Objectives In order to meet the needs of high-precision emergency positioning of low-cost terminals in disaster environments, we study the generation of grid-based enhanced positioning information and high-precision enhanced positioning methods of low-cost terminals in disaster environments.

Methods According to the characteristics of emergency positioning in disaster environment, a long-distance and large-scale reference station is used to generate BeiDou satellite navigation system（BDS） enhanced positioning information. At the same time, in view of the fact that the reference station network is destroyed and the error correction number cannot be generated, the atmospheric delay error of the client can be well eliminated by using the farther reference station to calculate the error correction number. Taking into account the calculation pressure of the client and reducing the server, on the basis of generating the undifferen‑ced error correction, the geometric distance between the virtual reference station and the long-distance BDS reference station and the satellite is fused to generate the BDS enhanced positioning observation value of the grid point. The enhanced positioning information of the grid point is broadcast by the data single communication method, which reduces the pressure of user data transmission, and can generate more dense grid points in the emergency positioning area according to the demand. The robust adaptive filtering model is used to improve the performance of emergency positioning in disaster environment. A quartile robust adaptive Kalman filtering method is proposed, which conforms to the low-cost terminal characteristics of disaster emergency environment. This method does not need to standardize the innovation vector. By using the quartile method to dynamically determine the threshold, the robust model is established and the varian‑ce expansion factor is obtained. According to the robust weight function, the weight of the observed value is adjusted reasonably, which can effectively eliminate the gross error and improve the positioning accuracy.

Results The measured observation data are used to verify and analyze the BDS enhanced positioning algorithm in disaster emergency environment. The results show that in the static mode, the planar accuracy and elevation accuracy of the low-cost terminal can both be controlled within 0.2 m, which is an improvement of 58.9% and 52.4% respectively compared to the conventional enhanced positioning algorithm; in the dynamic mode, the improvements in the east, north, and up directions of the low-cost terminal are 0.09 m, 0.14 m, and 0.26 m respectively.

Conclusions The experimental results show that the proposed method can achieve real-time dynamic centimeter-level high-precision positioning in disaster environment.