Research On Kalman Filter Algorithm In GPS Precise Point Positioning Using Un-differenced Phase

Precise Point Positioning (PPP) uses un-differenced dual-frequency pseudorange and carrier phase observations along with 1GS precise orbit and satellite clocks, for stand-alone precise geodetic point positioning (static or kinematic) with centimeter precision. Kalman filter is an extremely effective and versatile procedure for combining nosiy sensor outputs to estimate the state of system with uncertain dynamics. Kalman filter algorithm is unbiased and minimum variance which is better than the least square root estimation method in GPS dynamic system.It is discussed how to linearize the observation equation, how to select the state covariance matrix value based on the GPS un-differenced carrier phase precise point positioning's measurement and state equations. Standard Kalman filter is the same with linear observation and state equations. The extended recursive Kalman filter is suggested for debasing linearization bias for the unlinear measurement equations. To estimate ambiguity value together with position and velocity, a Kalman filter with ambiguity parameters is set up.Along with the Kalman filter processed, divergence will happen for the limited computer word or the filter model error. Different methods are brought forward for the radiation causes. Sage Adaptive filter and Kalman filter with adaptive factor method and UD decomposing algorithm are proposed and programmed. The second algorithm has strong detective and characters of simple computation.Physical systems are often subjected to unexpected deviations or failures such as blunders in the measurement model or errors in the dynamic model. These deviations will inversely invalidate the results and any conclusions based on them. It is therefore important to have a method in order to maintain an accurate and reliable solution. The innovation-based (predicted residuals) quality control approach is applied to detect the blunders in GPS navigation and kinematic positioning.All the algorithms should be realized in GPS user DSP chip. Data analysis center sends precise satellite and clock information in a certainty format to users all in region by GSM or GPRS communication method. In order to retrench the data quantity, eight step Chebyshelv polynomial function parameters and ls rate precise clock corrections are broadcasted. User receivers can get in-viewed satellites' three dimension position and velocity with these information. And user position and velocity will be estimated with the precise satellite information and un-differenced carrier phases and pseudoranges.