Using Interval Type-2 Recurrent Fuzzy Cerebellar Model Articulation Controller Based on Improved Differential Evolution for Cooperative Carrying Controller of Mobile Robots

Abstract

Mobile robot is widely utilized in various fields such as navigation control, obstacle avoidance and object carrying. For keeping away from obstacles to avoid collision and preventing object carrying from dropping down, we propose a state manager (SM) designed to assist the mobile robots so that they can switch operation between wall-following carrying (WFC) and toward goal carrying (TGC) by different external condition. In this controlling model, interval type-2 recurrent fuzzy cerebellar model articulation controller (IT2RFCMAC), embedded with a modified evolutionary optimization and dynamic grouping differential evolution (DGDE), is implemented for WFC and TGC. By adopting reinforcement learning strategy, mobile robots equip with adaptively wall-following control to make cooperative carrying control in real.