Research On Scheduling Of Container Yard Cranes And Its Simulation

The competitiveness of container terminals is affected by many factors, including the geographical environment, allocation of port resources, information technology, and business management. Among them, yard is one of the most important links in the operation of the container terminal, which is the temporary buffer for the import and export of containers. As we know, the efficiency of yard operation has a direct effect on the customer service and the throughput of the port, while the efficiency of the yard is mainly decided by the work efficiency of yard cranes. Due to the random and dynamic arrival of the containers, the random time of completing the work and the complexity of dispatching yard cranes, how to allocate and optimize the given number of yard cranes is significant. As we know, the rational and efficient scheduling of yard cranes can reduce the time of container trailer waiting yard crane for service, increase the work efficiency of quay crane, and compress the time of container ship in the port.This thesis generalizes the existing system simulation modeling methods, and selects the method based on discrete event modeling. After several assumptions based on many researches on port, a multi-objective dynamic scheduling model using Nonlinear Mixed Integer Programming for yard cranes is established. To resolve the NP-hard problem concerning yard cranes allocation, a greedy algorithm is then deployed, which is realized by using the computer programming in java language environment. Subsequently, a yard cranes scheduling model is built up based on AnyLogic platform and three groups of the scheduling strategies is designed. Finally, this paper combines systematic analysis, qualitative analysis and quantitative analysis method to authenticate the feasibility and creditability of the model and the algorithm. To be specific, the main methods are as follows:firstly, by running simulation of the instance which is given initial configuration of yard cranes, it verifies the high efficiency of greedy algorithm; secondly, by comparing the results from different designed scheduling strategies, it shows the quality of them in the case of different container workloads; thirdly, by doing research on how to determine the reasonable number of yard cranes, it obtains the allocation scheme which is economic and efficient.The decision-making tools for yard cranes dispatching developed in this thesis can be used to solve the practical problems such as initial allocation, work sequence making and job scheduling of yard cranes repeatedly and rapidly, which can provide decision support for the yard dispatcher, and offer a new way for yard infortionization.Compared with previous researches in this field, this paper has following feathers:(1) A dynamic scheduling model for multi yard cranes is established based on a rolling time period;(2) A more comprehensive study for yard cranes scheduling is developed, which considers not only the efficient scheduling for yard cranes, but also the initial required number of yard cranes and the number of yard cranes in each container block, to optimize yard cranes scheduling in the whole process;(3) The built nonlinear programming for yard cranes scheduling problem is solved by combining greedy algorithm and simulation method;(4) The tool platform for yard cranes dynamic scheduling based on AnyLogic has good scalability, and we can solve the majority of similar resource allocation problems by changing a small part of the model. In practice, after entering the number of containers in each container block, selecting different total number of yard cranes and different scheduling strategies, and running the simulation, yard operator can opt and make the optimum scheduling plan for the actual situation.