Simplified Solidity Square Root Cubature Kalman Filtering About BD Navigation Receivers

This paper focuses on reducing the dynamic positioning of the algorithm complexity,reducing the amount of calculation,improve computational efficiency,solving the dynamic positioning model mismatch,use a single large model errors,issues affecting the state noise and measurement noise of the dynamic positioning are non-Gaussian white noise.This paper bases on the three issues' s research,the main contents and innovations are as follows:1.The satellite-based navigation of the carrier state equation is linear,square root volume and robust Kalman update its volume when the point after passing through the state transition matrix and zero weighting,you can use the standard KF algorithm update,measurement update the process is still using SCKF,the square root of a simplified type sound volume Kalman algorithm(SSCKF),the algorithm is designed to solve the dynamic navigation positioning calculation volume,low efficiency problem,simulation and experimental data indicate quite SSCKF and SCKF accuracy,and solver time reduced by 25% compared with SCKF algorithm can effectively reduce the computational complexity,improve efficiency of the algorithm2.Based on the SSCKF,combined with variable dimension IMM thought,proposed a simplified type sound Cubature Kalman square root of variable dimension IMM algorithm for conventional interactive multi-model set coverage is not comprehensive and the every model due to excessive competition,etc.This paper interacts with different dimensions models,such as uniform acceleration and a uniform model,with parallel filtering,both measured by residual calculate the corresponding likelihood function,updating the results of the two models share the weight of the filter,the final weighted and as a result the entire output of variable dimension model,the state input values the next time the sub-models do not use filtering results on its own at a time,but with the overall output of variable dimension interaction model transformation matrix obtained by multiplying dimension value,so it can ensure that all the time state of the input worth accuracy.3.For dynamic navigation,the state noise and measurement noise generally exhibit characteristic non-Gaussian white noise,Gaussian sum Simplified SCKF algorithm is raised to analyze the dynamic navigation kurtosis value from the pseudo-noise measurement and correlation coefficients obtained which presents non-Gaussian white noise of conclusion,if the actual movement will remain non-Gaussian noise forced as Gaussian white noise to deal with,it will affect the filtering accuracy,with multiple Gaussian white noise as a child Gaussian term,using its weight to approximate non-Gaussian white noise,colleagues limit the total number of non-Gaussian term each time to ensure that each time solver efficiency,sports measured data proved that the The research content of this article has important guiding significance for the integrity of algorithm can effectively suppress the influence of non-Gaussian white noise,and enhance the stability of the algorithm filtering accuracy.