9409715 Feron The two most useful properties of a good control system design are robustness and performance. Performance is the ability for a control system to satisfy a number of specifications, usually based on some inputs and outputs. Robustness is the property that a system perform well even in the presence of uncertainties. This research will develop a method that combines the theory of stability multipliers with numerical convex optimization to handle the problem of robust performance for control systems analysis and synthesis. Preliminary investigations suggest that this technique should perform considerably better than existing analysis and synthesis methods for robust control by solving a much wider class of problems. The use of interior-point optimization techniques combined with the newly developed concept of Linear Matrix Inequality may lead to small amounts of computations, especially suitable for fast, interactive design tools. ***