Research On Environmental Perception,Recognition And Leader Following Algorithm Of The Quadruped Robot

The quadruped robot only need discrete foothold rather than continuous path when it is running, so it can move across more complex terrain and has the even better movement flexibility and environmental suitability compared with the wheeled robot and tracked robot. According to the number of legs, the legged robot can be divided into several classes including one-legged robot, biped robot, quadruped robot, hexapod robot and multi-legged robot. The quadruped robot imitates the motion type of quadruped animals, so it runes smoother than biped robot and avoids the complex mechanism of the multi-legged robot, at the same time, it can walk smoothly in complex terrain using static gait and run at high speeds in flat terrain using dynamic gait. According to the complexity and stability, the quadruped robot is a better choice in the field of legged robot.As the gradual development of the research on quadruped robot, the research focus has shifted from the structural design and motion stability analysis to how to improve the adaptability in complex environments. As a result, achieving the ability of sense surrounding and improving the autonomy is a dominant research direction and development trend. This dissertation focuses on the environmental perception system construction, terrain analysis, path planning, navigator recognition following. The main work and contributions are summarized as follows:(1) The environmental perception system based on 3D laser range finder (or 2D laser range finder+pan-tilt)+monocular camera+TOF camera is constructed. The depth information of the environment is obtained using laser range finder and TOF camera, and the laser range finder can get the rough measurement of large-scale terrain and the TOF camera can get the subtle information. Based on the depth information of the environment, the terrain is represented with the digital elevation model (DEM) and recognized by the slope, roughness and step height of the grid. The roughness is obtained through the dispersion between the slope plane in which the grid locates and its 8-neighbor height points, which avoids the detection error when using the variance of the terrain height.(2) To improve the path planning efficiency and robot security, the incremental A* search algorithm (IA*), the A* algorithm having obstacle grids extending (EA*) and its incremental form IEA* are proposed respectively. The IA* algorithm firstly searches an optimal path based on A* algorithm, then a new route from the current path to the new goal projection is added to generate a suboptimum route incrementally. In comparison with traditional method solving path planning problem from scratch, the IA* enables the robot to plan path more efficiently. The EA* extends the obstacle by means of increasing grid g-value, which makes the route far away from the obstacle and avoids blocking the narrow passage.(3) The accurate identification of navigator is the premise of following, so in order to ensure the recognition stability, the reflecting marker which can be easily distinguished from the environment through the intensity of reflected light of the laser range finder is wear by the navigator. Through coordinate mapping, the navigator location detected by the laser range finder can be projected into the image coordinate, so further judgment can be done to check the navigator location using image processing algorithm. The path connecting the robot and the navigator is planned by IEA* algorithm.(4) The single leg model is researched in order to realize the motion control of the quadruped robot, through this model, the relationship between the robot foot and the rotation of joints can be obtained. The servo control of the joint actuator can be realized through the single leg servo controller which is designed for the quadruped robot, then, the walking function of the quadruped robot can be realized through controlling the joint angle. The difficulty of following path is how to control the turning performance of the robot, so the turning control for the quadruped robot is studied, and through adjusting the motion control strategy the path following function is achieved.