臺灣博碩士論文加值系統

本研究主要建立低成本的可傾車輛之車身動態姿態估測系統。現行傾窄型多輪可傾車輛高速過彎時，必須側傾車身以抵抗車輛過彎時產生之翻覆力矩而過彎，若駕駛者操作不當，可能有翻覆之危險。若配備車身動態姿態估測系統，可以用於開發車輛翻覆預警系統，以便警示駕駛者或介入控制車速與傾角，使駕駛者免於翻車之危險。
目前車用姿態感測器通常只提供車身三軸的角速度，無法直接呈現車身姿態，必須透過估測技術來獲得車身三軸的角度。但車輛之使用環境複雜，姿態之估測精度常會受到車體震動與環境影響，因此為一種高難度的技術。
本研究以低成本的市售九軸感測器量測車身三軸的角速度為基礎，搭配四元數法，巴特沃斯濾波器，以及Mahony互補濾波器之誤差修正法，建立車輛動態姿態估測法則，目標為在雜訊環境下，獲得高精度的車身動態姿態估測值。
研究成果為一套以微處理器為主的資訊系統，具備量測、參數估測、以及顯示車體動態資訊之功能。該系統安裝於一部三輪可傾機車上進行實測，實車測試時速為25 km/hr，各項行使模式之測試結果顯示，估測誤差值均在合理範圍，且誤差值呈現高斯分布。其中，S型繞錐行駛測試模式之車身動態roll角估計值之誤差平均值為-0.257度， 標準差1.7063度。Yaw角誤差平均值2.5071，標準差3.1864度。定圓繞錐之roll角度誤差平均值-0.2914度 標準差3.8346度，Yaw角誤差平均值-3.5427標準差3.9687度，以上測試結果初步驗證本研究所建立的系統的可行性。
除了參數估測技術，本文所建立的量測系統及UI介面，可透過藍芽與手機連線，成功地實現以車身姿勢圖，即時(real time)呈現車身動態姿態之功能，且可記錄儲存資料。在未來車輛控制之研究中，可透過此一行車資訊系統，以手機顯示實驗數據與擷取車身資料，不僅增加便利性，還可提升可傾車輛實驗時的安全性。

The topic of this research is to develop a vehicle dynamics estimation and information system. A tiltable vehicle can be tilted its body by driver to create an anti-rollover torque in high speed cornering maneuvers. However, if the driver is not skillful enough, it will take the risk of rollover because of over-tilting operations. Thus a vehicle body posture estimation system could be useful for developing rollover alarming system for the vehicle.
Nowadays, the common used low-cost sensors for estimating vehicle body dynamics are the 9-axes sensor that provide angular speeds, accelerations, and magnetism on the x-, y, and z-axis. This research is focused on utilizing a low-cost nine-axes sensor to develop the vehicle dynamics estimation and information system. In order to apply such a sensor to determine vehicle posture on-line, a series of techniques should be studied to estimate the body angles on the 3 axes, such as the Quaternion, Butterworth filter, and Mahony complementary filter for error modification. The objective is to obtain high accuracy vehicle dynamic posture estimations under high noise conditions.
The main results of this research was a micro-processor-based information system with capability of measuring, estimating and displaying vehicle dynamics information. This system was equipped on a three-wheeled tiltable motorcycle to verify its performance. The test results are as follows.
For the S-pattern drive tests, all the measurement errors showed Gaussian distribution. The roll angle was with an average measurement error of -0.257 and a standard deviation of 1.7063 in degree, while the yaw angle was with an average measurement error of 2.5071 and a standard deviation of 3.1864.
As for the Constant radius cornering drive tests, the measurement errors also showed Gaussian distribution. The roll angle was with an average measurement error of -0.2914 and a standard deviation of 3.8346 in degree, finally, the yaw angle was with an average measurement error of -3.9687 and a standard deviation of 3.1864. all these results have shown the feasibility of the estimating system.
Besides parameter estimating, a UI interface was also established in the controller, which could communicate with cell phones through Bluetooth to display vehicle dynamic posture using posture picture, and storage the recorded data on-line .

摘要 I
謝誌 V
目錄 VI
表目錄 IX
圖目錄 X
第1章　緒論 1
1.1 研究背景與動機 1
1.2 文獻回顧 2
1.3 論文架構 3
第2章 系統架構與感知器校正 4
2.1 系統架構 4
2.2 感知器規格與實驗器材介紹 6
2.2.1 MPU6050介紹 6
2.2.2 HCM5883L介紹 7
2.2.3 VBOX 儀器 7
2.3 感知器參數設定 8
2.4 訊號濾波 9
2.4.1 BUTTERWORTH濾波器 10
2.4.2 BUTTERWORTH驗證 11
2.5 系統截止頻率設計 12
2.5.1 巴特沃斯驗證 13
2.6感知器校正 14
2.6.1 加速度計校準方法 14
2.6.2 陀螺儀校準方法 16
2.6.3 磁力計校準方法 17
第3章 行車資訊系統開發 18
3.1 行車資訊系統開發目的 18
3.2 行車資訊系統軟體架構 18
3.3 ANDROID簡介 19
3.4 通訊方式選擇 20
3.5藍芽資料解析方法 20
3.6 圖形介面與功能展示 21
第4章 可傾車輛車身姿態估測 26
4.1姿態估測參考方法 26
4.1.1 地理座標系 26
4.1.2 車體座標系 27
4.1.3旋轉矩陣 27
4.1.4 歐拉角法 28
4.1.5 四元數法 29
4.2四元數與旋轉矩陣轉換 30
4.3 車體姿態估測方法 32
4.3.1 誤差修正方法 32
4.3.2 四元數更新 36
4.3.3 姿態更新流程 37
第5章 可傾車輛車身姿態驗證 38
5.1動態測試-S型繞錐 38
5.2動態測試-定圓繞錐 42
5.3車身姿態結論 45
第6章 結論與建議 47
參考文獻 49
作者簡介 51

[1] R.Mahony, T.Hamel, and J.-M.Pflimlin. , 2008 , Nonlinear complementary filters on the special orthogonal group. Automatic Control, IEEE Transactions on, 53(5):1203 –1218.
[2] Gregory G. Slabaugh,Computing Euler angles from a rotation matrix. http://www.staff.city.ac.uk/~sbbh653/publications/euler.pdf.
[3] Ken Shoemake, 1985 ,Animating Rotation with Quaternion Curves Proc. SIGGRAPH, pp. 245 - 254..
[4] Matthew Leccadito,2013,A Kalman Filter Based Attitude Heading Reference System Using a Low Cost Inertial Measurement Unit, Virginia Commonwealth University
[5] WALTER T. HIGGINS, JR ,1975,A Comparison of Complementary and Kalman Filtering, IEEE Arizona State University Tempe, Ariz. 85281.
[6] Greg Welch, Gary Bishop,2006,An Introduction to the Kalman Filter, Department of Computer Science University of North Carolina at Chapel Hill Chapel Hill, NC 27599-3175.
[7] James Diebel ,2006,Representing Attitude: Euler Angles, Unit Quaternions, and Rotation Vectors, Stanford University Stanford, California 94301–9010.
[8] Ken Shoemake,1993,Quaternions, Department of Computer and Information Science University of Pennsylvania Philadelphia, PA 19104.
[9] Ken Shoemake,1994 ,Euler angle conversion.
[10] SOH Madgwick,Estimation of IMU and MARG orientation using a gradient descent algorithm, 2011 IEEE International Conference on Rehabilitation Robotics Rehab Week Zurich, ETH Zurich Science City, Switzerland.
[11] 徐嘉偉，應用互補濾波器於慣性感測之運動與姿態角估算，碩士論文，國立臺北科技大學，機電整合研究所。
[12] 徐偉群，模糊卡爾曼濾波器應用於載具姿態判定，碩士論文，國立臺灣海洋大學，機械與輪機工程學系。
[13] 余宗儒，2008，三軸加速規應用於人體姿態估測，碩士論文，國立中正大學，電機工程所。
[14] 高忠福，2006，使 用 平 面 加 速 規 及 磁 場 感 測 器 之 姿 態 估 測 器，碩士論文，國立交通大學，機械工程學系研究所。
[15] http://www.3dgep.com/understanding-quaternions/
[16] IMU Errors and Their Effects. http://www.novatel.com/assets/Documents/Bulletins/APN064.pdf
[17] https://developer.android.com/reference/android/view/View.html
[18] https://developer.android.com/reference/android/bluetooth/package-summary.html
[19] https://hom-wang.gitbooks.io/quadcopter/content/05_SensorCorr.html