Research On Chengshantou Short-term Vessel Traffic Flow Prediction Based On Extreme Learning Machine

With the rapidly development of maritime traffic,China's coastal has become one of the busiest traffic regional in the world.The number of vessels sailing through China's coastal waters has gradually increased,the trend of maritime traffic accidents increasing brought up with traffic safety,marine environmental pollution and other issues at the same time.Short-term traffic flow forecast is one of the key researches of traffic flow prediction,timely and accurate short-term vessel traffic flow forecasting information is the key to guarantee the safety navigation,smooth flow and effective operation of traffic.Based on the summary of short-term traffic and vessel traffic flow forecasting model,the existing traffic flow prediction algorithms tend to have low accuracy,slow convergence,unstable performance,In order to improve this situation,the Extreme Learning Machine,which is widely used in road short-term traffic flow forecasting is introduced to forecast vessel short-term traffic flow in Chengshanjiao waters.Firstly,this thesis introduces the basic knowledge of traffic flow theory and the basic concept of ship short-term traffic flow forecasting,gives the evaluation index of short—term traffic flow forecasting model.Secondly,use the data software obtains data of Chengshanjiao from the MSA AIS website,and data filtering.Normalized collected data to avoid influence of data fluctuation on Short-term traffic flow prediction model based on ELM and improve the prediction accuracy.Then,based on the mathematical tool of MATLAB,establish the short-term traffic flow prediction model based on ELM neural network,and introduce the programming realization.Finally,apply the ELM prediction model established in this paper to forecast the short-term traffic flow in Chengshanjiao traffic precaution area.The results show that these models can predict well for the short-term traffic flow in Chengshanjiao.Furthermore,analysis the predict outcomes and errors indicate that ELM neural network model can effectively avoid the inherent defects of the traditional neural network,The application of ELM has the advantages of small error,fast operation time and high precision for short-term traffic flow prediction.Fulfill the requirements of vessel short-term traffic flow forecasting.