The Study On The Accuracy Of Shipborne AIS Information And The Establishment Of Information Correction And Error Prewarning System

In recent years, the rapid development of international economy has led to great development of international trade, and following the heavy freightage, the low cost and high efficient transportation has got much favor from the major trade companies. With the increasing quantity and tonnage of ships and the raising speed and traffic density, the world important navigation channels are getting more and more jammed, and marine perils take place frequently, which threatens the lives at sea and the marine environment. Following the more and more serious problems and the rapid development of navigation technology, the newly developed navigation and communication system AIS-Automatic Identification System came into the world, which is a solution for above problems.Since the launch of AIS, the system has got widely favored in navigation field for its great advantage, but it has its own defects, either. For example, static data may be inaccurate when input and may also be falsified in voyage, the accuracy of dynamic data from other sensors cannot be validated, and the accuracy of data form the sending equipment can't be analyzed at the receiving part, etc. If the accuracy of transmitted data could be judged at the receiving end, then those ships whose information doesn't meet the requirements could be put forward correction and prewarning, the function of AIS would be fully benefited. This thesis is committed to solve the above problems in the following tow aspects.Firstly, analyzing the accuracy of static data. The thesis fully analyzes the error of AIS static data, such as MMSI, ship's name, call sign and IMO number, proposes a method by the sample database built on the SQL Server 2008 database platform to search and analyze the accuracy of received AIS static data, and put forward correction and prewarning to the officer or VTS operator, etc.Secondly, analyzing the accuracy of dynamic data (mainly ship position data). With building Kalman filter mathematical model, the result of Kalman filter is got by MFC class library on Visual C++6.0 platform. Comprehensively considering the GPS error and actual test results could determine the threshold value of accuracy of position information. The warning system mentioned in the thesis warns the officer of the ships whose position errors exceeds the threshold value, which catches the officer's care and reminds officer to issue reasonable instructions to avoid collision.