Responsive Motion Control for Robot Soccer Navigation Using Adaptive Social Force Framework

Abstract

This paper presents a modified Social Force Model (SFM) for navigation control of a soccer robot application. We modified the way of determining the parameter value of the gain factor, , of the SFM using the Fuzzy Inference System (FIS), so that the value of the gain factor, , is adaptive. The purpose of the gain factor adaptation is that the robot can move responsively but not over-reactive when it encounters an obstacle at high speed, which is a weakness of SFM with fixed parameters. Modification of SFM parameters using FIS is hereinafter referred to as the Fuzzy-based Social Force Model (F-SFM). We used this technique on a soccer robot with an omnidirectional drive platform with three motors. As an experiment, several modifications to the FIS rules were made and compared to the SFM with fixed parameters. The simulation-based experimental results show that the proposed method outperforms the SFM method with fixed-parameters, and the computation time does not differ significantly so that it can be applied for real implementation.