Application of statistical techniques in modeling and optimization of a snake robot

摘要

SUMMARY This paper provides a general framework based on statistical design and Simulated Annealing (SA) optimization techniques for the development, analysis, and performance evaluation of forthcoming snake robot designs. A planar wheeled snake robot is considered, and the effect of its key design parameters on its performance while moving in serpentine locomotion is investigated. The goal is to minimize energy consumption and maximize distance traveled. Key kinematic and dynamic parameters as well as their corresponding range of values are identified. Derived dynamic and kinematic equations of n -link snake robot are used to perform simulation. Experimental design methodology is used for design characterization. Data are collected as per full factorial design. For both energy consumption and distance traveled, logarithmic, linear, and curvilinear regression models are generated and the best models are selected. Using analysis of variance, ANOVA, effects of parameters on performance of robots are determined. Next, using SA, optimum parameter levels of robots with different number of links to minimize energy consumption and maximize distance traveled are determined. Both single and multi-criteria objectives are considered. Webots and Matlab SimMechanics software are used to validate theoretical results. For the mathematical model and the selected range of values considered, results indicate that the proposed approach is quite effective and efficient in optimization of robot performance. This research extends the present knowledge in this field by identifying additional parameters having significant effect on snake robot performance.