The Gravitational Search Algorithm (GSA) is novel metaheuristics approach inspired by the laws of gravitation and motion. In GSA, a set of agents, called masses, searches the design space in order to find the optimal solution by simulation of Newtonian laws of gravity and motion. In this work, a standard and an improved GSA approach based on a quasi-oppositional approach are presented and tested on a magnetic pole design benchmark. Results indicate that the performance of proposed improved GSA on the magnetic pole design is better than the that of classical GSA.