Modular wheeled mobile robot

This paper describes transputer-based design and application of fully decentralized estimation and control algorithms for a modular wheeled mobile robot (WMR). Fully decentralized estimations achieved by using a distributed and decentralized extended information filter (DDEIF). Optimal control is obtained locally using reduced order models with minimized communication requirements, in a scalable network of control nodes. The WMR is a modular vehicle system with nonlinear kinematics and distributed means of acquiring information. Its general kinematic model is constructed by using plane motion kinematics to derive forward and inverse kinematics for a generalized simple wheeled vehicle. This is then used to develop a modular decentralized kinematic model, which is combined with a fully decentralized control algorithm to provide a WMR control system. The issue of which WMR modules need to communicate and the information they need to exchange is then considered. Transputer architecture is used as the basis for hardware and software design as it supports the extensive communication and concurrency requirements characteristic of modular and decentralized systems. All the software is written in Parallel ANSI C and consists of two main parts: a configuration program and a nodal program which is loaded at each module. Pseudocode is used to demonstrate how the software achieves concurrency, modularity, and local internodal communication. The advantages of a modular WMR vehicle include scalability, application flexibility, low prototyping costs, and high reliability.