Research Of The Positioning For Indoor Positioning And Tracking Based On UWB

With the increasing demand for real-time high-precision indoor positioning, the traditional indoor location technology, such as infrared, bluetooth, ultrasound, RFID, have not been meet the needs of people in terms of the power consumption and the positioning accuracy. Ultra-wideband(UWB) technology compared to narrowband technology has many advantages: high transmission speed, low power, anti-interference ability and penetration. It can achieve good positioning performance in a few obstacles, but if there are many obstacles between the target and the detectors, the performance of orientation would decrease. Therefore, the study on the way that significantly reduce the non-line of sight(NLOS) transmission errors has far-researching meaning.This paper studied the characteristics of ultra-wideband technology, positioning system function of each part, the common types of localization and tracking algorithm; Secondly, the paper studied the UWB signal transmission channel model of indoor, simulation and analysis the characteristics of IEEE 802.15.4a channel transmission models at different short distance scenes; Finally, in order to achieve high-precision positioning for the objectives in the unknown scenarios, the paper has been divided into the following two aspects:On one hand, considering the target is stationary: the detection in an environment for LOS and NLOS coexistence, N-CHAN algorithm was designed, which not only retains the advantages of CHAN algorithm: when the noise is Gaussian distribution, it has a smaller amount of computation and high positioning accuracy, but also achieves high accuracy positioning in non-Gaussian noise scenarios. It can identify LOS and NLOS transmission and reduce errors based on the different characteristics of signal transmission in LOS and NLOS, the simulation proved that the positioning accuracy is significantly better than CHAN algorithm; When the path between target and the detector is NLOS within the detection area, using geometric relationship of TOA and the centroid Taylor algorithm, N-Taylor algorithm was designed, which can detect which path has the larger relative error, and at the same time can effectively eliminate the NLOS error and solve the problem of divergence. Simulation results show that the algorithm is obviously superior to CHAN-Taylor algorithm and Taylor algorithm in positioning accuracy, the number of iterations and the convergence.On the other hand, considering the target is moving: analyzing the characteristics of current various types of tracking algorithm, comparison of various algorithms by simulation shows: PF algorithm is most suitable for this article. combined with the actual demand, JZ-PF algorithm was designed. Simulation results show that the algorithm can effectively reduce tracking error and has good robustness under different sizes of Rayleigh distribution noise.Finally, considering whatever the object is stationary or moving, FFP algorithm was designed, which uses the RSSI measurements and could identify the target whether is moving, whether the path between the target and probe is LOS, simulation show that the algorithm can locate in a complex indoor environment and the accuracy could reach centimeter level by MATLAB.