Geomagnetic indoor positioning method capable of actively adapting to change of magnetic field

Abstract

The invention relates to a geomagnetic indoor positioning method capable of actively adapting to change of a magnetic field. The method comprises the following steps: training a position prediction model, measuring a geomagnetic sequence of a current walking position through communication equipment with an inertial sensor, and inputting the geomagnetic sequence into the trained position predictionmodel after filtering processing to obtain a predicted coordinate position at an initial moment. A pedestrian dead reckoning method is used for reckoning the current position of a pedestrian. When apart position of the environment magnetic field is changed; the changed magnetic field position is judged and updated through pedestrian dead reckoning with short time, high precision and high data updating rate; the problems that errors are accumulated and deviate from the actual position due to the fact that the pedestrian dead reckoning use distance is too long are solved, meanwhile, remeasurement modeling, modeling, training and the like do not need to be conducted on the whole magnetic field again, the positioning precision is high, and the model adaptability is high.

Description

technical field [0001] The invention belongs to the technical field of indoor positioning, and relates to a geomagnetic indoor positioning method that actively adapts to changing magnetic fields. Background technique [0002] Since the geomagnetic field is a spontaneous magnetic field generated by the rotation of the earth, compared with other indoor positioning technologies, geomagnetic indoor positioning has the natural cost advantage of zero deployment. Indoor positioning will still continue to work normally. However, measuring magnetic field data in buildings consumes a lot of labor costs, and at the same time, geomagnetism is easily disturbed by the external environment. [0003] In the context of the widespread application of artificial intelligence, a large number of indoor positioning methods use neural networks to process features and finally output positions. Facts have proved that the neural network has a very strong ability to extract features, and has a wide r...

AI Technical Summary

Problems solved by technology

However, because the geomagnetism is easily changed by the environment, when the neural network is applied to the geomagnetic indoor positioning, it is easy to appear in the changing area (the environmental magnetic field is changed due to the change of the placement of some items), and the positioning result has a large error, which leads to the drifting of the positioning point.

Images