Research On State Evaluation And Fault Diagnosis Of The Protection Trip Circuit And Equipment In Power System Based On Data-driven

In the context of the ever-increasing complexity of China’s power system,if the protection system cannot quickly and selectively remove faults,it may lead to the occurrence of cascading faults,resulting in serious economic losses.The operating circuit(OC)and the circuit breaker(CB)operating mechanism of the power system protection trip equipment still lack complete self-checking measures,and it is difficult to monitor their operating status online.Therefore,the protection trip circuit and equipment abnormality has become an important cause of incorrect operation of the protection system.In order to effectively solve the above problems and improve the operation reliability and maintenance management level of the protection system,this thesis systematically studied the power system protection trip circuit and equipment online monitoring,state evaluation,and fault diagnosis technology,and specifically carried out the following aspects:(1)Based on the opening/closing coil current(CC)waveform characteristics of OC,an OC online monitoring method is proposed,which can monitor the operating status and circuit integrity of OC in real-time,identify the branch where the disconnection fault is located,and reveal the abnormality of the CB auxiliary contact.Based on the dynamic fault tree algorithm,an OC failure model considering design redundancy and condition-based data is established,and an online OC state evaluation method is proposed.(2)A fault diagnosis method for CBs based on optimized affinity propagation clustering is proposed.The improved similarity matrix is constructed using the waveform signal of the opening/closing CC,and the improvement of the traditional clustering algorithm is realized by the introduction of the parameter optimization process.The similarity coefficient is defined to determine the type of sample failure,which effectively solves the problem that traditional methods cannot identify samples with unknown failure types,and this method has a higher diagnostic accuracy rate when there are fewer training samples.(3)A data-driven method for determining the maintenance priority of CBs is proposed.Combining the opening/closing CC waveform signal and the CB failure severity indicators,the CB maintenance feature vector is established,and the of ranking function the feature vector is calculated through the fuzzy C-means clustering and the learning to rank algorithm.This method can obtain a reasonable maintenance strategy when there is little prior knowledge,and effectively reduces the influence of factors such as data randomness and measurement uncertainty on maintenance decisionmaking.(4)A predictive maintenance decision-making method for CBs based on real-time state evaluation and remaining useful life calculation is proposed.Through the CC waveform signal,the degradation index considering the measurement uncertainty is constructed,and then the CB degradation model considering the time-varying uncertainty of the degradation process is established.The Sequential Bayesian algorithm combined with historical records and monitoring data is used to realize the real-time update of the CB state evaluation and remaining useful life calculation.The objective function of predictive maintenance decision that considers economy and reliability is obtained,which can provide data basis for the condition-based maintenance of CBs.This thesis focuses on the research of online monitoring,state evaluation,fault diagnosis,and prediction of the power system protection trip circuit and equipment.From the three aspects of break maintenance,planned maintenance,and conditionbased maintenance of the protection trip circuit and equipment,data-driven algorithms are applied to realize the knowledge mining and information combination of historical data and online monitoring data.On this basis,the power system protection trip circuit and equipment online state evaluation and fault diagnosis system is designed,and a complete power system protection trip circuit and equipment state evaluation and fault diagnosis research framework is formed,which provides technical support for optimizing equipment operation and maintenance scheme and effectively improves the level of equipment management.The research in this thesis is of great significance for maintaining the reliability and stability of power grid operation.