Modifications of the ISIM Algorithm for Bound Constrained Function Minimization

The improved simplex method (ISIM) based on the Nelder-Mead search (NMS) is possible to find a better search direction by combining the information of function values of the simplex. In this paper, modifications of the ISIM algorithm is described for bound constrained function minimization. Two criterions are put forward to deal with bounded problems for a simplex may flip into an unfeasible region. Simulation shows the feasibility of the modified algorithm