Study On Mobile Miniature Self-reconfigurable Robot System

The mobile self-reconfigurable robot consists of several unit robots, which can construct distributed robotic system in discrete state as well as interlink with each other to be a self-reconfigurable robot. It has the both the advantages of distributed robot that has large moving range and high efficiency, and the advantages of self-reconfigurable robot which have various functions and strong adaptive ability, making it can achieve complicated tasks and operations which can not be completed by fixed-configuration robot, especially applicable to variable environment and complex tasks, such as space exploration, disaster rescue, leakage point detection, as well as to education and service fields in which researches are of both high practical and theoretical values to extend application background of this kind of robot.The structure of the unit robot is of high importance to the accomplishment of system self-reconfiguration as well as to the adaptive ability when each unit robot works independently, which requires it not only can be easily linked to each nearby unit, but also can conquer small obstacles independently. To achieve this we proposed and produced a new variable structure miniature unit robot, which uses the self-corrective pin-hole structure to improve the successful rate of self-configuration and error tolerance, the double-track envelope structure to strengthen the ability of avoiding overturning and adapting to the ground, and the variable structure achieved through controlling the angle between two tracks to increase its obstacle-crossing performance, which make it achieve a balanced point among overall dimension, butting ability and adaptability.Since its self-configuration is based upon real-time and accurate detection of attitude and miniature unit robot has serious demands in volume, power consumption and cost, we researched in integrated navigation of miniature unit robot system using low-cost MEMS inertia components, GPS for civil use and photoelectricity-code disc. Combined with requirements of double-track driving system and self-reconfiguration, we built the model of navigation system of the miniature unit robot and merged multi-sense information using extended Kallman filtering algorithm to achieve accurate combination navigation which combined dead reckoning and position updating, which lay foundation to complete its function and extend its working range.The key problem of mobile self-reconfigurable robot technology is to complete the changing process from discrete state to assemble state automatically and reversibly. For the difficult"discrete to assemble"state change process, we researched in dynamic self-reconfiguration strategies of multi-mobile-robot. We divided the process into three steps: aggregate, alignment and butt; using time consumption, power rest and system optimization as index, we separately proposed corresponding strategies: dynamic aggregating optimized model based upon system balance, alignment algorithm based upon predicting chasing and automatic butting strategy based upon sight servo.In the end, we assembled the electromechanical system of the variable-configuration unit robot, compiled control program and interactive program, finishing the experimental platform of the variable-configuration unit robot. Then we planned system experiments verifying the mobility and trafficability of both the integrated robot and unit robot, in which major verification experiments subjected on the automatic butting of nearby units and self-reconfiguration process from discrete state to integrated state, which proved that the design of variable-structure unit robot is reasonable, that the integrated navigation algorithm is effective and that the self-reconfigurable strategies of multi-mobile-robot is correct.