DISTRIBUTED CONTROL OF MICROROBOTS FOR DIFFERENT APPLICATIONS

摘要

During the last years, the need for large and complex technical systems has become obvious. Examples are manufacturing cells, transport systems consisting of many vehicles, or robots working together to reach a common goal. As a consequence, the control architectures for these systems are not longer able to guarantee the well known properties of modularity, fault-tolerance, integrability and extendibility. On that account, the former centralized architectures are often replaced by distributed control mechanisms. During the last three years, the University of Karlsruhe has developed an infrastructure for distributed robot systems. This way, the problems of communication, coordination and cooperation between the system components, the task allocation, optimization and deadlock-avoidance are managed by the architecture KAMARA (KArlsruher Multi Agent Robot Architecture) itself. Using these concepts, the planning system for the Karlsruhe Autonomous Mobile Robot KAMRO was realized in a very short time. To prove the general use of the infrastructure, the control architecture for the microrobots MINIMAN is being developed at IPR by use of these new concepts. One robot consists of two manipulators driven by three piezoactuators, and it can move by use of three piezolegs. In the application field of these robots (medicine, biology and chip production), it is often necessary that the microrobots cooperate to perform difficult tasks. Due to the complexity of the system, the interactions between the robots, and the possibility of dead-locks during the task allocation, the KAMARA architecture was used to solve these problems in an efficient way. In this paper, the results of the work and some experiments with the new system are described.