随着现代电力系统的迅猛发展,电网结构和运行方式日益复杂,对状态估计的实时性和准确性也提出了更高的要求。为此,该文提出一种基于深度神经网络的电力系统快速状态估计,通过相关性分析筛选出该状态估计模型的输入量测集,进一步利用海...随着现代电力系统的迅猛发展,电网结构和运行方式日益复杂,对状态估计的实时性和准确性也提出了更高的要求。为此,该文提出一种基于深度神经网络的电力系统快速状态估计,通过相关性分析筛选出该状态估计模型的输入量测集,进一步利用海量历史数据建立基于深度神经网络的状态估计模型。当电力系统的实时量测更新时,将强相关量测输入已建立的状态估计模型中快速获得系统状态的估计结果。通过在IEEE标准系统和某实际省网进行算例仿真表明,所提方法的估计精度和鲁棒性均优于传统加权最小二乘(weighted least square,WLS)和加权最小绝对值估计(weighted least absolute value,WLAV);并且该方法的在线计算时间受系统规模影响较小,由实际省网的仿真结果可知,其计算效率较WLS和WLAV分别提升1.43和27.2倍。展开更多
The increasing penetration of renewable energy resources with highly fluctuating outputs has placed increasing concern on the accuracy and timeliness of electric power system state estimation(SE).Meanwhile,we note tha...The increasing penetration of renewable energy resources with highly fluctuating outputs has placed increasing concern on the accuracy and timeliness of electric power system state estimation(SE).Meanwhile,we note that only a fraction of system states fluctuate at the millisecond level and require to be updated.As such,refreshing only those states with significant variation would enhance the computational efficiency of SE and make fast-continuous update of states possible.However,this is difficult to achieve with conventional SE methods,which generally refresh states of the entire system every 4–5 s.In this context,we propose a local hybrid linear SE framework using stream processing,in which synchronized measurements received from phasor measurement units(PMUs),and trigger/timingmode measurements received from remote terminal units(RTUs)are used to update the associated local states.Moreover,the measurement update process efficiency and timeliness are enhanced by proposing a trigger measurement-based fast dynamic partitioning algorithm for determining the areas of the system with states requiring recalculation.In particular,non-iterative hybrid linear formulations with both RTUs and PMUs are employed to solve the local SE problem.The timeliness,accuracy,and computational efficiency of the proposed method are demonstrated by extensive simulations based on IEEE 118-,300-,and 2383-bus systems.展开更多
目前在电力系统中无法保证相量量测单元完全覆盖的情况下,状态估计需要采用相量量测单元(phasor measurement unit, PMU)与数据采集与监控(supervisory control and data acquisition, SCADA)混合量测进行传统非线性状态估计,但是SCADA...目前在电力系统中无法保证相量量测单元完全覆盖的情况下,状态估计需要采用相量量测单元(phasor measurement unit, PMU)与数据采集与监控(supervisory control and data acquisition, SCADA)混合量测进行传统非线性状态估计,但是SCADA数据精度低,含有较多不良数据,同时混合数据需要迭代求解,会导致计算效率低且存在截断误差。针对该问题,文章提出了一种基于堆叠去噪自编码器(stack denoising autoencoder, SDAE)与极限学习机(extreme learning machine, ELM)伪量测建模的电力系统高容错快速状态估计方法。其将含有不良量测的SCADA量测数据作为SDAE-ELM伪量测模型的输入,节点电压实部与虚部作为输出,根据历史数据进行训练得到伪量测值与伪量测误差模型,训练完成后得到精度较高的伪量测;将伪量测与PMU量测一起进行快速的线性状态估计。仿真结果表明,所提方法在保证估计精度的基础上,提高了计算效率,验证了所提方法的有效性。展开更多
This paper introduces a robust sparse recovery model for compressing bad data and state estimation(SE),based on a revised multi-stage convex relaxation(R-Capped-L1)model.To improve the calculation efficiency,a fast de...This paper introduces a robust sparse recovery model for compressing bad data and state estimation(SE),based on a revised multi-stage convex relaxation(R-Capped-L1)model.To improve the calculation efficiency,a fast decoupled solution is adopted.The proposed method can be used for three-phase unbalanced distribution networks with both phasor measurement unit and remote terminal unit measurements.The robustness and the computational efficiency of the R-Capped-Ll model with fast decoupled solution are compared with some popular SE methods by numerical tests on several three-phase distribution networks.展开更多
目的:探讨创伤性臂丛神经损伤的磁共振成像(MRI)三维循环相位稳态采集快速成像(three-dimensional fast imaging employing steady state acquisition with cycled phases,3D-FIESTA-C)序列和三点法非对称回波水脂分离成像(iterative de...目的:探讨创伤性臂丛神经损伤的磁共振成像(MRI)三维循环相位稳态采集快速成像(three-dimensional fast imaging employing steady state acquisition with cycled phases,3D-FIESTA-C)序列和三点法非对称回波水脂分离成像(iterative decomposition of water and fat with echo asymmetry and least squares estimation,IDEAL)序列特征及诊断价值。方法:对32例创伤性臂丛神经损伤患者进行术前MRI 3D-FIESTA-C及IDEAL序列检查后,再行图像后处理及诊断,总结臂丛神经损伤的MRI特征,将诊断结果与手术探查结果进行比较,评价术前MRI 3D-FIESTA-C联合IDEAL序列检查在诊断臂丛神经损伤中的作用。结果:12例患者术后证实共有39条节前神经损伤,3D-FIESTA-C序列显示出38条,其中31条神经根影像提示消失或离断,7条神经根丝减少、迂曲,无法连续追踪至椎间孔,同时可伴有神经根袖变钝、脊膜囊肿形成、硬脊膜增厚及脊髓形态信号异常改变。术前MRI 3D-FIESTA-C序列诊断臂丛神经节前损伤的灵敏度为97.5%,特异度为100%,准确率达98.3%。31例120条节后损伤中共12例45条节后神经断裂,其中7条伴纤维瘤形成,IDEAL序列显示42条节后神经根断裂,其他损伤表现为神经增粗,扭曲,走行僵直,周围结构水肿紊乱等。术前IDEAL序列诊断臂丛节后损伤的灵敏度为96.7%,特异度为100%,准确率为97.1%。结论:MRI3D-FIESTA-C联合IDEAL序列检查可清晰显示节前及节后臂丛神经损伤情况,对臂丛神经损伤的诊断符合率较高,可作为临床首选的无创性影像学方法。展开更多
文摘随着现代电力系统的迅猛发展,电网结构和运行方式日益复杂,对状态估计的实时性和准确性也提出了更高的要求。为此,该文提出一种基于深度神经网络的电力系统快速状态估计,通过相关性分析筛选出该状态估计模型的输入量测集,进一步利用海量历史数据建立基于深度神经网络的状态估计模型。当电力系统的实时量测更新时,将强相关量测输入已建立的状态估计模型中快速获得系统状态的估计结果。通过在IEEE标准系统和某实际省网进行算例仿真表明,所提方法的估计精度和鲁棒性均优于传统加权最小二乘(weighted least square,WLS)和加权最小绝对值估计(weighted least absolute value,WLAV);并且该方法的在线计算时间受系统规模影响较小,由实际省网的仿真结果可知,其计算效率较WLS和WLAV分别提升1.43和27.2倍。
基金supported by the National Key Research and Development Program of China under Grant 2018YFB0904500。
文摘The increasing penetration of renewable energy resources with highly fluctuating outputs has placed increasing concern on the accuracy and timeliness of electric power system state estimation(SE).Meanwhile,we note that only a fraction of system states fluctuate at the millisecond level and require to be updated.As such,refreshing only those states with significant variation would enhance the computational efficiency of SE and make fast-continuous update of states possible.However,this is difficult to achieve with conventional SE methods,which generally refresh states of the entire system every 4–5 s.In this context,we propose a local hybrid linear SE framework using stream processing,in which synchronized measurements received from phasor measurement units(PMUs),and trigger/timingmode measurements received from remote terminal units(RTUs)are used to update the associated local states.Moreover,the measurement update process efficiency and timeliness are enhanced by proposing a trigger measurement-based fast dynamic partitioning algorithm for determining the areas of the system with states requiring recalculation.In particular,non-iterative hybrid linear formulations with both RTUs and PMUs are employed to solve the local SE problem.The timeliness,accuracy,and computational efficiency of the proposed method are demonstrated by extensive simulations based on IEEE 118-,300-,and 2383-bus systems.
文摘目前在电力系统中无法保证相量量测单元完全覆盖的情况下,状态估计需要采用相量量测单元(phasor measurement unit, PMU)与数据采集与监控(supervisory control and data acquisition, SCADA)混合量测进行传统非线性状态估计,但是SCADA数据精度低,含有较多不良数据,同时混合数据需要迭代求解,会导致计算效率低且存在截断误差。针对该问题,文章提出了一种基于堆叠去噪自编码器(stack denoising autoencoder, SDAE)与极限学习机(extreme learning machine, ELM)伪量测建模的电力系统高容错快速状态估计方法。其将含有不良量测的SCADA量测数据作为SDAE-ELM伪量测模型的输入,节点电压实部与虚部作为输出,根据历史数据进行训练得到伪量测值与伪量测误差模型,训练完成后得到精度较高的伪量测;将伪量测与PMU量测一起进行快速的线性状态估计。仿真结果表明,所提方法在保证估计精度的基础上,提高了计算效率,验证了所提方法的有效性。
基金supported in part by the National Key Research and Development Plan of China(No.2018YFB0904200)in part by the National Natural Science Foundation of China(No.51725703).
文摘This paper introduces a robust sparse recovery model for compressing bad data and state estimation(SE),based on a revised multi-stage convex relaxation(R-Capped-L1)model.To improve the calculation efficiency,a fast decoupled solution is adopted.The proposed method can be used for three-phase unbalanced distribution networks with both phasor measurement unit and remote terminal unit measurements.The robustness and the computational efficiency of the R-Capped-Ll model with fast decoupled solution are compared with some popular SE methods by numerical tests on several three-phase distribution networks.
文摘目的:探讨创伤性臂丛神经损伤的磁共振成像(MRI)三维循环相位稳态采集快速成像(three-dimensional fast imaging employing steady state acquisition with cycled phases,3D-FIESTA-C)序列和三点法非对称回波水脂分离成像(iterative decomposition of water and fat with echo asymmetry and least squares estimation,IDEAL)序列特征及诊断价值。方法:对32例创伤性臂丛神经损伤患者进行术前MRI 3D-FIESTA-C及IDEAL序列检查后,再行图像后处理及诊断,总结臂丛神经损伤的MRI特征,将诊断结果与手术探查结果进行比较,评价术前MRI 3D-FIESTA-C联合IDEAL序列检查在诊断臂丛神经损伤中的作用。结果:12例患者术后证实共有39条节前神经损伤,3D-FIESTA-C序列显示出38条,其中31条神经根影像提示消失或离断,7条神经根丝减少、迂曲,无法连续追踪至椎间孔,同时可伴有神经根袖变钝、脊膜囊肿形成、硬脊膜增厚及脊髓形态信号异常改变。术前MRI 3D-FIESTA-C序列诊断臂丛神经节前损伤的灵敏度为97.5%,特异度为100%,准确率达98.3%。31例120条节后损伤中共12例45条节后神经断裂,其中7条伴纤维瘤形成,IDEAL序列显示42条节后神经根断裂,其他损伤表现为神经增粗,扭曲,走行僵直,周围结构水肿紊乱等。术前IDEAL序列诊断臂丛节后损伤的灵敏度为96.7%,特异度为100%,准确率为97.1%。结论:MRI3D-FIESTA-C联合IDEAL序列检查可清晰显示节前及节后臂丛神经损伤情况,对臂丛神经损伤的诊断符合率较高,可作为临床首选的无创性影像学方法。
