A well-planned electric transmission infrastructure is the foundation of a reliable and efficient power system, especially in the presence of large scale renewable generation. However, the current electricity market designs lack incentive mechanisms which can guarantee optimal transmission investments and ensure reliable integration of renewable generation such as wind. This paper first proposes a stochastic bilevel disjunctive program for optimal transmission investment based on the newly proposed theoretical H-R-G-V incentive mechanism. The upper level is a profit-maximization problem of an independent transmission company (Transco), while the lower level is a welfare maximization problem. The revenue of the Transco is bounded by a regulatory constraint set by the regulator in order to induce socially optimal investments. The application of the H-R-G-V mechanism allows the regulator to ensure social maximum transmission investments and helps to reduce transmission congestion and wind power spillage. The transmission investment under the H-R-G-V mechanism is modeled as a stochastic bilevel disjunctive program. To solve the developed mathematical model we first propose a series of linearization and reformulation techniques to recast the original model as a stochastic mixed integer linear problem (MILP). We exploit the disjunctive nature of the reformulated stochastic MILP model and further propose a Bean decomposition algorithm to efficiently solve the stochastic MILP model. The proposed decomposition algorithm is also modified and accelerated to improve the computational performance. The computational performance of our MILP modeling approach and modified and accelerated Bean decomposition algorithm is studied through several examples in detail. The simulation results confirm a promising performance of both the modeling approach and its solution algorithm.

精心规划的电力传输基础设施是可靠高效的电力系统的基础，尤其是在大规模可再生能源发电的情况下。但是，当前的电力市场设计缺乏激励机制，该机制不能保证最佳的输电投资并确保可靠地整合可再生能源，例如风能。本文首先基于新近提出的理论HRGV激励机制，提出了一种用于最优输电投资的随机双层分离式计划。上层是独立输电公司（Transco）的利润最大化问题，下层是福利最大化问题。Transco的收入受到监管机构设定的监管约束的约束，以吸引社会上的最佳投资。HRGV机制的应用使监管机构能够确保社会最大程度的输电投资，并有助于减少输电拥堵和风电泄漏。HRGV机制下的输电投资被建模为随机的两级析取程序。为了解决已开发的数学模型，我们首先提出了一系列线性化和重构技术，以将原始模型重铸为随机混合整数线性问题（MILP）。我们利用重新构造的随机MILP模型的析取性质，并进一步提出了一种Bean分解算法来有效地解决随机MILP模型。还对提出的分解算法进行了修改和加速，以提高计算性能。通过几个示例详细研究了我们的MILP建模方法以及改进和加速的Bean分解算法的计算性能。仿真结果证实了建模方法及其求解算法的良好前景。