Studies On Technology Of Localization And Path Planning Based On Tank Mobile Robots

Autonomous mobile robot is a hotspot filed of AI researches. The technology of localization and path planning is not only base and crucial element, but also the symbol that a robot's intelligence is high or low.The component and sensing system is induced in this paper in detail.There are two kinds of localization approaches, including absolutely localization approach by GPS and relatively localization approach by encoder. The two localization approaches are not enough accurate, so they are often combined to use. Gained data by GPS is filtered through the variable dimension Kalman filter technology, then the data after filtered and the gained data through relatively localization approach is to optimum linearity combined. The aim is to further accuracy of localization.Artificial potential field (APF) is a common approach in all robot path planning approaches. An improved algorithm based on APF is proposed in this paper. The improved algorithm can overcome the two existent problems in traditional APF and can improve the efficient and make the path better. Genetic algorithm (GA) can seek an optimum solution in the complex obstacles. A fitness function that is different form traditional GA, and simulation have been done. The experiment results indicate that the algorithm is good of real-time and quickly, and is fit for path planning of Tank mobile robot, and is fit for applying under complex environment.The formation control of AMB based on Ad-hoc networks has been researched in this paper to apply it to military filed. Ad-hoc networks is applied in many filed for its convenience and rapidness. The traditional APF algorithm is improved. A special formation is maintained in some require through communication gravitation.