鲁棒混合0-1多项式规划

Robust mixed 0-1 polynomial programming

针对含误差数据的混合0-1多项式优化问题,给出一种鲁棒优化方法,以提高其最优解的鲁棒性。该方法先将原问题转化为混合0-1线性规划,并在最坏情况下给出混合0-1线性规划的鲁棒对应模型,随后利用该鲁棒对应模型求解原优化问题。数值试验表明,该方法所求出的最优解具有良好的鲁棒性。

Considering that the data of a mixed 0-1 polynomial optimization problem are with error, a robust optimization method is proposed to improve the robustness of the problem＇s optimal solution. By the proposed method, the original problem is transformed into a missed 0-1 linear programming model firstly, then, in the worst case, a robust optimization counterpart is put forward, and finally, based on this corresponding robust optimization counterpart, the original problem can be solved. Numerical experiments show that the optimal solution worked out in this way is of good robustness.