This paper introduces a new control strategy for heterogeneous multi-robots systems dedicated to industrial logistic setups. This control strategy is based on both distributed intelligence and machine learning and inv...This paper introduces a new control strategy for heterogeneous multi-robots systems dedicated to industrial logistic setups. This control strategy is based on both distributed intelligence and machine learning and involves three parts: the rigid formation controller, the perception system and the path planner. Our controller is event-based and thus its control-coordination strategy can be self-adaptive and applied to real dynamic environment. During the navigating process, the multi-robots system derives the environment model, performs the path planning process that guaranties both the transportation constraints and the obstacle avoidance. For the validation, both simulation and real robot experiments are performed. The results show that the developed control strategy can be well used for realistic logistics applications.展开更多
Network localization is a fundamental problem in wireless sensor networks, mainly in location dependent applications. A common family of solutions to this problem is the range-based network localization. The resulting...Network localization is a fundamental problem in wireless sensor networks, mainly in location dependent applications. A common family of solutions to this problem is the range-based network localization. The resulting localization algorithms are noise sensitive and thus lacking in terms of robustness. Our contribution provides an algorithm which is robust to measurement errors. We propose an analytical tool to analyze the effect of range errors in the final location and use a distributed method to solve the noisy range localization problem.展开更多
文摘This paper introduces a new control strategy for heterogeneous multi-robots systems dedicated to industrial logistic setups. This control strategy is based on both distributed intelligence and machine learning and involves three parts: the rigid formation controller, the perception system and the path planner. Our controller is event-based and thus its control-coordination strategy can be self-adaptive and applied to real dynamic environment. During the navigating process, the multi-robots system derives the environment model, performs the path planning process that guaranties both the transportation constraints and the obstacle avoidance. For the validation, both simulation and real robot experiments are performed. The results show that the developed control strategy can be well used for realistic logistics applications.
文摘Network localization is a fundamental problem in wireless sensor networks, mainly in location dependent applications. A common family of solutions to this problem is the range-based network localization. The resulting localization algorithms are noise sensitive and thus lacking in terms of robustness. Our contribution provides an algorithm which is robust to measurement errors. We propose an analytical tool to analyze the effect of range errors in the final location and use a distributed method to solve the noisy range localization problem.
