Research On The Post-Processing Software Of GPS Kinematic Positioning

Real-time kinematic positioning (RTK) is one of the most important applications and research fields of GPS, used to obtain the real-time position information of any moving target by building a wireless communication link for data transmission between the reference station and rover. However, in some applications of GPS kinematic positioning that we have no need to get the real-time location information, we can bring the collected data back to the indoors for processing, called Post-Processing of kinematic positioning. Also, the use of the precise ephemeris and better error correction model makes the Post-Processing has certain advantages in terms of positioning accuracy. Track is the most commonly used software for Post-Processing of kinematic positioning, however, it runs on UNIX or LINUX operating system, it’s installation is complex and do not have visual interface.For the above needs and shortcomings, this thesis studied the source code of Track, learned the theory of single epoch GPS positioning, then developed a Windows operating system based prototype Post-Processing software of GPS Kinematic Positioning with friendly interface. We use some measured GPS data for program testing, analyzed the positioning accuracy and the ability of ambiguity resolution of our prototype program. We also discussed the influence on positioning accuracy casued by different baseline length, and tried to find out the relationship between the length of observation time and accuracy.The cores of our prototype program are the integer ambiguity resolution and the single epoch Kalman filter. It includes the following functional modules:File reading module, Parameter setting module, Ambiguity resolution module, Kalman filter module, Other auxiliary function module like report output module, transformation of coordinate system, transformation of time system, display, database, etc.We used eight groups of measured data to test our program, these data are collected from either IGS or SICHUAN CORS. The first four sets of data share the same sampling rate(1HZ) and time of observation (1hour), their baseline length are5km,42km,123km and235km, used to analyze the capability of ambiguity resolution in our program and position precision of different baseline length. The later four sets of data share the same sites and starting time, their length of observation time are15min,30min,45min,60min, used to analyze the relationship between the time of observation and accuracy.We extracted the result of ambiguity and coordinate information calculated by Track and our prototype program, compared their mean value and standard deviation of coordinate results in all epochs to each other, and also compared the mean value to the theoretical value, finally we get the conclusions as follows:The resolved integer ambiguity of our prototype program is roughly consistent with the result calculated by Track, some of the ambiguity has a diffrence of1-2cycles. Using Track to deal with1HZ GPS data, the positioning precision of short baseline can reachs millimeter; For long-short baseline, the accuracy of the horizontal direction reachs millimeter, and the accuracy of the elevation accuracy reachs1-2cm; for medium-long baseline, the accuracy of the horizontal direction can reachs1-2cm, elevation accuracy reachs2-3cm; the positioning accuracy of long baseline up to centimeter level.Our program generally realized the post-processing function of kinematic positioning, the positioning accuracy of short baseline can reachs1-2cm; for long-short baseline, the accuracy is2-3cm; the accuracy of medium-long baseline reachs cm level; the accuracy of long baseline is decimeter level. The horizontal accuracy is superior to the elevation accuracy in Post-Processing. The shorter the baseline, the higher the precision. The positioning accuracy is hard to guarantee when the observation time is insufficient, over30minutes of initializing observation and15minutes of new satellites makes the accuracy reachs to cm level. Once the length of observation time meets the demand, the positioning precision can not be improved by extending the time of observation.