Research On Port Container Truck Dispatching Based On Multi-Agent

Container trucks(referred to as CTs)are important facilities in the port dispatching and unloading operation process,and important parts of port operation scheduling,which have important impacts on port loading and unloading efficiency.The traditional truck loading and unloading operation is carried out in a fixed loading line according to a static plan.This method leads to inefficient work and inefficient usage of the truck.In port loading and unloading operation scheduling,the optimization of real-time truck scheduling is an important part of improving operation efficiency.This paper focus on the improvement of the efficiency of the CTs operation,and reviews relevant researches on the real-time dynamic card scheduling of the territory,which has very important practical significance and application value for improving the comprehensive competitiveness of the port.Firstly,through the analysis and summary of the characteristics of the existing port CTs scheduling,the functions and problems in the dynamic real-time CTs scheduling are clarified,so as to "loaded operation"(i.e.both flows of CTs from and to the shore is loaded).Aiming at the loading and unloading operation of the container,a multi-agent system design idea is introduced.Combined with the ant colony algorithm,a distributed parallel CTs scheduling algorithm model is proposed.Secondly,according to the characteristics of its operation,the framework model and problem model between multi-agent systems are designed.In the function design of Agent,the intelligent algorithm is split and put into the intelligent module of multi-agent body for the purpose of parallel computing.Through the information collaboration between Agents,the intelligent algorithm is optimized.This multi-agent parallel computing method is used to realize the scheduling operation,which greatly improves the processing efficiency and can improve the effectiveness of the CTs scheduling.In the solution model of intelligent algorithm,the residual coefficient of pheromone and the critical value of pheromone are designed,and a simulated annealing algorithm is introduced in the update selection,so that it has a larger solution space in the early stage and solves the better solution.Finally,the intelligent modules designed in the multi-agent body are realized.Through the distributed parallel computing of multi-agent body,a complete solution for the real-time scheduling problem of the CTs is proposed.In this paper,the proposed solution is applied to the actual operation of port loading and unloading operations,and the centralized and distributed computing efficiency is compared and analyzed.It is proved that the multi-agent distributed computing has a good improvement on the timeliness of the card scheduling,and it also confirms.The algorithm can achieve the goal of "loaded operation ".Through the intelligent optimization algorithm design of distributed parallel computing,the real-time dynamic card scheduling and the "loaded operation" goal in the process of truck scheduling are realized,which greatly reduces the stagnation time of port ships and the empty driving of trucks.The rate realizes the rational use of port resources and has practical significance and application value for real-time dispatching of port cards.