基于神经网络的电力系统在线稳定性监测与控制OA
Online Stability Monitoring and Control of Power Systems Based on Neural Network
针对电力系统电压稳定性分析与改善问题,提出一种基于人工神经网络的在线稳定性监测与控制的方法.采用最小特征值及其对控制变量的敏感度作为稳定性指标,并利用线性规划技术在不同负荷工况下生成最优无功功率控制方案,构建对应的输入—输出训练样本,用于训练多层前馈神经网络.在改进的IEEE 30节点测试系统中验证所提出的方法,结果表明,优化控制后,最小特征值从0.194提高到0.214,电压曲线也有明显改善.与传统线性规划法相比,所提出的方法的计算时间大大缩短,仅需0.21 s,适合在线监测与控制.
Aiming at the problem of analyzing and improving the voltage stability of power systems,a method for online stability monitoring and control based on artificial neural network is proposed.The minimum eigenvalue and its sensitivity to the control variables are used as the stability indexes,and the linear programming technique is utilized to generate the optimal reactive power control scheme under different load conditions.The corresponding input-output training samples are constructed for training the multi-layer feedforward neural network.The proposed method is validated in the improved IEEE 30-node test sys-tem.The results show that after the optimized control,the minimum eigenvalue increases from 0.194 to 0.214,and the voltage curve also improves significantly.Compared with the traditional linear programming method,the calculation time of the pro-posed method is greatly shortened to only 0.21 s,which is suitable for online monitoring and control.
陆俊;蒯文科;胡丹;蒋明;杨奚诚
国网安徽省电力有限公司信息通信分公司,安徽,合肥 230061国网安徽省电力有限公司信息通信分公司,安徽,合肥 230061国网安徽省电力有限公司信息通信分公司,安徽,合肥 230061国网安徽省电力有限公司信息通信分公司,安徽,合肥 230061智慧电网数字协同技术安徽省联合共建学科重点实验室,安徽,合肥 230088
信息技术与安全科学
人工神经网络负载流电压稳定性电力系统在线监测与控制
artificial neural networkload flowvoltage stabilitypower systemsonline monitoring and control
《微型电脑应用》 2026 (2)
23-27,5
国家自然科学基金项目(61502008)
评论