In this work, firstly, a decomposition technique is proposed and applied to solve the Security Constrained Unit Commitment problem (SCUC). A recently developed model for optimal short term programming of power generation in thermoelectric plants including transmission constraints is adopted. Real gap is computed at the final solution. Secondly, a method to analyze results and redesign the system structure is also proposed. By applying this method, critical lines and generators can be detected, and then, network improvements can be suggested by either eliminating or duplicating lines or generators. With the aim of studying the efficiency of the proposed method, two systems of different sizes are considered. The efficiency of the proposed technique is compared with the direct resolution of the original model and a linearly approximate model. The proposed decomposition technique presents a good performance: optimal solutions are obtained within small real gaps and efficient computing times. In power systems, the SCUC problem is solved every 60 minutes or less; therefore, the resolution time sought must be much less than that. Without applying the decomposition technique, the resolution of the SCUC problem takes several hours to reach an acceptable gap; and instead, when applying the decomposition technique; smaller gaps can be obtained in a few minutes.

在这项工作中，首先，提出了一种分解技术，并将其应用于解决安全约束单元承诺问题（SCUC）。采用了最近开发的用于优化热电厂短期发电规划的模型，该模型包括传输约束。实际差距是在最终解决方案中计算的。其次，提出了一种分析结果并重新设计系统结构的方法。通过应用此方法，可以检测到关键线路和发电机，然后可以通过消除或复制线路或发电机来提出网络改进建议。为了研究所提出方法的效率，考虑了两个不同大小的系统。将所提出技术的效率与原始模型和线性近似模型的直接分辨率进行比较。所提出的分解技术表现出良好的性能：在较小的实际间隙内以及有效的计算时间内获得了最佳解决方案。在电力系统中，每60分钟或更短的时间内解决一次SCUC问题。因此，所需的解决时间必须远远少于此。不使用分解技术，解决SCUC问题需要几个小时才能达到可接受的水平。相反，当应用分解技术时；几分钟即可获得较小的间隙。解决SCUC问题需要花费几个小时才能达到可接受的水平；相反，当应用分解技术时；几分钟即可获得较小的间隙。解决SCUC问题需要花费几个小时才能达到可接受的水平；相反，当应用分解技术时；几分钟即可获得较小的间隙。