With the development of natural science and technology, bionic robots are gradually becoming popular in different workplaces, and bionic quadrupedal robots are one of the most widely used categories at present. In recent years, people have been looking for ways to optimize the gait of bionic quadrupedal robots, so that they can complete their work more quickly and efficiently while adapting to environmental changes. Based on recent research and innovations, the existing mechanisms of the robot have been improved and different designs have been introduced to enhance its capabilities, such as improved balance, increased speed, or better obstacle navigation. At the same time, multi-objective optimization and deep reinforcement learning have been used to train the robot's gait through intelligent algorithms. Additionally, different optimization schemes have been tested using various simulation environments to ensure excellent motion control across different terrains. In this paper, by collecting gait optimization solutions in recent years, it is found that existing techniques have greatly reduced gait instability and improved the overall performance of the robot while maintaining efficient robot motion, which greatly satisfies the usage requirements.
Read full abstract