In view of the two problems often encountered in seeking solutions with the basic ant colony algorithm， that is， the search ending up in becoming locally optimum and the time taken for convergence becoming globally optimum， the ant colony algorithm is improved with respect to the transfer probability and information modification in the global range， so that its ability of searching for the globally optimum solution and the speed of converging at the globally optimum solution is improved. The improved algorithm has been applied to solving the problem of single stage transmission network expansion planning， and the corresponding mathematical model and solving methods are developed. The calculation results of a sample system show that the approach is feasible and effective.