Abstract: In order to solve the problem of low positioning accuracy in the severe non-line-of-sight (NLOS) substation environment, based on the time-of-flight (TOA) ranging method, a positioning algorithm combining Taylor series and unscented Kalman filter is proposed to improve positioning estimation accuracy. The algorithm first improves the initial value of the Taylor algorithm, the threshold filtering and weight calculation of the position estimated value are solved by the weighted least squares method (WLS) to reduce the number of iterations while ensuring the positioning accuracy. Then, for the problem of unstable prediction of the lower label coordinates in the NLOS environment, the Unscented Kalman Filter (UKF) algorithm is introduced for de-noising, and the estimated value of Taylor series is used as the observed value of the UKF algorithm to improve the estimation accuracy of UKF prediction. The test results show that the root mean square error of this method is reduced by at least 5.81% under the static experiment. Under the dynamic experiment, the average positioning error of path tracking can be reduced by more than 37.5%.