首页|基于相似日拓扑辨识的发电调度潮流数据自动校核方法

基于相似日拓扑辨识的发电调度潮流数据自动校核方法

扫码查看
受负荷状态以及发电情况的波动性影响,潮流数据的校核误差难以控制,为此,提出基于相似日拓扑辨识的发电调度潮流数据自动校核方法.首先,通过电网状态矩阵,计算得到各全纯函数幂级数系数的逼近状态后,利用帕得逼近算法确定电网状态相似日.然后,以确定的相似日为基准,原大规模的拓扑节点潮流等效问题转化为一系列可以并行求解的小规模子问题形式,并将电网发电调度计划中的惯量信息引入到梯度下降法中.最后,采用自适应矩估计的方式对电网拓扑中潮流参数的分布情况加以辨识,求解双随机性质的线性潮流拓扑分布矩阵,实现对潮流数据的校核.在测试结果中,设计方法对电网中各节点电压幅值、相角以及有功功率的校核结果表现出较高的稳定性和可靠性.
Automatic Checking Method For Power Flow Data Of Power Generation Dispatching Based On Similar Day Topology Identification
Due to the fluctuation of load status and power generation,it is difficult to control the verification error of power flow data.Therefore,an automatic verification method for power generation scheduling power flow data based on similar day topology identification is proposed.Firstly,by using the grid state matrix,the approximate states of the power series coefficients of each holomorphic function are calculated,and then the Pade'approximation algorithm is used to determine the similarity days of the grid state.Then,based on the determined similar days,the original large-scale topology node power flow equivalence problem is transformed into a series of small-scale model problems that can be solved in parallel,and the inertia information in the power grid generation scheduling plan is intro-duced into the gradient descent method.Finally,the distribution of power flow parameters in the power grid topology is identified using adaptive moment estimation,and the linear power flow topology distribution matrix with double random properties is solved to verify the power flow data.In the test results,the design method showed high stability and reliability in verifying the voltage amplitude,phase angle,and active power of each node in the power grid.

automatic verificationpower grid state matrixcoefficients of the power series of a holomorphic functionPad approxima-tion algorithmgradient descent methodadaptive moment estimationlinear power flow topology distribution matrix

刘旭斐、秦文超

展开 >

云南电网有限责任公司 昆明 661599

自动校核 电网状态矩阵 全纯函数幂级数系数 帕得逼近算法 梯度下降法 自适应矩估计 线性潮流拓扑分布矩阵

2024

网络新媒体技术
中国科学院声学研究所

网络新媒体技术

CSTPCD
影响因子:0.208
ISSN:2095-347X
年,卷(期):2024.13(6)