GNSS/SINS Integrated Navigation Algorithm with ARIMA Model Assistant Factor Graph

When working in the substation, the construction personnel may be near the dangerous area or stray into the non-working area for various reasons, which may lead to the injury of the staff or affect the normal work of the substation. Therefore, the real-time personnel location information in the substation is carried out to ensure the safety of construction personnel. To solve the problem of unavailability of GNSS due to strong magnetic field interference in a specific area of substation, GNSS/SINS positioning method based on Autoregressive Integrated Moving Average (ARIMA) model assistant factor graph is proposed in this paper. This method can calculate the GNSS information at the time of interruption after the GNSS signal is interrupted, model and optimize the pseudo GNSS location information, and use the factor graph algorithm to jointly optimize the pseudo GNSS information and SINS information. When GNSS works normally, GNSS data and SINS data are directly modeled and fused, and the factor graph incremental smoothing algorithm is used to fuse GNSS information and SINS information, so as to obtain accurate location information of construction personnel. This paper verifies the feasibility of the algorithm through the construction personnel positioning experiment. The experimental results show that the proposed algorithm can realize the positioning information prediction after GNSS signal interruption, and the location accuracy is better than 1 meter, meeting the accuracy requirements of real-time early warning.