作者: Jay Farrell
DOI:
关键词:
摘要: This project is concerned with accurately and reliably determining the state of a vehicle relative to specified trajectory (e.g., lane centerline). We are utilizing inertial navigation methods based on inexpensive solid sensors aided by external such as carrier phase differential GPS, magnetometers, roadway height. Due this integration sensors, reliability increased single sensor approach changes required infrastructure may be significantly decreased. projects objectives included preparation for participation in DEMO2002, research into INS GPS tracking satellites, use auxiliary (roadway height or magnetometers) aid integer resolution process. has been very interesting successful. Although Demo2002 was ultimately cancelled our preparations it were fruitful we did participate smaller demonstration PATH at Crows Landing. In preparing Demo2002, constructed much more robust prototypes hardware. include significant software re-writing that improved software. Overall, hardware now reliable easier install test vehicles. area aiding receiver, performed new algorithm design, but not implemented. expect would lead better satellite during brief interruptions signal. Implementation require extensive interactions receiver design team, which no longer possible due industry. ambiguity resolution, have developed algorithms information from non-GPS facilitate speed-up process resolution. evaluated two experiments UCR improvements ability correctly identify ambiguities epoch fewer than 6 satellites. also used Landing testing. Finally, worked generated an impressive set data shown Figures 12-25 Section 4.2. culmination project. The mounted GPS/INS bus already instrumented magnetometer demonstrated (1) control matched well both when only DCPGPS INS; (2) seamless transitions between control, manual control; (3) advanced maneuvers changes.