考虑多时段CVaR的供电公司购电优化模型

电力市场环境下，由于市场价格的不确定性和负荷需求的随机性，供电公司面临在多个市场间购电时需要综合考虑风险和收益的权衡问题。单时段条件风险价值（conditional value at risk，CVaR）可以衡量单一时段购电的风险和收益，但购电过程是一个动态的优化问题，因此将多时段CVaR风险收益模型应用于供电公司的购电组合中，建立以风险最小化为目标的多市场购电组合优化模型。应用该模型，对供电公司在多个时段多个市场的购电分配进行计算，并利用该模型分析不同置信水平和期望收益对购电分配决策的影响。考虑到风险对未来投资收益的长期影响，引入转移影响率，将上一时段的风险转移到下一时段，从而降低整体风险。算例结果表明了该方法的有效性，从而为供电公司的购电决策及其风险评估提供新的思路。

Optimal Model of Power Purchasing for Power Supply Company Considering Multi-period CVaR

In the power market, for the uncertainty of market price and the randomness of load demand, the power supply companies need synthetically consider the balance between risks and revenues when purchasing electricity in a number of markets. Single -period conditional value at risk （CVaR） can only measure the risks and revenues of single period, but the process of purchasing electricity is a dynamic optimal problem. So the muhi - period CVaR model is applied to the portfolio, and optimal power purchasing portfolio model for multiple markets in which the minimum risk is the objective function is proposed. Using the model, the allocation in different periods and different markets can be calculated. Furthermore, the effect of different con- fidence level and expected revenues on purchasing strategy is analyzed. Taking into account of the long - term impact of risks on the future investment income, the risk - transferring rate is introduced in order to transfer the risks of one period to the next period and reduce the overall risks. The calculation results show the validity of the multi - period CVaR model, and provide a new way for purchasing strategy and risk evaluation.