Dynamics and Compliance Control of a Linkage Robot for Food Chewing

Abstract

A chewing robot based on a six-bar linkage is developed to reproduce a variety of human chewing trajectories in 2-D space. In order to mimic human chewing behaviors for the purpose of objective measurements of food texture changes during chewing and preparation of food bolus, a compliance control is considered for the robot. In this paper, the dynamics of the robot is derived and validated through both simulations and experiments. Food models of peanuts and marshmallow are developed to simulate the food force-compression profiles. The compliance control of the robot is realized by a combination of an adaptive fuzzy velocity controller and a fuzzy adaptation mechanism for type of foods. By means of experimental comparisons between the chewing velocity and force results with food models, it is found that the compliance control proposed is able to control the chewing velocity in adapting to the food texture. The chewing robot developed is a valuable potential tool for evaluating food properties and bolus preparation in the field of food sciences.