| Summary: | 碩士 === 國立臺灣科技大學 === 機械工程系 === 92 === This thesis takes advantage of the relative motion sensor to assist indoor RF positioning system, and develops an indoor positioning algorithm based on recursive least squares method. Compared with the conventional positioning method based on known signal strength/distance models between the base stations and the receiver, the method proposed in this thesis used additional relative movement information and can be applied to applications where parameters of the signal strength/distance models are unknown or time-varying. Furthermore, the proposed method can improve the indoor positioning accuracy to be better than 1 meter.
The factors of the positioning errors are investigated through a series of computer simulations. The concept of Dilution of Precision(DOP) are derived to characterize the effect of indoor movement path choice to the overall system accuracy. Based on the DOP index, the path corresponds to the best positioning accuracy can be obtained.
Experiment results show that the relative motion sensor aided indoor RF signal strength positioning system can provide better positioning accuracy compared with conventional indoor position method based on known (pre-determined) signal strength/distance models. Furthermore, this method provide better RF signal noise immunity and more robustness for indoor positioning application. The experiment results also indicate that the factors effected the RF signal measurement accuracy include the orientation of antenna as well the design of RF signal strength measure equipment.
|
|---|
