The invention relates to a robot positioning method based on a wireless sensor network and a geomagnetic map. wireless sensor network nodes, during the movement of the robot, by activating A wireless sensor network node, use RSSI to obtain the observation model of wireless sensor network nodes and robots; then use PSO and multi-sensor information fusion to obtain the particle likelihood; at the same time obtain the weight of particles in the particle swarm optimization method; Resampling and fitness function judgment are used to obtain the position of the optimal particle, obtain a local geomagnetic map, and use the geomagnetic matching method to achieve precise positioning, and finally obtain the absolute position. The present invention is easy to associate data, no wrong association occurs during observation, and the robustness and accuracy of the network are improved; at the same time, a more realistic system state distribution prediction is obtained, which makes the positioning of the robot more accurate and reduces the calculation amount of geomagnetic matching. The real-time performance of the method is increased.