arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1708.07930math.OC
收藏

分布鲁棒多阶段随机最优化的研究

A Study of Distributionally Robust Multistage Stochastic Optimization

Jianqiu Huang, Kezhuo Zhou, Yongpei Guan

摘要:

本文研究了一个考虑分布鲁棒性的数据驱动的风险厌恶多阶段随机规划模型。我们优化RMSP的最坏情况下的分布内的一个模糊集的概率分布直接从历史数据样本构建。由于RMSP的目标函数具有多级嵌套的Minimax结构,使得RMSP难以求解,因此本文将其转化为一个包含期望和条件风险值(CVaR)的凸组合的确定性等价模型,并采用定制的随机对偶动态规划(SDDP)算法求解.当所收集的数据样本的大小增加到无穷大时,我们证明了具有分布鲁棒性的RMSP与传统的多阶段随机规划的一致性。此外,为了测试我们提出的模型和算法的计算性能,我们进行了数值实验的风险规避的热液调度问题,其结果表明,我们的RMSP框架的有效性。

英文摘要:

In this paper, we focus on a data-driven risk-averse multistage stochastic programming (RMSP) model considering distributional robustness. We optimize the RMSP over the worst-case distribution within an ambiguity set of probability distributions constructed directly from historical data samples. The proposed RMSP is intractable due to the multistage nested minimax structure in its objective function, so we reformulate it into a deterministic equivalent that contains a series of convex combination of expectation and conditional value at risk (CVaR), which can be solved by a customized stochastic dual dynamic programming (SDDP) algorithm in this paper. As the size of collected data samples increases to infinity, we show the consistency of the RMSP with distributional robustness to the traditional multistage stochastic programming. In addition, to test the computational performance of our proposed model and algorithm, we conduct numerical experiments for a risk-averse hydrothermal scheduling problem, the results of which demonstrate the effectiveness of our RMSP framework.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑