Modeling, identification and analysis of tractor and single axle towed implement system

Abstract

ABSTRACT Increased and sustained agricultural productivity is a key to meet the globally increasing demands for food and energy. Automation of agricultural machinery is one of the ways to improve the efficiency and productivity of various field operations. Because a field implement performs most of these operations, accurate implement guidance is needed to reduce production cost, increase yield, and improve sustainability. Model-based guidance controller design and virtual prototyping techniques can be used in automatic guidance controller development to improve the accuracy and robustness of the guidance controller while reducing the development time and cost. Hence, development and analysis of accurate tractor and implement system models are needed to support automatic tractor and implement guidance controller development. Real-time vehicle model simulation capability allows engineers and users to intuitively interact with the realistic virtual prototypes and to evaluate the performance of physical hardware. As the model complexity is increased to improve the model accuracy and/or fidelity, the computational need will also increases thus increasing the challenge to meet real-time constraints. In this regard, it is important to minimize the computational load to a Virtual Reality (VR)-based real-time dynamics model simulation system. In this dissertation, various strategies were investigated to reduce the computational burden on the dynamics model simulation so that real-time simulation could be achieved for increasingly complex models. A distributed architecture was developed for a virtual reality-based off-road vehicle real-time simulator to distribute the overall computational load of the system across multiple machines. Multi-rate model simulation was also used to simulate