Research on Flashover Voltage Prediction of Catenary Insulator Based on CaSO_(4) Pollution with Different Mass Fraction

Pollution flashover accidents occur frequently in railway OCS in saline-alkali areas.To accurately predict the pollution flashover voltage of insulators,a pollution flashover warning should be made in advance.According to the operating environment of insulators along the Qinghai-Tibet railway,the pollution flashover experiments were designed for the cantilever composite insulator FQBG-25/12.Through the experiments,the flashover voltage under the influence of soluble contaminant density(SCD)of different pollution components,non-soluble deposit density(NSDD),temperature(T),and atmospheric pressure(P)was obtained.On this basis,the GA-BP neural network prediction model was established.P,SCD,NSDD,CaSO_(4) mass fraction(w(CaSO_(4))),and T were taken as input parameters,50%flashover voltage(U_(50%))of the insulator was taken as output parameters.The results showed that the prediction deviation was less than 10%,which meets the basic engineering requirements.The results could not only provide early warning for the anti-pollution flashover work of the railway power supply department,but also be used as an auxiliary contrast to verify the accuracy of the results of the experiments,and provide a theoretical basis for the classification of pollution levels in different regions.

AC and Lightning Breakdown Voltage Prediction Based on PD Measurements

Partial discharge(PD)measurements are a standard method to determine insulation integrity since many years.For new equipment,the partial discharge level should be below a certain standardized level to be commissioned successfully.However,what is when a monitoring system detects upcoming partial discharges during the lifetime of an electrical equipment?Unfortunately,the discharge magnitude is not directly proportional to the remaining lifetime or the breakdown risk or breakdown voltage.Expert systems or experienced professionals can identify the PD defect root cause with a good certainty.This helps to determine the given risk.Nevertheless,clear risk quantification is missing.In this paper,a new approach is presented to predict the AC and lightning breakdown voltages of the equipment based on patterns from PD measurements.The method is validated with PD test data of several tip-plate configurations in air.A neuronal network is trained with these measurements.For control measurements with a different tip,it can be shown that the breakdown voltage can be predicted with an average failure of 5.3%for AC and 9.1%for lightning.

Study of predicting breakdown voltage of stator insulation in generator based on BP neural network

The breakdown voltage plays an important role in evaluating residual life of stator insulation in generator.In this paper,we discussed BP neural network that was used to predict the breakdown voltage of stator insulation in generator of 300MW/18kV.At first the neural network has been trained by the samples that include the varieties of dielectric loss factor tanδ,the partial discharge parameters and breakdown voltage.Then we tried to predict the breakdown voltage of samples and stator insulations subjected to multi-stress aging by the trained neural network.We found that it's feasible and accurate to predict the voltage.This method can be applied to predict breakdown voltage of other generators which have the same insulation structure and material.

The improved PAC method for a three-phase PWM grid-connected inverter

In this paper, a vector regulating principle of the phase and amplitude control PAC method for three-phase grid-connected inverters is presented.To solve the problem of heavy inrush current and slow dynamic response when system starts up, the starting voltage prediction control and the current feed-forward control are proposed and used, which improve the dynamic performance of the system in the PAC.The experimental results carried out on a three-phase grid-connected inverter proved the validity of the proposed method.

Prediction of voltage distribution using deep learning and identified key smart meter locations

The energy landscape for the Low-Voltage(LV)networks is undergoing rapid changes.These changes are driven by the increased penetration of distributed Low Carbon Technologies,both on the generation side(i.e.adoption of micro-renewables)and demand side(i.e.electric vehicle charging).The previously passive‘fit-and-forget’approach to LV network management is becoming increasing inefficient to ensure its effective operation.A more agile approach to operation and planning is needed,that includes pro-active prediction and mitigation of risks to local sub-networks(such as risk of voltage deviations out of legal limits).The mass rollout of smart meters(SMs)and advances in metering infrastructure holds the promise for smarter network management.However,many of the proposed methods require full observability,yet the expectation of being able to collect complete,error free data from every smart meter is unrealistic in operational reality.Furthermore,the smart meter(SM)roll-out has encountered significant issues,with the current voluntary nature of installation in the UK and in many other countries resulting in low-likelihood of full SM coverage for all LV networks.Even with a comprehensive SM roll-out privacy restrictions,constrain data availability from meters.To address these issues,this paper proposes the use of a Deep Learning Neural Network architecture to predict the voltage distribution with partial SM coverage on actual network operator LV circuits.The results show that SM measurements from key locations are sufficient for effective prediction of the voltage distribution,even without the use of the high granularity personal power demand data from individual customers.

Data-driven Predictive Voltage Control for Distributed Energy Storage in Active Distribution Networks

Integration of distributed energy storage(DES)is beneficial for mitigating voltage fluctuations in highly distributed generator(DG)-penetrated active distribution networks(ADNs).Based on an accurate physical model of ADN,conventional model-based methods can realize optimal control of DES.However,absence of network parameters and complex operational states of ADN poses challenges to model-based methods.This paper proposes a data-driven predictive voltage control method for DES.First,considering time-series constraints,a data-driven predictive control model is formulated for DES by using measurement data.Then,a data-driven coordination method is proposed for DES and DGs in each area.Through boundary information interaction,voltage mitigation effects can be improved by interarea coordination control.Finally,control performance is tested on a modified IEEE 33-node test case.Case studies demonstrate that by fully utilizing multi-source data,the proposed predictive control method can effectively regulate DES and DGs to mitigate voltage violations.