Study On The Route Identification And Simulation System Of Wangjiatan Reach In The Tail Of Three Gorges Reservoir

During the decline period of the Three Gorges Reservoir,some river sections in the changing backwater area will have shallow,dangerous and urgent obstruction flow patterns,which seriously affects the safety and efficiency of ship navigation and inhibits the full play of inland navigation benefits.The channel pattern of the Three Gorges Reservoir’s obstructed navigation section makes it difficult to be completely resolved by the huge amount of engineering such as mountain opening,rock splitting and restoring deep ponds,and the scheduling and water level fluctuations in front of the Three Gorges Dam cannot be adjusted significantly,so The obstruction characteristics of the Kuwei River section will exist for a long time and affect the smooth navigation of the ship.Based on the background of the Wangshangtan section of Changshou,this paper studies the route planning method under the background that the existing channel scale cannot be effectively improved to improve the navigation efficiency of ships.At the same time,in view of the fact that the automatic identification system cannot effectively collect ship navigation information,a track extraction method based on the principle of computer vision is proposed as a supplement to the original data of route planning.The main results of the paper are as follows:(1)Constructed a database of historical ship routes in the Wangjiatan section.By analyzing and studying the navigational conditions of the river section,the grouping conditions of the channel conditions combining the ship’s ship type and real-time hydrological information are obtained,and the route database is constructed based on the historical ship trajectory data.The saving of route data in the form of a database not only solves the problem of insufficient capacity under the traditional storage method,but also facilitates the exchange of data between different terminals.(2)Optimized background difference method suitable for inland river motion detection.The ship target detection test shows that the optimized background difference method can more accurately detect the ship’s moving target,reduce the interference of environmental factors such as light changes,and can accurately obtain the ship’s moving target in the video under the image coordinate system through the centroid tracking.Pixel location.(3)The track extraction method under video image calculation is proposed.Aiming at the situation that the automatic identification system can not effectively collect the navigation information of ships,a method of ship track extraction based on the principle of computer vision is proposed.The camera parameters calibrated by the Zhang Zhengyou calibration method are combined with the perspective transformation of the measurement area and the channel image to calculate the geographic location information of the track point.(4)A real-time dynamic route planning method for ships in the Wangjiatan section is proposed.Based on the DBSCAN clustering algorithm,the ship’s turning concentrated area is extracted and the entire channel is segmented based on this,and the empirical optimal route under the channel condition is optimized and calculated by the method of segment combination.The results show that the planned route has a shorter navigation time than the historical optimal route,which can improve the navigation efficiency of the study section.