For swarm intelligence optimization algorithm of solving high-dimensional and multi-modal functions, it is difficult to optimize the particles for each dimension and it is easy to fall into local extreme point. On the basis of analyzing the mechanism of quantum- behaved particle swarm optimization (QPSO) algorithm, the QPSO algorithm is improved: to compare each dimension of the previous generation particle with the later generation to optimize and to construct a new control function of the contraction-expansion coefficient. The experimental results show that the improved algorithm are significantly outperforms the QPSO algorithm in convergence accuracy and convergence rate. Specifically, it is of strong ability to avoid falling into the local optimum. It is very suitable for solving high-dimensional and multi-modal optimization problems.