A 3D stable trot of a quadruped robot over uneven terrains. (February 2017)

• Record Type: Journal Article
• Authors: Khorram, Mahdi; Moosavian, S Ali A
• Is Part Of: Proceedings of the Institution of Mechanical Engineers. Volume 231:Number 3(2017)
• Page Start: 555
• Page End: 573
• Publication Date: 2017-02
• Subjects: Quadruped robot -- dynamics stability -- trot gait -- uneven terrains -- 3D gait generation; Mechanical engineering -- Periodicals; 621.05
• DOI: 10.1177/0954406215617492
• Languages: English
• ISSNs: 0954-4062

Abstract

Legged robots have superior advantages rather than wheeled robots for moving over uneven terrains in the presence of various obstacles. The design of an appropriate path for the main body and legs is an important issue for such robots especially on the uneven terrains. In this paper, the focus is to develop a stable gait for a quadruped robot to trot on uneven terrains. First, a stability condition is developed for a whole-body quadruped robot over uneven terrains based on avoiding the tumbling. By using a simple model, a point with zero moments is calculated in the three-dimensional space. Then, the reference path of this point is determined so that the tumbling moments become zero. The path of the main body will be calculated by using an optimal controller. The main feature of the proposed gait generation framework is that the height of robot can change continuously and stably on uneven terrains. To evaluate the robot stability, the tumbling moments around diagonal lines are calculated and some methods are proposed to reduce these moments to improve the robot stability. The tip of swing foot is also planned to avoid any collision with the environment. The proposed method will be demonstrated using an 18-Degrees of freedom (DOF) quadruped robot in simulation and experimental studies. The experimental setup is a small-size quadruped robot, which is composed of a rectangular plate as its main body with four legs that each one has three active joints with DC servo motors. Obtained results reveal that the robot can trot on uneven terrains stably. Besides, the comparison with the previous methods approves the merits of proposed algorithm on uneven terrains.