A new hybrid algorithm was proposed for dynamically adjusting water volume to solve the optimization problem of secondary cooling water distribution. In this paper, considering the problem that the local search ability of SFLA is weak, we introduced the quantum particle swarm algorithm to improve the clustering accuracy and convergence. At the same time,supplemented with chaos mechanism,we enhanced the self organization of individuals to improve the algorithm global search capability. Experimental results show that the clustering effect of the SFLA with quantum particle swarm is good, has rapid constringency speed. And in the low times test,it is effective to improve the center porosity and corner crack of slab,intermediate crack and shrinkage were improved,the performance of the algorithm was substantially promoted.