The invention relates to an RFID
indoor positioning system and method based on ANNs (artificial neural networks). The RFID indoor positioning method comprises following steps of carrying out positioning area
data acquisition in an off-line stage, pre-
processing original data, establishing a training
database and constructing an ANNs indoor location positioning model; carrying out
noise filtering and data normalization pre-
processing on the
original data by means of a
Gaussian filtering principle in dependence on the
signal statistical data characteristic; obtaining real time signals in the on-line stage, pre-
processing data, inputting the pre-processed data into the ANNs indoor location positioning model constructed in the off-line phase in real time, and outputting the location of the object to be positioned. The RFID
indoor positioning system comprises a passive RFID tag, an RFID reader and a
computer terminal. Compared with the prior art, the RFID
indoor positioning system and method are advantaged in that the
noise filtering, pre-processing and positioning precision is high, the
Gaussian filtering process calculating time is short, the
noise jamming resistance is high, the cost is low and the positioning real-time performance is high and the like.