The goal is to increase fundamental understanding of intelligent and learning control theory and practice of this project. The specific objective is to explore intelligient control concepts by identifying, modeling, and controlling the combustion processes in both large gas turbines and automobile engines. A multidiscipinary approach will be used to concurrently make fundamental progress in control theory, develop real time learning and control algorithms, and conduct laboratory experiments to test ideas. Results form statistics, neural networks, artificial intelligence, knowledge-based systems, and biology will first be investigated. A hierarchy of controllers will then be developed: a human operator, a top-level discrete action controller, and lower level continuous controllers. Finally, the project will verify a validate the intelligent and learning control systems by a combination of theoretical exploration, computer simulation, and laboratory experimentation. This research offers the protentinal of substantial economic and societal benefits in real world applications. An intelligent engine control system that can optimize the proper air/fuel ratio on a real time basis can provide real advances simultaneously in both air pollution reduction and improved fuel economy.//
