A new RAIM algorithm based on multivariate cumulative sum and its improvement

“Snapshot” RAIM algorithm based on parity space can detect a larger bias fault in one step by using redundant BeiDou satellites, but it does not have the ability to detect a medium cumulative fault. First, a method that uses multivariate cumulative sum is designed to detect receiver positioning equations parity residual, and it derives multivariate cumulative RAIM (MCRAIM) alarm threshold, and then we compared the performances of the two kinds of algorithm with RAIM and MCRAIM. Secondly, the RAIM and MCRAIM are combined to be improved to a new algorithm that can detect not only a big mutational mean shift in one step, but also moderate continuous mean shift, and the Calculation formula of detecting Function and threshold will be given. Finally, the average run length indicator and fault detection performance of RAIM–MCRAIM combination algorithm are analyzed, and it proves that a combination of the two algorithms have more advanced features, it can make up for the defects of using either RAIM or MCRAIM individually.