溯源弹簧形变过程的断路器振动信号递归量化分析辨识方法

Circuit Breaker State Identification Method Based on Tracing Spring Deformation Process and Recurrence Quantitative Analysis of Vibration Signals

从锁止机构脱扣引起储能弹簧释能,经部件带动动触头运动再到静止的每个动作具有严格阶段特征,伴随断路器动作的机械振动展现了能量传递及设备健康状态。该文提出一种溯源弹簧形变过程的断路器振动信号递归量化分析方法,首先由高速相机捕捉断路器操动时储能弹簧的动作图像,通过计算机视觉跟踪动态提取反映弹簧形变特征帧,再依据特征帧时序划分操动过程;然后将不同阶段振动信号映射至高维相空间,经递归分析得到体现动力系统变化特征的递归图,并递归量化分析其纹理结构特征;最后利用支持向量机模型对正常及故障状态下的断路器振动特征样本进行分析辨识,对比结果证明,由弹簧释能时序细化振动信号特征分析有效提高了分类识别准确率。该文方法在断路器操动机构状态辨识中具有广阔的应用前景。

High-voltage circuit breakers are key electrical equipment for power grid control and protection.As the structure complexity and automation of transmission and distribution networks continue to increase,the requirements for reliable operation are also increasing.Each movement has strict stage characteristics for high-voltage circuit breakers with the spring-operated mechanism,from energy storage spring release caused by the tripping of the locking mechanism to the moving contact driven by components and then to the rest.The mechanical vibration accompanying the action of the circuit breaker indicates the energy transmission and health status of the equipment.This paper proposes a method for the recursive quantitative analysis of vibration signals in high-voltage circuit breakers,focusing on traceability to the spring deformation process.First,a high-speed camera captures the action image of the energy storage spring during circuit breaker operation.A specific target in the high-speed image sequence is dynamically extracted through a computer vision tracking algorithm(normalized cross-correlation graphic pyramid matching algorithm)to form a spring movement and displacement time curve.Four characteristic frames are defined for spring deformation:deformation start,maximum velocity,maximum deformation,and oscillation end.Secondly,according to the characteristic frame timing of the image sequence,the operation process is accurately divided into four stages:tripping,energy storage release,closing buffer,and oscillation braking.Each stage highlights the changing characteristics of signals,reflecting the correlation between the signal and the actual physical energy transfer state.Finally,vibration signals of different stages are mapped to a high-dimensional phase space.The recurrence plot(RP)reflecting the changing characteristics of the dynamic system is obtained through recursive analysis.In addition,the RQA feature sequence is obtained through recursive quantitative analysis(RQA),allowing for an accurate depiction of the relationship among spring discharge,vibration events,and component motion.Taking a ZN65−12 high-voltage circuit breaker as the research object,a fault simulation experiment platform is built,and the support vector machine model is used to analyze and identify vibration characteristic samples under normal and fault conditions.The experimental results show that the timing of spring energy release refines vibration signals.Feature analysis effectively improves the accuracy of circuit breaker status identification.High-speed image and vibration signal capture are both non-invasive measurement methods.With accumulating experimental data and further algorithm improvement,it is expected to find applications in circuit breaker fault monitoring and live testing.