Research On Traffic Flow Prediction And Ship Lock Scheduling Model Of Inland River

There are many rivers and abundant inland river shipping resources in Zhejiang province.With the rapid development of economy,its inland river shipping flow is also increasing,which exceeds the navigation capacity of the originally ship lock.Generally,congestion is due to the fact that the number of ships exceeds the original design capacity of inland waterway.Congestion is more concentrated in the lock location.In order to reduce the frequency of traffic congestion in the inland waterway of Zhejiang province and improve the environment of inland river transportation,it is necessary to study on the planning and scheduling.Traffic flow prediction of inland river provides the basis for inland navigation planning and dispatching.Therefore,it is necessary to study river vessel traffic flow prediction.The expansion or construction of the ship lock requires a large amount of manpower and material resources and occupies substantial social resources.Compared with it,and started from the perspective of scheduling management,the improvement of navigation capacity can reduce the resource occupation and have higher economic and social benefits.Based on the analysis and comparison of various traffic flow prediction methods,this paper establishes the GM-BP-Markov combined prediction model.Firstly,according to the characteristics of small capacity data sample,the GM(1,N)model is established with Grey Theory.Secondly,in order to improve its prediction effect on non-linear data,BP neural network is used to optimize it,and a GM-BP prediction model is established.Finally,in order to further improve the prediction of the GM-BP prediction model for the data with volatility,the GM-BP-Markov combination prediction model is established.The combined prediction model can predict the data accurately with the characteristics of "small sample,non-linearity and volatility".The joint scheduling model of upstream and downstream lock is also established in this paper.The model considers two sub-targets:average sluice space utilization rate and average sluice time,and classifies the sub-targets based on the number of passing ships.When solving the model,the realization process of each constraint condition is introduced,and the arrangement part of the gate chamber is described in detail.To verify the feasibility and validity of the model,MATLAB is used to simulate the model and the results show that the model has certain feasibility and validity.The GM-BP-Markov combined prediction model of ship traffic flow established in this paper is of certain reference significance for data prediction in other fields,which have the characteristics of "small sample,non-linearity and volatility".The joint scheduling model of upstream and downstream ship locks established in this paper can be a reference for improving the scheduling optimization of ship locks in Zhejiang province or other similar inland waterway,and can also provide certain theoretical basis for future research on similar problems.