An Approach to Nonsmooth Unconstrained Global Optimization Based on Filled Function

This paper introduces a filled function method to find better minimizers of nonsmooth unconstrained global optimization problem. Firstly, the definition of filled function is defined and a filled function is constructed. Then, the numerical properties of this filled function are discussed. And at last, an solution algorithm based on these theoretical properties is described and a numerical experimentation is reported. The preliminary computational results Confirm the usefulness and effectiveness of the approach.