Research on Dynamic Characteristic Simulation and Control Algorithm of a Modular Bipedal Robot

In the field of intelligent robotics, bipedal robots have become a key vehicle for the integration of artificial intelligence and mechanical technology, relying on their potential in the fields of service, medical and industrial applications. Aiming at the limitations of current bipedal robot technology, such as high manufacturing cost, large structure and insufficient functionality, this study develops a low-cost, modularized bipedal robot system with support for secondary development, which can effectively meet the practical needs of teaching practice, scientific research and engineering applications. The bipedal robot for undergraduate teaching practice and competitive competitions is used as the research object, and the modularized structure design is adopted to shorten the research and development cycle, prolong the life span, and reduce the cost. A physical prototype has been developed, and its modularized architecture supports the flexible configuration of functions. The core research includes: establishing a single-leg kinematic model (forward and inverse) based on the D-H parametric method, using SolidWorks to complete 3D modeling and motion simulation; optimizing and validating the gait control algorithm on the Nvidia Isaac Gym platform; and evaluating the algorithms quantitatively for their dynamic performance and motion stability during training.