Research Of Multi-objectctive Decision-making And Intelligent Optimization Of Containeter Terminal Layout Program

Container Terminal is an important node in the container transport system,Container Terminal construction projects need large investment and a longconstruction period. Terminal layout program involved many the economic andtechnical factors, while Terminal is a device serviced for the shipping companiesand cargo services. Improving production efficiency, reducing the ships stopped inport, enhancing customer service standards can stabilize the supply and improve thecore competitiveness of the port, this requires sufficient berths, yards andadvanced container handling equipment. But these measures may cause the initialinvestment too large, and low utilization of facilities and equipment will affect theeconomic efficiency of port enterprises; Because terminals with random productionprocess, discrete and uncertainty characteristics, therefore terminal layout problemis a multi-purpose, multi-factors of complex issues. This paper is about Containerterminal layout program of multi-objective decision-making and IntelligentOptimization, a number of new ideas and methods have been put award, the maincontents and results are as follows:(1) The PSO-BP neural network model of port container throughput forecastingwas established, the forecasted result of container throughput based the model candetermine the reasonable size of a container terminals;At the same time, variousforecasting methods was used to a certain container terminal, by comparativeanalysing the forecasting results, the results forecasted by PSO-BP model is moreaccurate;(2) A new mixed method was presented in this paper to determine the indicators'weight of valuation system, and the method is based on the Delphi method andFAHP method. Under this mixed method, we can get a weight adjustment quotiety ofindicators, accordingly, we can determine the ultimate right of all indicators, at last,a reasonable and workable comprehensive evaluation system was established;(3) A new terminal layout program evaluation method was put forward based on the discrete event system simulation. Through the simulation program running,different layout simulation outputs was transported to valuation system;(4) A statistics program was established based on DLL technology andembedded simulation system to realize the data exchange between system simulationand valuation system, accordingly, we can evaluate the layout program;(5) The layout program simulation and optimization models based on anon-groups iterative multi-objective Genetic Algorithms was established, anddislocation crossover operator was applied to improve the non-groups iterativeGenetic Algorithms;(6) In response to the initial simulation programs too few, a smart machinelearning model was put forward to solute the problem; The combination methods ofsimulation and optimization technology was applied in decision-making terminallayout of examples, the results show that the method has theoretical and practicalvalue.