Papers by Keyword: Multi-Agent

Abstract: The development of market economy has put forward new demands to the logistics. However, the traditional artificial decision has been unable to meet the demand of the market. Under the rapid development of Internet of things and information technology, the application of new technology in logistics vehicle scheduling makes it more reasonable, also, reduce the logistics cost. On the basis of traditional logistics vehicle scheduling process, the paper designs a framework of logistics distribution system from ontology view, combined with the application of ontology and agent in the logistics vehicle scheduling and material scheduling. Then, it sets up the vehicle scheduling system model based on the multi-Agent, to analyse the process of mutual cooperation between agents if they complete the task together. Meanwhile, the paper also establishes a communication mechanism based on the blackboard system, to further analyse the cooperation mechanism between Agents.

Abstract: Numerical control (NC) programming is the critical part for the NC manufacturing system. The NC manufacturing programming is the concentrated reflection of design of component-processing, which reflect directly the productive efficiency of NC manufacturing and the quality of products. This paper suggests that the Multi-agent as the theoretical basis and development platform to research the NC programming based on agent, in order to reach the autonomy of programming unit, the NC programming unit based agent as its the processing object, and self-organization of system. The NC programming based on Multi-agent can achieve the autonomy and functional maturation of each programming sub-unit, and can also achieve the coordination and cooperation among all the dependent units, therefore, the system processes the self-organization ability. One standard process template is created relay on this technology, in order to templatize and intelligentize the process flow and process data of NC manufacturing.

Abstract:

The cooperation is the most important for a multi-agent system (MAS). In order to research the cooperation, creating a multi-agent general cooperation model, that not only thinking about the cooperation solving way, but also thinking about the cooperation necessity and how to solve the conflicts, was a feasible method. However, the cooperation model might be inefficient. This paper suggested, based on the singular value decomposition (SVD), simplifying multi-agent cooperation data, and then improving the general cooperation model. According to the simulation results, the improved model had higher efficiency and can provide a more reliable tool for the study of the cooperation in MAS.

Abstract: This paper presents ideas that using the complex system theory to achieve combat effectiveness evaluation, and constructs an combat agent model and a potential field theory model, then take an example to accomplish the combat effectiveness evaluation based on multi-agent.

Abstract: For implementing the online control and simulation of TE process, a simulation and optimization platform is proposed based on multi-agent technology. Professional software such as G2, GAMS, Matlab, are wrapped as agents with multi-agent technology. Based on the simple object access protocol, it can control and optimize the online simulation procedure of TE process. Finally the practicability and validity of the proposed platform are verified through the application in the TE process simulation and optimization.

Abstract: This paper based on multi-agent technology, according to automobile safety airbag roll bag station production technology, at first adopt TS fuzzy neural regression network to distributed modeling for the controlled object, the methods of combining supervision learning and reinforcement learning are used, according to multi-agent external reinforcement signals and the value function of evaluate network ,using adaptive genetic co-evolution algorithm to optimize the action network, so can adapt to the mutational environment, and engineering application supply the proof about the effectiveness of the control strategy.

Abstract: To meet the demand of school's digitization, informationization, intelligent and practical,the design method of education model system based on Multi-agent is designed.We present the Multi-Agent cooperative processing model combining mobile Agent technology with the communication characteristics and modes of Multi-agent and the implementing technologies is researched in the education model.

Abstract: There are plenty of services on distribution communication network. Service classification and characteristics are given in details. Based on the proposed smart service-awareness model multi-agent Qos control model is analyzed. All of these are helpful to meet user requirements on distribution communication network.

Abstract: The question of metro safety is one of the hot areas of current research. Subway accident will not only cause a lot of casualties, but also caused widespread traffic jams, to which is very important to pay close attention. Through the analysis of potential safety hazard of metro interchange station and the distinguishing feature of the technology of multi-agent, this thesis pose a new application of multi-agent in emergency events of metro station.

Abstract: For the purpose of assisting emergency vehicles to move, the key is to put forward the application of Vehicle Operation Agent and Road Agent in sections to complete the dynamic route guidance process for an emergency vehicle combined with improved-Dijkstra algorithm. At last, by using Starlogo software and Matlab together, the paper completes the simulation of the emergency vehicle guidance process, to prove the effectiveness of the method.