Research On Fault-Tolerant Scheduling Based Road Mechanical Multi-agent Hybrid System And Simulation

With the development of the scientific technology and the increase of theproduction requirement, people pay more attention on new generation ofthe road machinery cluster. The intelligentized and systematic engineeringmachinery of cluster is not only an important trend of future developmentof road machinery, but also a focus of the researches to solve some seriousproblems in road constructions nowadays, such as the illogicalconfiguration of machinery of cluster, the lack or overstock of material,the low efficiency of construction, the real-time fault-tolerant scheduling,and the high intensity of work. Furthermore, the latest achievement ofartificial intelligence, multi-agent is applied to intelligentize theengineering machinery of cluster. This will let the cluster own moreagility, self-determination, and reliability than the cluster using groupcontrol. The technology of real-time fault-tolerant scheduling is the bestmethod to improve the fault-tolerant capability of the engineeringmachinery cluster, and avoid serious loss. This will provide an abundantbasis for raising the level of the intelligentization and systematism ofengineering machinery of cluster in our country, and later, the level of theintelligentization and systematism will reach to the international top level.The main research contents of this thesis are listed here:1) For the first time, the technology real-time fault-tolerant scheduling isapplied to hybrid control of the intelligentized machinery of clusterThis paper takes the intelligent construction machinery groupsmulti-agent system for example to introduce a new technology offault-tolerant scheduling. It predigests and abstracts large-sizemachinery group multi-agent into three layers of agent hiberarchy, anddefines main function of each single agent. Moreover, combining withthe intelligent fault diagnosis system the paper designs the appropriatetolerance technology, which takes main control agent as the core ofmulti-agent system. This technology significantly decreases thenumber of required agents by integrating passive and activeduplication, and improves the processor utilization by exploiting themaximum available processing time for scheduling aperiodic tasks,which is left unused by periodic tasks. Simulation results show theeffectiveness.2)Apply MAGE distributed agent platform to simulate supply chainsystem of the intelligentized machinery of cluster.Combining with the latest achievement of artificial intelligence andapplying a set of theories of agent software, this paper analyze anddesign concrete supply chain system of intelligentized machinery ofcluster, and make the simulation.