基于扩展合同网的协同故障诊断任务分配研究

RESEARCH ON TASK ALLOCATION OF COOPERATIVE FAULT DIAGNOSIS BASED ON EXTENDED CONTRACT NET

任务分配作为协同故障诊断技术的重要环节,任务分配机制的优劣直接影响到诊断效率和准确性。针对现有合同网协议存在分配效率低、任务量大时易造成单节点过于忙碌等问题,提出一种基于扩展合同网的协同故障诊断任务分配机制。首先利用Hash算法建立子故障任务与对应分配路径的分配知识库,为任务分配提供动态分配经验;其次将Dijkstra算法与合同网协议相结合,通过权衡故障诊断能力与路径权值来选择"性价比"高的任务分配节点。算法仿真表明,扩展合同网任务分配算法减少了通信量、提高了分配效率。

Task allocation is an important part of collaborative fault diagnosis technology,whether the task allocation mechanism is good or not directly affects the efficiency and accuracy of the diagnosis. For existing contract net protocol,the low allocation efficiency and the huge task load could easily lead to the problem such as over-busy single node.Thus,a task allocation mechanism of collaborative fault diagnosis is proposed based on extended contract net. Firstly,using Hash algorithm to establish allocated knowledge base which links sub fault tasks with the corresponding distribution path,providing expert distribution experiences for task allocation. Then,combining the Dijkstra algorithm with the contract net protocol,the task allocation node of high ″ cost-effective″ is chosen by balancing the capability of fault diagnosis and the path weight. Simulation results show that the extended contract net task allocation algorithm reduces the traffic and improve the allocation efficiency.