摘要
研究一种适合大型化学工业园区集中大负荷用户能源需求条件的电网整合平衡控制调度方案。通过对电网内的电源站母线负荷和园区内企业用电负荷进行一维矩阵建模,通过神经网络循环自博弈学习模式进行数据挖掘分析,实现一种适应该调度环境的平衡调度智能新算法。结果表明,与该园区早期使用的平衡调度方案相比,该算法在倒闸频率提升43.9%的前提下,实现无功增压比和电压峰值波动比分别下降71.2%和50.6%,同时实现7个电源站负荷情况最大偏差比提升3.8倍的控制效果。
A power grid integrated balanced dispatching scheme suitable for the energy demand conditions of centralized large load power users in industrial park was studied.Through one-dimensional matrix modeling of bus load of power station and power load of enterprises in the park, and data mining and analysis by neural network cyclic self game learning mode, an innovative algorithm of balanced dispatching suitable for dispatching environment was realized. Results show that compared with the balanced scheduling scheme used in the early stage of the park, the improved algorithm can reduce the reactive power boost ratio and the peak voltage fluctuation ratio by 71.2% and 50.6% respectively under the premise of increasing the switching frequency by 43.9%.At the same time, the maximum deviation ratio of seven power stations is increased by 3.8 times.
作者
李蒙赞
LI Mengzan(Electric Power Research Institute of State Grid Shanxi Electric Power Company,Taiyuan 030001,Shanxi China)
出处
《粘接》
CAS
2023年第2期117-120,共4页
Adhesion
关键词
工业园区
平衡调度
大负荷用户
神经网络
自博弈学习
industrial park
balanced scheduling
heavy load users
neural network
self game learning
