• CN: 11-2187/TH
  • ISSN: 0577-6686

Journal of Mechanical Engineering ›› 2023, Vol. 59 ›› Issue (6): 214-225.doi: 10.3901/JME.2023.06.214

Previous Articles     Next Articles

Modeling and Mechanical Parameters Identification of Interaction Mechanics between Rotary Robotic C Type Leg and Sandy Terrain

YANG Chuanxiao1, HU Xiong1, SUN Shibin1, GU Bangping1, DING Liang2, TANG Dewei2, GAO Haibo2, DENG Zongquan2   

  1. 1. School of Logistic Engineering, Shanghai Maritime University, Shanghai 201306;
    2. State Key Laboratory of Robotics and System, Harbin Institute of Technology, Harbin 150001
  • Received:2022-04-26 Revised:2022-08-25 Online:2023-03-20 Published:2023-06-03

Abstract: The establishment of mechanical model and parameter identification of the interaction between robot foot and sand are important constraint conditions and physical information for multi-modal perception and decision-making for legged mobile robot on sandy terrain surface. Since high trafficability and adaptability on sandy terrain surface can be achieved for the C-shaped legged configuration, the influences of positional posture and speed vector to the dynamic foot-terrain interaction mechanics for C-shaped leg under the condition of swing gait based on resistance force theory in terradynamics are fully considered, then a contact mechanics model between the C-shaped foot and sandy soil is established. Through rotational contact mechanics experiments between C-shaped leg and sandy terrain surface under three width group conditions, the data are extracted and variation laws of contact forces in horizontal and vertical direction to different attitude angles are analyzed. Hence, the linear expression is obtained through analytical derivation of the integral model, these unknown parameter matrixes are deduced item by item based on the recursive least square algorithm. Finally, based on the item-by-item matrix function indicating iterative input-output relations, the parameters identification results depending on the existing data sample size space are obtained. Compared with the experimental results, the errors of predicted vertical force and horizontal force generated in identification are 4.05% and 4.22% respectively, which verifies the accuracy and effectiveness of parameter identification. Some physical properties of sandy terrain can be revealed through these identified parameters, and the influences of leg geometrical features on mechanical models can also be indicated by these reference values.

Key words: C type leg, rotary interaction, terradynamics, recursive least square method, parameters identification

CLC Number: