An agent of behaviour architecture for unmanned control of a farming vehicle
Computers and Electronics in Agriculture
Hi-index | 0.00 |
Drive systems are usually modeled using a mathematical characterization of their physical phenomena. However, the difficulty in establishing a relevant model representation, in particular for electro-hydraulic systems, makes important the search for other modeling mechanisms that allow the combination of previously compiled system's knowledge with acquired experimental information. This paper, divided into two parts, describes the potential and possible drawbacks of integrating fuzzy learning mechanisms into a drive system that includes an electro-hydraulic actuator. First, experimental verification of the actuator's fuzzy modeling is presented in Part I of the paper, where the variable selection problem and the performance of the learning algorithm are discussed. In Part II, extensive experimental results employing the extracted fuzzy model and associated learning algorithm are presented. The feasibility and effectiveness of integrating fuzzy learning mechanisms into the actuator's control is also discussed.