In this paper, we present sufficient conditions for global optimality of a general nonconvex smooth minimisation model problem involving linear matrix inequality constraints with bounds on the variables. The linear matrix inequality constraints are also known as “semidefinite” constraints which arise in many applications, especially in control system analysis and design. Due to the presence of nonconvex objective functions such minimisation problems generally have many local minimisers which are not global minimisers. We develop conditions for identifying global minimisers of the model problem by first constructing a (weighted sum of squares) quadratic underestimator for the twice continuously differentiable objective function of the minimisation problem and then by characterising global minimisers of the easily tractable underestimator over the same feasible region of the original problem. We apply the results to obtain global optimality conditions for optinusation problems with discrete constraints.

In this paper we establish necessary as well assufficient conditions for a given feasible point to be a globalminimizer of smooth minimization problems with mixed variables.These problems, for instance, cover box constrained smooth minimizationproblems and bivalent optimization problems. In particular, ourresults provide necessary global optimality conditions for differenceconvex minimization problems, whereas our sufficient conditionsgive easily verifiable conditions for global optimality of variousclasses of nonconvex minimization problems, including the class ofdifference of convex and quadratic minimization problems. Wediscuss numerical examples to illustrate the optimalityconditions