To deal with the problem of premature convergence and slow search speed, this paper proposes a novel particle swarm optimization (PSO) called self-adaptive escape PSO, which is guaranteed to converge to the global optimization solution with probability one. Considering that the organisms have the phenomena of escaping from the original cradle when they find the survival density is too high to live, this paper uses a special mutation –escape operator to make particles explore the search space more efficiently. The novel strategy produces a large speed value dynamically according to the variation of the speed, which makes the algorithm explore the local and global minima thoroughly at the same time. Experimental simulations show that the proposed method can not only significantly speed up the convergence, but also effectively solve the premature convergence problem.