| The increasing number of bridges across navigable waters has led to a significant increase in the risk of ship-bridge collisions.In the past,passive anti-collision facilities could not effectively avoid ship-bridge collision accidents.Therefore,it is necessary to carry out research on active anti-collision of ship-bridge,reduce or even avoid shipbridge collision accidents,and ensure the navigable safety of ships in bridge area and the safety of bridges.According to the ship braking theory,the dangerous area of bridge area is divided and the simulation environment of bridge area is constructed.The identification model of abnormal ship behavior based on Bayesian classifier is established to realize the identification of abnormal ship in bridge area.On this basis,Kalman filter algorithm is used to construct a real-time prediction model of ship trajectory to track and predict the trajectory of abnormal ships in real time.Finally,an active bridge warning method based on the rapid rolling update of time sequence information is proposed to give realtime warning to ships with yaw,retrograde abnormal behavior or the risk of bridge collision in bridge area.Based on the warning method,a set of shore-based unattended ship bridge active collision prevention warning system is developed.The main contents include:(1)Dangerous zone division of bridge area based on characteristics of ship handling behavior.The braking mode of ships in bridge area is analyzed,and the calculation model of minimum safe distance of bridge pier under the influence of wind and current is established.Taking ship tonnage as independent variable,the minimum safe distance between the ship and the bridge pier is determined,and the classⅠ,Ⅱand Ⅲ danger zones of the ship in the bridge area are divided adaptively.(2)Identification of ship abnormal behavior in bridge area based on ship traffic flow historical trajectory analysis.The historical track of ships in bridge area is cleaned and clustered,the traffic flow characteristics of ships in bridge area is analyzed,and the typical navigation behavior of ships in bridge area is explored.A classifier is established according to the Bayesian classification algorithm,and the identification model of ship abnormal behavior in bridge area based on ship history data is proposed and verified.(3)The ship trajectory prediction model in bridge area is constructed.The Kalman filter method is used to include the influence of ship position error noise,and the position of the next moment is calculated according to the current ship motion information.The ship trajectory prediction model in bridge area was built based on the fast rolling update of target ship state information based on time sequence.The influence of time step on the accuracy of prediction results is studied,and the accuracy requirement of ship position prediction in bridge area is analyzed.The prediction time step is set according to the analysis results,and the real-time tracking and prediction of ship trajectory with abnormal behavior is realized.(4)This paper presents an active anti-collision risk warning method for shipbridges.The early warning scope and process of ship-bridge active collision avoidance risk are determined,and the overall framework of system design based on fast rolling update of time sequence information is constructed.The early warning model and algorithm are proposed according to each sub-module of the framework.Simulation experiments are carried out for typical Bridges of the Yangtze River and a practical system is constructed.The simulation results verify the accuracy of the proposed model and algorithm,and the running results of the practical system prove the reliability and practicability of the system design and research results.The proposed active anti-collision risk warning method for ship-bridges can be adapted to bridge areas with different traffic environments,and the practical system has been installed and tested on a bridge in the upper reaches of the Yangtze River which can adapt to various weather conditions,all-weather barrier-free,uninterrupted dynamic monitoring.The research results can provide theoretical basis,technical support and system solution for bridge active anti-ship collision warning. |