The objective of the present proposal is the development of some mathematical and computational tools that are appropriate to the next generation of CACSD (Computer Aided Control System Design) environments. These future CACSD packages will be radically different from the present packages in that they will truly be able to perform control systems synthesis and rapid prototyping, rather than just analysis and simulation. A key role in this is played by the integration of modeling, system identification, and robust control design. The long range goals address the following problems: (i) estimation of model accuracy from measured and prior data, (ii) tuning of the identification experiment for robust controller design, (iii) development of reliable and efficient algorithms, and (iv) development of an interactive object oriented environment which is sensible for a wide variety of users. Phase I of the proposed research is directed towards the feasibility of solving these problems. Phase II will extend the results of Phase I as well as develop the corresponding software primitives.
