Abstract: Particle swarm optimization(PSO) algorithm tends to suffer from falling into local optima and converging slowly in complex high-dimensional optimization problems. To solve this problem, a particle swarm optimization algorithm with the Hooke-Jeeves operator(HJPSO) is proposed. To avoid the best particle being trapped into local optima, a mechanism for judging whether the particle trap into local optima or not is added into the iteration of PSO. When the signs of premature and stagnation is detected, the Hooke-Jeeves operator is introduced to help the algorithm jump out of the likely local optima by pattern search for the current historical optimal location to the whole particle swarm. The experimental results of several benchmark functions show that the proposed approach has strong capability of preventing premature convergence, better convergence rate and robustness.