The Research On The Precise KINRTK Theory And Its Applications

It is well known that the pattern of conventional DGPS is that the data of fixed base station transmitted to rover station by radio. Most applications are met by the pattern, But the pattern can not meet the need of some kinematic users in some cases , for example, the voyages are formed into columns to ocean, the airplanes carried warship are landed in warship, the airplanes are formed into columns in air et al. For the DGPS data that the fixed base station transmitted cannot cover with largo range. If the pattern of conventional DGPS is still used in the applications, two problems are caused: Firstly, the time that the carrier cycle ambiguity is solved will be increased and the precision of the position will be decreased with the baseline increased; Secondly, the DGPS data will be costly, incredibility and even impossible. For the range of between the base station and the rover station will become longer and longer.To solve the problem of positioning relatively between the moving objects with high accuracy, we set up the reference station on the moving carrier, to solve the relative position of kinematic user to reference one by GPS carrier phase positioning On-The-Fly (OTF); It's named to On-the-fly GPS Positioning Relative to (?) Moving Reference (KINRTK). the solution of ambiguity , the repair of the cycle slip and the mathematical processing model et al are studied ,The main achievement of this paper are as follows:1. If several epoch data are used to solve the ambiguity by LS (Least Square)(the sampling rate of GPS receiver is commonly high, for example 1 OHz, when the receiver is used to kinematic measure), the invert of normal equation of GPS observation equations will be instable, Because the GPS satellites are high orbit, and the float solution of the ambiguity departured from the exact value because of the inevitably observation noise, So the right ambiguity is difficultly fixed. The normal equation of GPS observation equations and the veracity of ambiguity are improved by increased observation time before, So the efficiency of rapid position is decreased. For this situation, the method of Ridge Estimation, the method of spectrum correction iteration and the method of restricted condition are discussed to eliminate the ill condition of the normal equation; and the experiments results show, for single frequency GPS ,the initialization can be finished in sequence 5 epoches (the sampling rate of GPS receiver is 1s)and the right ambiguity can be achieved.2.The core problem of high accuracy kinematic to kinematic relative position is how to fast confirm the solution of carrier phase ambiguity. An integrate method of solving ambiguity was put forward on analyzed existed OTF methods, step solution is the basic idea of the method, the initial ambiguity was searched by the method with LAMBDA method firstly, and with combined method, experiments results show the method can improve the search speed.3. Sometimes two moving platforms do not apart far (less than 10km) in kinematic to kinematic relative position, For this situation, an new GPS OTF method was put forward for single frequency. If the routine algorithm of solution ambiguity was used, the formation of normal equation rank is large, thus the calculation burdened greatly, the algorithm can not meet the real-time need, but the method of QR decomposition is used, and then theambiguity parameter was apart from position parameter by matrix transformation;and the rank of solution ambiguity reduced greatly. So the method have the advantage of solving speed fast> easily calculating and on-the-fly ambiguity.4. It is well known that though Double Difference patter can eliminate most common error;the remains error of troposphere and ionosphere have influence on DD data;so the data is not white noise;but corruption. If the routine algorithm is used to process the data;there is a model error in its results. So the theory of Hm Filter is introduced;and its basic principle and robust were discussed.5. Three model of data processing were analyzed and compared;that is LS;Kalman Filter and Hw Filter,While analysing and comparing the three model of data processing;namely LS;Kalman Filter and Hn Filter;we can find that when y ->â–  oo;the recursion equation can simplify the Kalman Filter equation ,this shows that Kalman Filter is one special case of Hn Filter. Through the experiment contrast;we find that when the observation data has the character of white noise;three kinds of models are equivalent in precision. But when the observation data has the corruption;the Ha Filter can get the best precision and the LS takes second place;Kalman Filter can but get the worst precision.6. Cycle-slip is the peculiar problem of high precision positioning (high precision positioning must adopt the carrier phase observation,. and in nature cycle-slip is a peculiar problem of carrier phase measurement);so the detection and repair of cycle-slip is key question of GPS high precision positioning;and all methods of detection and repair of cycle-slip used in kinematic positioning are analyzed and state the advantages and shortcomings of these methods are compared by experiment results. On this basis;a new method of detection and repair of cycle-slip based on tri-difference which has a similar idea and process with the LS ambiguity resolution is introduced. Experiments results show that the new method has a better result in kinematic positioning.7. In GPS kinematic positioning;if the altitude difference of the base station and the rover station is large;then troposphere error would take considerable effect on positioning result;so the correction of troposphere error is necessary in high precision position. In past application;we mostly use Hopfield model and improved Hopfield model which are based on standard meteorological elements to calculate the correction. This paper puts forward a kind of a dynamic troposphere error correction model by utilizing the relation among temperature;atmospheric pressure;height;and the new model can correct the error according to the real-time height. The contrast results of some tests show that the new model can get better result.8. KINRTK is different to conventional DGPS whose precise position is known;so its theory differs from ones. It's discussed that single-point positioning on pseudo-range provides original coordinates;their precision and influence on KINRTK; then author draws the conclusion that it's same to integer ambiguity researching process and cycle slip disposal with each other; The only difference is that the initialed coordinate must be fixed and known in conventional DGPS. And the results of the practical data processing show that KINRTK algorithm's precision to cm level.