Controller Design for a Skid-Steered Robot and Mapping for Surveillance Applications

Abstract

Skid-steered robots, with their robust structure and manoeuvrability, are generally used as outdoor mobile robots. Both kinematic and dynamic modelling of these robots is difficult due to sliding and rolling inherent in general curvilinear motion. In order to improve motion and pose estimation, this paper proposes a kinematic and dynamic model for skid-steered mobile robots. A PID controller, tuned using Genetic Algorithm, based on the dynamic model is then proposed for accurate control of the skid-steered robot. The dynamic model developed enables motion planning for general planar motion. The coefficient of rolling resistance, the coefficient of friction, and the shear deformation modulus, all of which have terrain-dependent values are accommodated in this model. Surveillance bots are of great importance in protecting and saving human life. In this context, mobile and multi-functional robots which map their surroundings are adopted as a means to reduce environmental restructuring and the number of devices used to cover a given area. Skid-steered robots are robust and, therefore, are ideal for surveillance applications.