Ship Pipe Workshop Scheduling Problem Based On Multi-Agent System

In the process of shipbuilding, pipe manufacturing is an important link in the shipyardproduction organizations. Pipe manufacturing workload is relatively large, especially in theconstruction of the multi-ship, the more aggravated the pipe shop processing tasks, resultingin a pipe shop has become a bottleneck in the shipbuilding process, thereby increasing thelevel of production of the pipe shop for the entire shipbuilding has great significance. Pipeworkshop scheduling system is a vital part of the workshop production management, so itdirectly decided whether workshop can complete machining task through the reasonable useof limited manufacturing resources in a designated time, and the scheduling program affectsthe production level of the workshop in some extent. However, the majorities of ourshipping companies’ pipe processing plant is still used traditional mathematical models, orrely on the experience of the scheduling staff to production scheduling. In this condition,scheduling staff work volume, low efficiency, and due to larger human disturbance factors,slow dynamic response, low level of use of enterprise information; complete theoptimization and integration of system very difficultly. So, research of pipe workshopscheduling problem has great significance.In this paper, according to domestic and international results, proposed schedulingsystem based on Multi-Agent to solve the pipe workshop scheduling problem, andestablished pipe workshop scheduling model based on Multi-Agent. First of all, through theanalysis of theory of Multi-Agent system and pipe workshop production organization andscheduling status, constructed a scheduling system model based on Multi-Agent. This modelconsidered the characteristics of the pipe processing organization fully. Finally, according tofuse optimization algorithm which can better solve static scheduling problem andnegotiation mechanism based on the contract net, designed a hybrid negotiation mechanism,which prompted system no matter what state get more optimal scheduling scheme. Onchoice of the optimization algorithm, article selects the particle swarm algorithm to improvethe performance through joining the mechanism of variation and dynamic inertia weight.In order to verify the validity of the proposed scheduling system developed based onMulti-Agent pipe workshop scheduling system. First, combined actual managementcondition of the pipe workshop and characteristics of Multi-Agent system. Then, discussedaspects of the system requirements analysis, business model, static model, dynamic model, architecture and database design. Finally, developed pipe workshop scheduling systembased on Multi-Agent, and on this basis, verified the superiority and effectiveness of pipeworkshop scheduling system based on Multi-Agent through an example.