Rearch On Distinguishing The Error AIS Data At Inland Waterway

AIS(Automatic Identification System)is an equipment to assistant ship navigation.With the development of inland waterway transport,it has been widely used in the operation of inland river ships.The ship navigation devices are mainly composed of the radar,but AIS that as an assisting navigation equipment still work as an irreplaceable role.For example,when radar has a signal blind area on the turning of Channel,AIS could help it get the information of ships.When Ships transmit AIS information on very high frequency(VHF),its time interval is according to the provisions of international maritime organization.When the AIS information is transmitting,its signal strength gradually weakened.And then the interference of radio waves and obstacles will lead to information losing.In addition,affected by ship sensors and other hardware devices,AIS information ship send may become error data.That will affect ship navigation safety and regulation.To put forward effective AIS information screening method,ensure high quality of AIS data is great important for inland waterway transport safety.In order to explore a better method of AIS data screening,and statistic characteristics of error AIS data,main work includes:On the Base of characteristics of Wuhan section channel in the Yangtze River,AIS data acquisition platform was put up on the roofs of three pontoon boats near the Baishazhou Bridge,Wuhan Yangtze River Bridge and Tianxingzhou Bridge.The information collected transmit to the server through the network,and then parse the received data and store it in the SQL server 2008 database on the server.The data will be used in data screening.According to the state of AIS data reception,determined the object of study is AIS historical data in Baishazhou channel.Due to the ship track is affected by the Yangtze river water level,get AIS data is more advantageous to identify according dry period,flood level period and common water period.The three typical water level periods of a year are April,August and October.Obtain AIS data as sample from the database on the three month in 2014.The vessel track is different in ascending and descending,and course Angle is in a fixed range.So extracting huge amounts of AIS data from the database,then static course Angle data of Baishazhou channel,determine course Angle range in ascending and descending,subdivide the sample data in six parts.Build fuzzy c-means clustering model based on Euclidean distance and Mahalanobis distance,DS Evidence Theory model respectively on screening,then discriminate AIS wrong information and verify the applicability and effectiveness of the two methods.According to the statistics data discrimination,to compare the two methods of fuzzy c-means clustering,and then compare the better method with DS evidence theory on the results of discrimination.Then get a better AIS data discrimination method.This method will be used to identify the AIS information.Finally,static the characteristics of AIS error information base on space and time.