The linearization method of Pshenichny is known as a reliable, efficient and grobally convergent method for structural optimization problems. But it suffers from inefficiency of one-dimensional search where several times of structural analysis are required.
Constraint approximation technique of Prasad is incorporated into the algorithm to improve such inefficiency. Several examples show that accurate constraint approximation is possible by the method of Prasad and overall computational efficiency can be improved more than 10%.