Abstract: Based on the underground WLAN system in coal mine, a self-made wireless mobile terminal was used to scan AP nodes to obtain MAC address and signal strength information RSSI, and the data of environmental information such as temperature, humidity and gas concentration were collected at the same time, the data was transmitted through the network to the ground server. The linear least squares model was established by using multi-lateral location, and the underground tunnel was divided into several segments, the piecewise estimated weighted values were used to obtain the matrix of dynamic weighted values, the real-time positioning data of underground personnel was calculated by piecewise weighted least squares algorithm, and finally the environment database and personnel trajectory database of underground personnel were generated. Experiment results show that the average positioning error of the system is less than 1.88 m. By using the modified weighted matrix and the piecewise weighted least squares algorithm, the positioning stability and high precision of the system can be guaranteed when the electromagnetic environment is changed.