Indoor Robot Positioning Using an Enhanced Trilateration Algorithm

Abstract

This paper presents algorithms implemented for positioning a wheeled robot on a production floor inside a factory by means of radio-frequency distance measurement and trilateration techniques. A set of radio-frequency transceivers located on the columns of the factory (anchors) create a grid with several triangular zones capable of measuring the line-of-sight distance between each anchor and the transceiver installed in the wheeled robot. After measuring only three of these distances (radii), an enhanced trilateration algorithm is applied to obtain X and Y coordinates in a Cartesian plane, i.e., the position of the robot on the factory floor. The embedded systems developed for the anchors and the robot are robust enough to establish communication, select the closest anchors for measuring radii, and identify in which of the grid zones the robot is located.

A decision tree‐based NLOS detection method for the UWB indoor location tracking accuracy improvement

“…However, in case of high noise values and NLOS signal path, the unobstructed first path F 1 can be buried in the noise . The calculation of estimated FPath i in NLOS condition is depicted in Equation : where F 2 , and F 3 are the first path amplitudes in multipath propagation. N is the preamble accumulation count value, and A is the predefined constant of 113.77 dBm for the pulse repetition frequency R f of 16 MHz or 121.74 dBm for a R f of 64 MHz.…”

“…N is the preamble accumulation count value, and A is the predefined constant of 113.77 dBm for the pulse repetition frequency R f of 16 MHz or 121.74 dBm for a R f of 64 MHz. In the traditional approach, we need manually compare the resultant of FPath i with the estimated RSL to identify LOS/NLOS condition . With the information of the channel impulse response CIR , the pulse repetition frequency used R f , and the received RX preamble symbols RX , the calculation of estimated RSL , is depicted in Equation .…”

“…In the traditional approach, we need manually compare the resultant of FPath i with the estimated RSL to identify LOS/NLOS condition. 9,10 With the information of the channel impulse response CIR , the pulse repetition frequency used R f , and the received RX preamble symbols RX, the calculation of estimated RSL, is depicted in Equation (8).…”

“…The common localization method is based on the trilateration algorithm . This algorithm requires three or more reference points to determine the location of an object.…”

“…The common localization method is based on the trilateration algorithm. [8][9][10][11][12][13][14] This algorithm requires three or more reference points to determine the location of an object. In the study of Musa et al, 15 we proposed another localization method, a dulateration-based indoor location tracking (Dilacak) algorithm, which used to track a moving object using only two reference points.…”