Organizers
Luke Achenie
, Vladik Kreinowich
and Kaj Madsen
University of Connecticut at Storrs, USA
email: achenie@engr.uconn.edu
University of Texas at El Paso, USA
email: vladik@cs.utep.edu
Technical University of Denmark at Lyngby, Denmark
email: km@imm.dtu.dk
In many practical problems (such as process design, modeling, identification, optimization, and control) there is a need to (a) solve systems of equations and inequalities, and/or (b) optimize some performance measure. The results obtained by conventional algorithms are either local or cannot be guaranteed. Deterministic global optimization methods have evolved to the point that obtaining global solutions for certain classes of problems are almost routine. Even then, since the results cannot be validated (due to round-off and limits on machine precision), some global solutions may be missed. Interval analysis provides guaranteed approximations of the set of all the actual solutions of the problem. This ensures that no solution is missed. In recent years researchers in many areas of science of engineering - e.g., in Chemical Engineering - have used interval analysis to locate, e.g., all roots of a system of equations, including those that have been missed using conventional approaches. Interval analysis has also been employed very successfully in global optimization. The state of the art global optimizers such as alpha-BB incorporate some interval analysis ideas. This minisymposium is devoted to the theory and application of interval analysis in validated computing and deterministic global optimization. We particularly encourage successful industrially-relevant applications. We also encourage problems that can evolve as benchmark case studies for future research.