| 引用本文: | 郝广涛,韩学山,梁 军,贠志皓,董晓明,张学清.广域环境下基于Q型因子学习方法的电网节点聚合规律?[J].电力自动化设备,2015,35(1): |
| HAO Guangtao,HAN Xueshan,LIANG Jun,YUN Zhihao,DONG Xiaoming,ZHANG Xueqing.Grid node aggregation law based on Q-factor learning in wide-area environment[J].Electric Power Automation Equipment,2015,35(1): |
|
| 摘要: |
| 大规模可再生能源的并网以及电力市场解除管制的改革，使传统集中式的节点电压在线安全分析、调度与控制方法面临困境。对此，在电网运行状况全景过程化可观测条件下，提出了渐进学习的电网节点聚合的新调控理论，通过电网“局部自治与集中调控互融”的调控方式解决电网的调控问题。提出了基于Q型因子学习的电网节点聚合规律的挖掘方法，根据传统节点电压方程及发电机、负荷的等值模型得到广域环境下节点电压向量与电势之间的解析关系，推导出连续2个量测时刻下节点电压幅值变化的影响因子，并对其进行过程化的Q型因子学习，进而得到电压幅值具有一致变化的节点聚合规律。通过对德州电网的仿真分析，验证了所提方法的准确性及有效性。 |
| 关键词: 电力系统 全景过程化可观测 渐进学习 局部自治与集中调控互融 节点聚合 Q型因子 模型 广域测量 调控 |
| DOI： |
| 分类号: |
| 基金项目:基金项目：国家自然科学基金资助项目（51177091） |
|
| Grid node aggregation law based on Q-factor learning in wide-area environment |
|
HAO Guangtao1, HAN Xueshan1, LIANG Jun1, YUN Zhihao2, DONG Xiaoming2, ZHANG Xueqing3
|
|
1.Key Laboratory of Power System Intelligent Dispatch and Control of Ministry of Education，Shandong University，Ji’nan 250061，China;2.State Key Lab of Control and Simulation of Power Systems and Gerneration Equipments，Department of Electrical Engineering，Tsinghua University，Beijing 100084，China;3.Jinan Municipal Electrical Power Company，Ji’nan 250012，China
|
| Abstract: |
| As the grid-connection of large-scale renewable energy and the reform of deregulated electricity market make the centralized online node-voltage security analysis，dispatch and control even difficult，a regulation theory of node aggregation based on the progressive learning in the condition of panoramic process observability for power system operation is proposed，which adopts the power grid dispatch mode of “interacted local autonomy and central control”. A method based on Q-factor learning is proposed to mine the law of node aggregation，which，based on the traditional node voltage equation and the equivalent models of generator and load，obtains the analytical relationship between the node voltage phasor and the electric potential in the wide-area environment，deduces the influencing factor of voltage magnitude variation during two successive measurements，and achieves the aggregation law of nodes with same direction of voltage magnitude variation through the process of Q-factor learning. The simulative analysis for Dezhou Grid verifies the correctness and effectiveness of the proposed method. |
| Key words: electric power systems panoramic process observability progressive learning interacted local autonomy and central control node aggregation Q-factor models wide area measurement dispatch |
