Automatic parameter identification for distributedly placed modular robots

Describes a concept of a distributively placed modular robot system and a methodology of parameter identification of the system. The robot system consists of modules which have monofunction (e.g. sensor or actuator) and radio communication. Depending on tasks, the system is constructed by attaching modules to an object or in a working area distributively. Then the modules achieve a task by their cooperation. In our algorithm, the parameters are identified by executing motions with actuators, and observing the changes of the state with sensors. The motion which we call testing motion influences accuracy of identification because of sensing errors. Accounting for this, especially for the case of a mobile robot and a ceiling camera module, the design of modules' testing motion is made by evaluating the influence of the camera's sensing error and dead-reckoning's locating error. The effectiveness of the algorithm is shown in the simulation and a basic experiment.

[1]  Toshio Fukuda,et al.  The meaning of functional reconfiguration under the dynamic environment and system behavior , 1996, Proceedings of IEEE/RSJ International Conference on Intelligent Robots and Systems. IROS '96.

[2]  Hiroshi Mizoguchi,et al.  Robotic room: Symbiosis with human through behavior media , 1996, Robotics Auton. Syst..

[3]  Masafumi Hashimoto,et al.  Dynamic Object-Transportation Control Method by Multiple Mobile Robots , 1995 .

[4]  H.F. Durrant-Whyte,et al.  Kinematics for modular wheeled mobile robots , 1993, Proceedings of 1993 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS '93).