An Application of Fuzzy Forecasting Algorithms to Indoor Positioning Systems for Sensor Networks

• Main Authors: Ze-Hao Chen, 陳澤豪
• Other Authors: 李俊賢
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/nmc82u

碩士 === 國立臺北科技大學 === 電機工程系研究所 === 100 === GPS is one of the most common positioning system technologies in locating and navigating object. However, the main disadvantage of GPS is that the satellite signals may be easily blocked by buildings. This condition may result in the low positioning accuracy when the GPS is applied in indoor positioning. For this reason, other positioning technology such as infrared, ultrasonic, or radio technology with its unique physical characteristics has been employed to solve the problems encountered in indoor positioning. This paper focuses on implementing the fuzzy logic system to forecast the location of an object and combining the neural network to improve the accuracy of indoor positioning. In short, the low positioning accuracy in indoor positioning may be caused by instrumental errors, especially the effects by non-line of sight (NLOS). Essentially, NLOS results in errors when the signal is obscured during signal transmission. The purpose of this research is utilizing the forecast location algorithm based on the fuzzy logic system to reduce the NLOS error in indoor positioning. Furthermore, the procedure of this method is designed in the following steps. First, the line of position (LOP) algorithm is used to calculate the position of mobile node. Then, the coordinate of position calculated by LOP is incorporated into neural network to reduce positional errors. In addition to achieve forecast location algorithm, the various walking speeds of an elder is integrated into the fuzzy logic system to estimate the coordinate of mobile node on the next time. The simulation results indicated that the instrumental and NLOS errors were significantly reduced.