A Generic Framework for Flexible Agent-Based Intelligent Control

点击翻译按钮获取中文摘要

The research involves a multidisciplinary team, already working together, with heavy experience in control systems, artificial intelligence, power systems, human performance, neural networks, and vision/CAD systems to design a Generic Framework for Flexible Agent-Based Intelligent Control (IC). This objective will be accomplished via three goals: 1. Defining the requirements for an IC framework based on the common themes across existing UTA testbeds in manufacturing material handling, power systems simulation, human performance/telerobotics, automated surface finishing, flexible real-time digital control implementation, and autonomous vehicle navigation and control; 2. Development of a generic IC framework applicable in all these testbeds to confront design and control problems involving complex nonlinear systems, hierarchical levels of abstraction and time scales, cognitive modeling and learning, human interface at several levels, man-in- the-loop, coordination of multiple agents, sensor fusion, real-time control system reconfiguration, and failure recovery. The framework will be object-oriented, modular, and agent-based; and 3. Evaluation of the generic IC framework by: application to the testbed areas, verification of ability to support current architectures (e.g. NASREM, NGC), and development of a case-study- based graduate course in intelligent control. Extensive industry involvement will be provided through existing Industry Advisory Committees and company membership programs at ARRI and HPI, ongoing industry contracts, and other UTA programs. //