Adaptive control using multiple models, switching, and tuning

Abstract

The paper describes the efforts to increase substantially the scope of adaptive control by using multiple models, switching, and tuning. Both deterministic and stochastic systems are considered and stability and convergence problems are discussed. While deeply rooted in conventional adaptive control theory, the multiple model approach provides greater flexibility in design, and in many cases may result in faster, more accurate, and more robust systems. Practical applications in flight control, process control, the control of a flexible transmission system as well as a solar power plant, where the approach has already proved to be successful, are also described.