the contact between the foot and the ground is the only input to the system.Due to its strong coupling and nonlinearity,the foot-ground interaction can complicate the model and reduce the control domain,thus affecting the dynamic performance of the robot.This paper proposes a point contact foot-end design to simplify the foot-ground contact model and improve the dynamic performance of the bipedal robot.To achieve this,we built a bipedal robot prototype platform and integrated a hardware and software control system.Furthermore,we established a kinematic and dynamics model for the robot and developed a highly dynamic and robust control method based on the identification of model parameters.The main focus of this paper is to study the aforementioned techniques to improve the performance of bipedal robots:1.The prototype platform construction and hardware system integration of the bipedal robot system.The topology and leg configuration of the bipedal robot was designed based on the bionic mechanism of the ostrich,and the entire prototype system was built.The hardware control system such as the joint actuator,core controller,inertial measurement unit,real-time input/output human-machine interaction unit,and power management system is integrated with the prototype platform to achieve high bandwidth and strong real-time interaction between the controller and each unit.2.The kinematic model and dynamic model of the bipedal robot are established.By analyzing the topology and geometric characteristics of the robot and establishing the.coordinate transformation relationship between different coordinate systems based on the MDH method,the positive and negative kinematic mapping relationship between the working space of the robot’s foot end and the joint space is further deduced;the kinetic characteristics of each linkage are recursively modeled by the Newton-Euler method to establish the single-leg kinetic model,and the floating base dynamic recursion method is further introduced The modeling of the dynamics of the robot is based on the Newton-Euler method,which lays the foundation for the model-based control of the bipedal robot.3.Robot single-leg dynamics parameter identification.Based on the established kinetic symbolic model,the minimum inertia parameter set is obtained by linearization transformation;the excitation trajectory is designed and optimized,and the identification data are collected and processed under this excitation trajectory;the least squares method is used to find the minimum inertia parameter values,and the accuracy of the experimental results is verified under the excitation trajectory and the verification trajectory to provide more accurate model parameters for the kinetic-based motion control.4.Highly dynamic bipedal robot motion control method based on hierarchical control framework and experimental validation.A hierarchical control framework is designed for highly dynamic motion control,including a state estimation algorithm to calculate the robot body state with joint encoder and inertial measurement unit data,an optimal plantar force optimization algorithm based on center-of-mass dynamics and friction cone constraints,and a whole-body motion control strategy based on multi-task priority,and the control algorithm is verified by high dynamic experiments,multi-terrain adaptation experiments and antidisturbance experiments.The robustness and high dynamics of the control algorithms are verified through high dynamic experiments,multi-terrain adaptation experiments,and antidisturbance experiments.