The interior point cutting plane method（IPCPM） possesses both the advantages of cutting plane method and interior point method， so it is suitable for solving discrete optimization problems of large scale system. But in research it is found that if the problem has multiple solutions， the optimal solutions of interior point method will converge to the interior of the optimal face， and cutting planes can not be generated correctly due to the failure of identification of optimal base， which lead to the failure of IPCPM. A new general optimal base identification method is presented in this paper to solve the problem， which improves the robustness of the algorithm. The simulation results of IEEE test systems indicate that the improved algorithm can properly handle various types of optimal solutions and significantly enlarge the application field of IPCPM.