基于L形算法的多阶段电网规划

提出了一种计算精度能够满足工程需要，而计算效率显著提高的多阶段网络规划新算法．该方法在灵敏度分析方法上构建输电网多阶段网络规划的线性近似模型，采用Ｌ形算法对该模型进行分解处理，将一个多阶段的高阶电网规划问题转化为多个规划子事件进行降价迭代计算，最终收敛到原始规划问题的满意解．所提方法不仅可以显著提高求解速度和减少内存消耗，而且可以获得与其他方法等效的网络优化扩展方案．通过一个测试算例，验证了该方法的有效性和可行性．

New Approach for Multi-Stage Power Network Planning Based on L-Shape Algorithm

Aiming to improve the solution efficiency of multi stage planning means, it expounded a new approach to reduce greatly computing burden while maintaining essential decision precision. The new approach is mainly built on a sensitivity estimation model via LP initiated by Garver etc. and soluted by using L shape algorithm technique, by which the original complex multi stage expansion optimization problem with high dimension constraint matrix can be converted into several simple planning child events with low dimension constraint matrix, and a satisfactory alternative can be achieved by an iteration solution. The new method demonstrates that it has rapid computing speed and less memory demand and forth and was adequately verified by a reduced actual power system testing example.