Geomagnetic indoor positioning method based on multilayer long short-term memory network

The invention provides a geomagnetic indoor positioning method based on a multilayer long short-term memory network. For the problem of how to combine geomagnetic sequence information with a recurrentneural network to improve the positioning precision, the method is divided into two stages of geomagnetic indoor database establishment and model learning positioning. Establishment of the geomagnetic indoor database is divided into two aspects, one aspect is collection of indoor geomagnetic track signals, the other aspect is establishment of the geomagnetic indoor database, and in the establishment process, a KNN and DTW fusion algorithm is adopted to solve a proper geomagnetic indoor database establishment form. In the model learning and positioning stage, a geomagnetic indoor database is learned from a short-term memory network, and a learned long-term and short-term memory model suitable for parameters can be used for positioning. According to the invention, the signal source only uses geomagnetic signals, in