| Recent years, demand for indoor positioning is booming around people. The traditional GPS can achieve excellent positioning performance in outdoor environment, but the signal drawback of harding to penetrate building, surrounding multipath and NLOS interference make it can’t accomplish positioning effectively. To solve this problem, a large number of indoor positioning technologies emerged, such as infrared ray, Bluetooth, ultrasonic and RFID. However, these technologies have a negative performance on accuracy and power consumption. Under this background, this thesis studies the UWB(Ultra Wide Band) and MEMS(Micro Electro Mechanical System), trying to combine the two technologies to realize complicate indoor navigation.Study on UWB positioning technology. After mastering the principle of UWB positioning technology and common location algorithms, this paper takes a relatively system research on the key factors that affect the UWB positioning accuracy in indoor, including UWB sensor layout scheme and UWB signal penetrability. The test results show that, the localization precision of center region is higher than which in edge region; Among a certain range number of sensors, the more sensors, the higher is the positioning accuracy, but it contributes little to the accuracy when sensors exceed the range. In the aspect of penetrability, the UWB signal has a strong ability to penetrate the common materials such as wood, glass, plastic and so on, but has almost no penetrability to iron materials. Something need to special note is that human body has a strong absorption capacity to UWB signal.Research on pedestrian dead reckoning(PDR) based on motion model. The PDR needs exact step number, step length and heading information. This article designed a new gait detection method based on analysis of traditional gait detection. Aimed at the case of disturb steps emerged in gait detection, an effective rejection scheme was proposed. The test showed that the counting precision could reach 100%; the relative error of step length based on frequency model is below 10%. When determining the pedestrian heading, regards the short time integration of angular velocity as the criterion to charge the heading is changed or not, if it changed, name the former heading plus the heading alteration as the current heading, otherwise name the former heading as the current heading. This kind of method can not only avoid the accumulative error emerged among angular velocity integration, but also could wipe out the shortage of the poor real-time of the magnetometer. The final PDR overall test showed that the gait detection scheme and step length method are accurate enough, but still some error among heading determines.Finally, this paper proposed a hybrid positioning system combined UWB and PDR after a deep analysis on each of them. When the UWB signal is disturbed and can’t locate effectively, the PDR could provide a temporary auxiliary position result; conversely, the UWB positioning could correct PDR in a large degree when PDR drift after a long time working.For the system construction, this paper adopt MATLAB to compile code and process data. The experimental results showed that the fusion UWB and PDR for indoor application can significantly improved the stability and continuity of the system. |
PDF Full Text Request