Integrating decentralized coordination and reactivity in mas for repair-task allocations

Abstract

Task allocation is an important research area for multi-agent systems (MASs). In a large system of systems, multiple MASs are connected through the network, and decentralized coordination among MASs is vital. In general, it takes time to coordinate task allocations. However, when a task has to be done within a short time, it is necessary to start the task execution immediately. In this paper, we present a new task-allocation algorithm that reconciles decentralized coordination and reactivity. We consider scenarios where multiple causes of future agent failures are created simultaneously and consecutively, and if they are not removed by repair actions within limited times, some agents become out of order with high probability. In this paper, we show that the combination of decentralized coordination and reactivity significantly increases (more than doubles) the average numbers of successful repairs when the time available for decision-making and repairing is short.