Sailing Time Prediction Algorithm For Controlled Waterway Ship Based On The Routes Matching

Because of the complex operating environmental condition in controlled waterway,ships can only sail through one way under the command of signal tower. Currentcommanding methods are artificial based method and the automatic identificationsystem (AIS) based intelligent commanding method, both of which generate commandsby estimating potential time that ships may take from upstream control section anddownstream control section respectively to upstream limit mark, where the commandsystem assumes that all the ships sail through the same route. However, in practice, thechoice of routes depends on several factors including water level, flow rate and shipload, etc, thus the actual sailing time and the estimated time may vary significantly,which severely affects the accuracy of the commanding system and reduces thewaterway traffic efficiency. Therefore, how to improve the efficiency and safty ofcontrolled river waterways has important theoretical and practical significance.In this thesis, Lu Zhou waterway administration’s Shen Beizui signal station isused as the platform to explore a best route in line with current environmentalconditions from a large number of historical routes stored in the database, andfurthermore to study a sailing time prediction algorithm, and to improve the estimationaccuracy of ship sailing time, which could enhance the performance of the developedintelligent commanding system. The main contributions of the work can be summarizedas follows:①By collecting the ships information by making use of AIS and established thehistory database, analyzed the collected data to highlight the sailing strategies for inlandnavigation vessels.②Proposed a sailing time prediction algorithm for controlled waterway shipbased on the routes matching with detailed description.③In terms of routes fitting method, aiming to achieve a fitting ship route, theleast squares method, cubic spline interpolation and B-spline fitting method have beenstudied.④By extracting navigation feature point of the route to be fitted, we developedthe optimal matching route by using equivalence relation based fuzzy clustering methodfuzzy and selecting proper cut set.⑤The actual historical data are applied to simulate route fitting and route matching and the simulation results confirmed the effectiveness of the proposedalgorithm.