need more branching operations, especially with many variables and constraints. However, the numerical is because the eliminated solution space is also ignored during B&B procedure. Moreover, if the optimum point is not found in the first subspace, it needs to use B&B procedure twice.
4.2. The iteration number in each branching process In each branching process, it utilizes the simplex method to find the optimum solution without integer limit. Thus, the number of iterations in each branching process is also an important index, which affects the performance of both methods. The numerical results of iteration number are shown in Table 3. Observing these results, we find that our method does not decrease the iteration number much for the cases with few constraints. However, for the cases with many constraints, our method needs much less iterations than B&B method. This is because our method reduces the constraints of any sub-problem and then decreases the iterations of the simplex method. The reduction is more significant for cases with more constraints.