Identification and Control of an Automated off Highway

The main objective of this study was to develop an automated agricultural vehicle guidance system that can be easily transplanted from vehicle to vehicle. The proposed solution to this problem is to first perform a tractor model identification, and then use a pole placement technique to place the closed loop dominant poles in their desired locations. One of the most difficult aspects of designing a controller for vehicle guidance is arriving at a good model of vehicle lateral dynamics. This study presents a new approach for identifying the lateral dynamics of an automated off-highway vehicle. A second order model is proposed to represent the vehicle lateral dynamics. An Iterative Learning Identification (ILI) method is used to identify the model parameters. Simulation and experimental results show the convergence of parameters with arbitrarily chosen initial estimations. The estimation results are compared to other traditional identification methods: least squares estimation and gradient based adaptive estimation. The results highlight the practical benefits of the ILI approach – i.e. that it can be performed in a relatively small section of field and therefore done prior to actual usage or engagement with crops iii To my family iv Acknowledgements First and foremost, I would like to thank my advisor, Dr. Alleyne, for all his guidance and inspirations. I started to understand 'control' in a more practical sense than solving exam problems. You helped me grow as a researcher, to find out a problem, and to search for a way of solving. Above all this, you help me find and understand the weakness of myself, and your caring and encouragement showed me a way of improvement. Next in line, I want to thank my peers in Alleyne's research group, who are all among the most intelligent, hardworking, and fantastic people I've ever met. First of all, I would like to thank Yangmin for being such a nice lab mate and a good friend, for offering technical support and for sharing good time together. Thanks also goes to Kira for helping me revising my presentation in the late evening, discussing with me the questions I have on research and helping me setting the timelines for my thesis writing; to Sandipan for helping me understanding difficult paper and theorem; to Dave for your insight in technical issues; to Neera for your help in both research related and beyond; to Vikas, Bin, and Tim, for your willingness to discuss and …