A Research On Optimization Of Logistics Enterprise Transportation Scheduling System Based On Multi-agent Technology

Transportation scheduling system is a matter of multi-factor, multi-level,large-scale and complex system; it is a distributed, nonlinear, time-varying stochasticsystem in different time and different spaces. For logistics companies, transportationscheduling cost accounted for the vast majority of the proportion of the total cost.Multi-Agent (Multi-Agent) technology has autonomy, distribution, coordination,self-organizing ability, learning ability and reasoning ability. We have introducedMulti-Agent technology to build a transportation scheduling system architecture. Dueto the logistics and transportation scheduling environment uncertainty, we apply fuzzystochastic modeling methods and intelligent optimization technology to solve fuzzymodeling and optimization problems for complex transportation scheduling systembased on Multi-Agent technology in uncertain environment. Therefore, reducingtransportation cost optimizes logistics and transportation scheduling for improving theeconomic efficiency has very important practical significances.First, according to the characteristics of the logistics enterprise transportationscheduling system, using relevant theories of multi-agent technology and modelbuilding method, we construct an architecture for logistics enterprise transportationscheduling system. The architecture is mainly composed of five independent Agentmodule, namely coordination Agent, transportation orders Agent, transportationscheduling Agent, vehicles Agent and road network information Agent. According tomodeling method based on single-agent architecture model and the different behaviorand function of each Agent, we have designed its internal structure model for theseAgents.Then, through the application of fuzzy stochastic modeling techniques and modeling method based on fuzzy membership function, combined with the actualsituation of logistics and transportation scheduling for transportation schedulingsystem architecture Agent transportation scheduling module, we constructed a fuzzytarget based on triangular fuzzy numbers transportation scheduling model andfull-fuzzy fuzzy constraints. By constructing fuzzy goals and fuzzy constraints andmore extreme ranges and membership functions, the whole fuzzy problem istransformed into model and re-use of assembly functions leads to the integration offuzzy membership functions and multi-objective fuzzy constraint. According to thetheoretical basis of the proposed ecological optimization algorithm Geneticalgorithms and basic operating procedures, the model of iterative optimizationsolution is built.Finally, according to the actual situation of transportation scheduling system, theabove method is applied to the modeling and optimization for a cement logisticsenterprises of transportation scheduling system to construct a full fuzzy programmingmodel based on triangular fuzzy number, and then fuzzy membership functionoptimization methods are used, combined with MATLAB software, using geneticalgorithm optimization iterations to solve the approximation optimal solution at acertain level of possibility for fuzzy goals and fuzzy multi-constraints transportationscheduling model which is built on the text. The results illustrate transportationscheduling model is suitable for logistics enterprises transportation schedulingrequirements, and the use of genetic algorithms to optimize the resulting solution ismore feasible and practical. The scheduling scheme will be optimized andtransportation cost will be reduced so that economic efficiency of enterprises can beenhanced.