A novel technique is proposed in this paper to optimize the Particle Swarm Optimization (PSO) algorithm. It is named Informed Particle Swarm Optimization (IPSO). A new treatment is added to the conventional PSO which eliminates blind searching in the conventional PSO. In the proposed algorithm, each particle will search it's around by a variable radius before following the gbest and pbest. It makes the proposed algorithm faster in searching the search space and better in finding the optimum point. The radius which each particle can will be decreases look around during the optimization by a nonlinear function. Because of the non blinding search, in the proposed algorithm, probability of falling in the best local is significantly decreased. The proposed algorithm is applied on some benchmarks and simulation results show advantages of the proposed IPSO.