Improvement of Indoor Positioning Accuracy of Ultra-wide Band Sensors by Using Big Bang – Big Crunch Optimization Method

Abstract

Ultra geniş band teknolojisi, birçok iç mekân konum belirleme sisteminde başarılı çözümler sunan, diğer yöntemlere kıyasla daha iyi performans gösteren, gelişmekte olan bir teknolojidir. Bu çalışmada, ultra geniş band (Ultra Wide Band-UWB) sensörler kullanılarak bir iç mekân konum belirleme sistemi geliştirilmiş ve kullanılan ek algoritmalarla, standart donanımların sağladığı doğruluk düzeyi arttırılırken aynı zamanda ortalama hatayı azaltmak hedeflenmiştir. Bu amaçla Büyük Patlama-Büyük Çöküş (Big Bang-Big Crunch veya BB-BC) optimizasyon yöntemi deneysel iç mekân konumlandırma sistemine uygulanmış ve ölçüm doğruluğu üzerindeki olumlu etkisi yapılan testlerle kanıtlanmıştır. Test alanı olarak 7.35 m × 5.41 m boyutlarında 39.76 m 2 'lik bir alan seçilmiş ve özel olarak tasarlanmış bir tavan sistemine yerden 2.85 m yüksekliğe üç farklı UWB alıcı yerleştirilmiş ve 182 adet test noktasından 60 sn. süreyle toplam 10.920 ölçüm alınmıştır. Ölçüm sonuçları Büyük Patlama-Büyük Çöküş optimizasyon algoritması ile düzeltilerek, ortalama hatası önceki 20.72 cm değerinden 15.02 cm'ye düşürülmüş, böylelikle ölçüm sonuçlarının doğruluğu arttırılmıştır. Ultra-wide Band technology is an emerging technology that offers successful solutions in many indoor positioning systems and performs better than other methods. In this study, an indoor positioning system using Ultra-wide Band (UWB) sensors was developed and it was aimed to increase the accuracy level of the standard equipment with the additional algorithms used while reducing the average error. For this purpose, the Big Bang-Big Crunch (BB-BC) optimization method has been applied to the experimental indoor positioning system and the positive effect on the measurement accuracy has been proved by the tests made. An area of 39.76 m 2 was selected as a test area of 7.35 m × 5.41 m and three different Ultra-wide Band receivers were installed at a height of 2.85 m on a specially designed ceiling system and a total of 10.920 measurements were taken from 182 test points for 60 seconds. By correcting the measurement results with the Big Bang-Big Crunch optimization algorithm, the average error was reduced from the previous 20.72 cm to 15.02 cm, thus the accuracy of the measurement results were improved.

Improving non-line-of-sight situations in indoor positioning with ultra-wideband sensors via federated Kalman filter

“…One of the approaches used for enhancing indoor positioning accuracy with UWB sensors is the time of flight (TOF) calculation combined with the big bang-big crunch optimization method. The results showed a significant improvement of 27.5% in positioning accuracy [9].…”

“…In this study, Q was chosen as a 4×4 corner matrix with a value of 0.0003. Then, the produced state estimation (𝑥 ̂ 𝑖 ) (10) and error covariance matrix (𝑃 𝑖 ) (11) are updated with the Kalman gain (𝐾 𝑖 ) (9).…”

Accurate indoor positioning with ultra-wide band sensors

“…Therefore, the studies shifted in this direction. The average positioning error value of 24 cm was first reduced to 20.72 cm by making the necessary calibrations and developing a ceiling system by placing UWB sensors at the appropriate height [29].…”

“…TREK1000 UWB sensors were used in the first studies to eliminate the offset value and all the results were tested by calculating a common offset value. In [29], a single offset value was obtained.…”

“…By using this offset value, it was possible to reduce the average positioning error value from 20.72 cm to 15.02 cm [29]. Afterwards, the studies have been developed to reduce offset and noise effects in order to reduce the average positioning error to around 15 cm.…”