不確定環境におけるファジィVaR基準に基づく収益・リスク評価モデルの構築

在JSPS项目的第二年，我们将按照申请JSPS奖学金时制定的研究计划继续在相关领域开展工作。在此期间，我们主要关注几种风险度量的发展，复杂数学模型的构建以及我们的方法在单位承诺和投资组合选择问题中的应用。主要研究目标已经完成，研究成果已在SCI期刊和国际会议上发表。在过去的一年中，我们的成果可以总结如下：1。单元承诺问题的研究：首先，我们将前人提出的单节点模型改进为多节点系统，以获得更现实的决策。对数学模型进行了改进，以处理多节点单元承诺中必要的传输互联约束，并将多目标问题统一为一个函数，从而得到问题的唯一解。并对求解算法进行了相应的改进，以解决复杂的优化问题。其次，最近我们刚刚完成了一项新的研究，构建了一个基于单位承诺的双层电力交易模型。该模型同时处理发电公司和市场运营商的优化问题。我们已经将这个结果提交给IEEE Trans。论电力系统。投资组合选择问题的研究：与其他博士合作。我们实验室的同学们，在这个领域我们也有一些有趣的结果。本文建立了两个新的模糊环境下的投资组合选择模型：第一个模型利用基于技术分析的模糊双随机变量来帮助管理者进行证券未来收益的预测和决策，第二个模型解决了投资者在投资组合选择中的退出策略问题。这两项研究对投资组合问题都是新颖的，并且与已有的文献比较显示出了有效性。我们对多节点单元承诺问题的研究还获得了第十届管理工程国际学术研讨会“优秀论文奖”。少

In the second year of JSPS program, we continue to work in the related fields according to the research plan made when apply the JSPS scholarship. During this period, we mainly focused on the developments of several risk measurements, the constructions of complicated mathematical models and the applications of our approaches in unit commitment and portfolio selection problems. The main objects of the research have been achieved, and we have published the research results in both SCI journals and international conferences.Our results in the past year can be summarized as follows :1. Research on unit commitment problem : First, we improve the single-node model proposed in our previous study to a multi-node system to ferret more realistic decisions. The mathematical model is improved to handle the transmission interconnection constraints which are necessary in a multi-node unit commitment, and we unify the multi-objective problem as one function to obtain a unique solution of the problem. … More Moreover, the solution algorithm is modified accordingly to solve the complicated optimization problem. Second, recently we just finished a new study which constructs a unit commitment-based bilevel electricity trading model. This model handles the optimizations of both the generation company and the market operator simultaneously. We have submitted this result to IEEE Trans. On Power Systems.2. Research on portfolio selection problem : Collaborate with other PhD. students in our lab, we also have some interesting results in this filed. We build two novel portfolio selection models in fuzzy environment : the first study uses technical analysis-based fuzzy birandom variables to help the managers in both security future returns forecasting and decision-making, while the second one addresses the investors' exit strategy in portfolio selection problem. Both of the studies are novel to portfolio selection problem and show effectiveness when comparing with existing literatures.Our research on multi-node unit commitment problem also won the "Excellence paper award" in the Tenth International Symposium on Management Engineering. Less