Abstract: In view of the taylor algorithm’s no convergent problem caused by the error between initial value and real value, a new combined positioning algorithm based on LSM and taylor series is proposed. Firstly, the measured values of time difference are processed by the LSM and then the results are used as the initial value of taylor algorithm, the final positioning result is provided through the method of taylor series expansion iterative. The acquisition method of measurement value of time difference is proposed, moreover, the affect of parameters on the performance of localization algorithm is discussed. Simulation results show that the algorithm not only has high accuracy, robust performance, but also overcomes the divergence of taylor algorithm.