用ZigBee無線感測網路做為室內自我調整誤差定位方法之研究

A Self-Adjustment Error Indoor Positioning Method Using ZigBee Wireless Sensor Networks

黃俊元(Chung-Yaun Huang)

指導教授 ： 黃國鼎

國立虎尾科技大學/電機資訊學院/電機工程研究所在職專班/碩士(2012年)

https://doi.org/10.6827/NFU.2012.00153

摘要

使用ZigBee無線感測網路做室內定位研究中，主要是利用待測點接收訊號強度指標(Received Signal Strength Indicator, RSSI)判斷與參考點間距離的特性，再利用三邊定位法來判斷待測點的位置。所以定位前必須先對待測環境建立接收訊號強度與距離之環境模型。但在過去研究中建立環境模型，並未考慮在不同環境實際障礙物之影響與不同時間點下環境模型之訊號強度衰減與距離時變之特性。為了減少環境中因障礙物造成的影響，本研究將提出一接收累積較長時間的RSSI值來做統計平均之循環收集法。實驗中證實可以在不同待測環境中建立更符合其地理特性因素影響考慮下之模型曲線。為符合環境影響時變特性，本文之定位研究實驗中，均會先執行循環收集法來建立更符合該時間與空間下之環境模型曲線。本研究選擇使用三邊定位法，此方法在待測環境空間最少必須佈置三個參考節點，此三個參考節點所圍成之區域，稱為三角形定位區域。待測環境空間多設計為矩形形狀，一般都會在四個角落佈置參考節點。因為四個角落佈置方式會有四個三角形地定位區域形成，待測節點會有二個不同的三角形定位區域可選擇，造成不知選擇哪一個三角形定位區域較為合適之糢糊性產生。所以本文提出一個新的三角形定位區域選擇法，解決上述問題。由於本研究所使用的實驗模組為PCB式天線，隨著不同角度接收到的RSSI值會跟著變化，進而影響待測節點無法選擇到最佳三角形定位區域情形，本文將提出一調整天線角度的方式來解決PCB式天線所造成的影響。實驗結果顯示符合預期之效益。因RSSI值之接收具有時變的特性，當利用已建立之環境模形曲線來估算之後待測節點與參考節點的距離時一定會產生誤差。當此誤差過大時，會影響到三邊定位法估算待測節點位置時，無法估算出明確座標點的情形產生。所以，本文提出自我調整誤差三邊定位法來改善無法定位的情形。最後，本文將實際選擇半室內及室內環境空間來進行定位實驗。從結果顯示，最大定位誤差小於85cm，平均誤差約為35cm，誤差變動率小於19cm。本研究提出的室內自我調整誤差定位方法，確實在ZigBee短距離定位應用中獲得顯著的改善，且確實可在不同環境定位中有效的提高定位的準確度。

關鍵字

ZigBee ； 訊號強度指標 ； 三邊定位法

並列摘要

In the ZigBee-based indoor positioning systems, the received signal strength indicator (RSSI) is the main idea to estimate the distances from reference nodes to target node. Then trilateration method is mostly considered to evaluate the target position. As the result, the system must create the environment model of the relation between RSSI and distance before positioning targets. The previous studies create the environment models with unconcern for the possible change of obstructers in the same environment and the time-variant feature of RSSI values from the same reference. The environment model after created could result in the decrease of positioning accuracy as performs the positioning method. In order to improve the weakness, a cycling RSSI collection method with long-term average is proposed to create the environment model that closes the actual environment feature. In the different experimental environments, the results show the environment models created by the proposed method have good performance in comparison with them created by the previous one. Therefore, the proposed method must be performed to update the environment model before positioning targets each time. trilateration method is considered to position targets in our positioning system. It needs to select three reference nodes around the target in order to evaluate the target position. In convenience, the coverage area around the three reference nodes, called triangular position area. In practice, the position environment is usually formed to rectangular and installed four reference nodes in all corners. There are four triangular position areas appeared in the rectangular position environment and then a target exists two possible triangular position areas to select as performs trilateration method to position. In order to avoid the ambiguity to select a proper triangular position area, a new selection method is proposed to overcome the above problem. In addition, due to our ZigBee experimental equipment’s antenna built on PCB, the RSSI value is affected by the direction between transmitter and receiver. It will not only mistake the selection of triangular position area but also worsen the position accuracy. A programmable antenna adjustment method is proposed to solve the antenna problem of experimental equipment. Due to the time-variant feature of RSSI values, to estimate the distance from reference node to target node always exists some error according to the environment model. If the error of estimated distance between reference and target nodes is too serious, the target position cannot be evaluated to a closed value when the trilateration method is performed. A modified trilateration method with self-adjustment error is proposed to obtain a closed value of target position. Finally, the proposed positioning system is experimented in the actual spaces of E.E. building at NFU. The experiment results show that the all position errors are less than 85cm, the average position error is 35cm, and the error deviation is 19cm. Obviously, they always have obviously improvements for positioning accuracy in the different environments.

並列關鍵字

ZigBee ； RSSI ； Trilateration method

參考文獻

[1] A. Krukowski, D. Arsenijevic, “RFID-based Positioning for Building Management Systems,” Circuits and Systems (ISCAS), pp. 3569-3572, May 30 2010-June 2 2010.

[2] E. Zhang, W. Jiang, Y. Kuang, M. A. Umer, “Active RFID Positioning of Vehicles in Road Traffic,” Communications and Information Technologies (ISCIT), pp. 222-227, 12-14 Oct. 2011.

[3] W. M. Yeung, J. K. Ng, “An Enhanced Wireless LAN Positioning Algorithm based on the Fingerprint Approach,” TENCON 2006. 2006 IEEE Region 10 Conference, pp. 1-4, 14-17 Nov. 2006.

[4] C. B. Lim, S. H. Kang, H. H. Cho, S. W. Park, J. G. Park, “An Enhanced Indoor Localization Algorithm Based on IEEE 802.11 WLAN Using RSSI and Multiple Parameters,” Systems and Networks Communications (ICSNC), pp. 238-242, 2-27 Aug. 2010.

[5] H. Yang, G. Yoon, D. Han, “Floor Accuracy Improvement of Wireless LAN based Large Scale Indoor Positioning,” Intelligent Radio for Future Personal Terminals (IMWS-IRFPT), pp. 1-2, 24-25 Aug. 2011.