即時車輛導航系統之GPS/INS/GIS整合設計

Abstract

本論文提出一套以卡爾曼濾波器整合全球定位系統與慣性導航系統的即時車輛導航系統，再利用地圖匹配演算法消除來自卡爾曼濾波器的誤差，且找到正確的道路與位置。此硬體架構整合一個加速規與一個陀螺儀量測二維位移、速度、車頭方向，同時設計了兩個微算器以擷取全球定位系統接收器與慣性量測元件之資料，然後傳送到主處理器運算。接著主處理器利用卡爾曼濾波器來估算最佳導航資訊。此外本論文提出一個修正的地圖匹配演算法以調整車頭方向。最後最佳導航資訊將會顯示在使用者之圖示介面。實驗結果將證實提出的方法對於車輛行經都市區、地下道、樹林茂盛的地區時是有效且可實現的。

In this thesis, a GPS/INS navigation system integrated by Kalman Filter is proposed. Furthermore, a map-matching algorithm that eliminates the error of GPS/INS and finds the correct road and position on the map is also applied to this real-time vehicle navigation system. One accelerometer and one gyro are integrated in the system to obtain 2-D displacement, velocity, and heading direction. Two microprocessors are designed to capture both the signals of GPS receiver and IMU set, and then transmit data to the main processor. After that, the optimal navigation information of GPS/INS is estimated in the main processor by the Kalman filter. In addition, a modified map-matching algorithm is adopted to adjust the heading direction of vehicle. Finally, the optimal navigation information would be shown in a graphical user interface. The experiment results would demonstrate the effectiveness and applicability of the proposed methods as the vehicle runs in the urban area, underground passage, and wooded area.