基于电磁-结构力场的换流变压器绕组结构参数优化及故障诊断OA
Winding Structure Parameter Optimization and Fault Diagnosis of Converter Transformer Based on Electromagnetic-structural Force Field
换流变压器是高压直流输电系统电能变换与传输的关键装备,其运行可靠性对电网整体稳定性至关重要.为此围绕换流变压器绕组结构优化与故障诊断开展系统研究,提出了多场耦合建模-结构参数优化-绕组故障诊断的完整方法体系.首先,构建了电磁-结构力场耦合的换流变压器仿真模型,分析谐波激励下振动特性变化规律;其次,以振动最小值作为优化目标,选取绕组高度、厚度和间距等6个关键结构参数,采用梯度提升回归树模型结合贝叶斯优化算法进行优化,使最大振动位移较初始设计降低11.93%.进而,针对优化后绕组仍可能发生的故障提出了基于振动信号的诊断方法,该方法通过相关性分析获得了最佳的油箱壁面测点位置,构建了绕组典型故障仿真模型,并建立了基于TST-XGBoost的故障诊断模型,最终实现了98.33%的故障识别准确率.研究成果不仅为换流变压器的减振设计提供了有效的方法,也为绕组状态评估与智能诊断提供了理论基础与技术支撑.
Converter transformer is the key equipment for power conversion and transmission in HVDC transmission system,and its operation reliability is very important to the overall stability of power grid.In this paper,the structural op-timization and fault diagnosis of converter transformer windings are systematically studied,and a complete method system of multi-field coupling modeling+structural parameter optimization+winding fault diagnosis is proposed.Firstly,the simulation model of converter transformer with electromagnetic-structural force field coupling is constructed,and the variation law of vibration characteristics under harmonic excitation is analyzed.Secondly,the minimum vibration is taken as the optimization objective,and six key structural parameters,such as winding height,thickness,and spacing,are se-lected.The gradient boosting regression tree model combined with Bayesian optimization algorithm is used for optimization,so that the maximum vibration displacement is reduced by 11.93%compared with the initial design.Fur-thermore,a diagnosis method based on vibration signal is proposed for the faults that may still occur in the optimized winding.By using the method,the best position of the tank wall measuring point can be obtained through correlation analysis,the typical fault simulation model of the winding can be constructed,the fault diagnosis model can be estab-lished based on TST-XGBoost,and the fault recognition accuracy of 98.33%can be realized.The research results not only provide an effective method for the vibration reduction design of converter transformer,but also provide theoretical basis and technical support for winding state evaluation and intelligent diagnosis.
袁发庭;鲍国苗;费瑞昊;肖明月;桂童童;唐波
湖北省输电线路工程技术研究中心(三峡大学),宜昌 443002||三峡大学电气与新能源学院,宜昌 443002三峡大学电气与新能源学院,宜昌 443002三峡大学电气与新能源学院,宜昌 443002三峡大学电气与新能源学院,宜昌 443002三峡大学电气与新能源学院,宜昌 443002湖北省输电线路工程技术研究中心(三峡大学),宜昌 443002||三峡大学电气与新能源学院,宜昌 443002
换流变压器多场耦合谐波激励贝叶斯优化故障诊断
converter transformermulti-field couplingharmonic excitationBayesian optimizationfault diagnos
《高电压技术》 2026 (7)
3338-3349,中插21-中插23,15
国家自然科学基金(52307179)湖北省教育厅科研计划项目(D20231205)湖北省高等学校优秀中青年科技创新团队计划项目(T2024003).Project supported by National Natural Science Foundation of China(52307179),Science and Technology Research Project of Education Department of Hubei Province(D20231205),Outstanding Young and Middle-aged Sci-tech Innovation Team Project of Universities in Hubei Province(T2024003).
评论