In recent years, applications of artificial intelligence (AI) techniques in fault diagnosis high-voltage circuit breakers (HVCBs) have gained wide attention. real applications, how-ever, HVCBs work the normal state most time. Therefore, problem imbalanced monitoring data is prevalent, which threatens generalization capability AI-based methods, resulting poor diagnostic performance. To address this problem, an oversampling method called Density-weighted Minority Oversampling (DWMO) was...

Mechanical fault diagnosis of a circuit breaker can help improve the reliability power systems. Therefore, new method based on multiscale entropy (MSE) and support vector machine (SVM) is proposed to diagnose in high voltage breakers. First, Variational Mode Decomposition (VMD) used process breaker's vibration signals, reconstructed signal eliminate effect noise. Second, calculated selected as feature vector. Finally, vector, identification classification are realized by SVM. The constructed...

The fast spectrum kurtosis (FSK) algorithm can adaptively identify and select the resonant frequency band extract fault feature via envelope demodulation method. However, FSK method has some limitations due to its susceptibility noise random knocks. To overcome this shortage, a new is proposed in paper. Firstly, we use binary wavelet packet transform (BWPT) instead of finite impulse response (FIR) filter bank as segmentation Following this, Shannon entropy each calculated. appropriate center...

The fast spectrum kurtosis (FSK) algorithm can adaptively identify and select the resonant frequency band extract fault feature by envelope demodulation method. However, in practical applications, source may be located different bands; plus noise interference, weak side of compound is not easy to identified FSK. In order improve accuracy spectral analysis method, a modified method based on maximum correlation deconvolution (MCKD) proposed. According possible characteristic frequencies,...

As high-voltage circuit breakers (HVCBs) are directly related to the safety and stability of a power grid, it is great significance carry out fault diagnoses HVCBs. To accurately identify operating states HVCBs, novel mechanical diagnosis method HVCBs based on multi-feature entropy fusion (MFEF) hybrid classifier proposed. MFEF involves decomposition vibration signals into several intrinsic mode functions using variational (VMD) calculation by integration three Shannon entropies. Principle...

The periodic impacts are regarded as the typical characteristics of local defect wind turbine bearing. For this reason, it is significant to extract from original measured vibration signal with background noise interferences during identification process. purpose effectively solving issue, an innovative diagnostic frame based on Lkurtogram guided adaptive empirical wavelet transform (LGAEWT) and purified instantaneous energy operation (PIEO) put forward. Within frame, L-kurtosis indicator...

Abstract In this study, the rotor loss as well temperature variations due to static air‐gap eccentricity (SAGE) and inter‐turn short circuit (RISC) in synchronous generators is presented. Different from previous studies, study investigates loss/temperature changes under not only single SAGE fault RISC but also combined faults (CF) that are composed of these two faults. Detailed formula derivation on hysteresis (HL) eddy current (ECL), together with rise varied types different operating...

The key step of bearing fault diagnosis is to select a suitable resonance frequency band, so as filter out interference components the maximum extent and retain information in band. Kurtogram algorithm can locate band well, which has been widely researched applied recent years, produced many derivative algorithms. statistical indicators used by these methods identify features are divided into time domain indicators. Time more sensitive single accidental impact components, while easily...

A gas relay in the transformer, an important non-electricity protection equipment, only triggers stem leaf spring for serious alarm when baffle rotates to a certain angle, which cannot judge transformer fault event of heavy before running state and development process, this paper puts forward kind based on panel corner monitoring methods. The partial structure is improved, signal extracted by installing Hall sensor back baffle. On basis, develops test stand capable simulating characteristics...

Mechanical characteristic parameter can be used to evaluate circuit breaker performance, traditional test method for measuring cannot applied on-line, also the application process is complicated. To solve this problem, a new Timing parameters extraction from vibration signal of in opening proposed. First, based on principle spring operating mechanism, event caused by components impact analyzed. Then, order eliminate noise, variation mode decomposition adopted decompose signal, reconstruct...

The high voltage circuit breaker (CB) play an important role in the power systems, so fault diagnosis of CB has great significance. A new method based on timing parameters and Fuzzy C-means clustering (FCM) is proposed this paper. First, cause vibration signal analyzed theoretically, it found that events will be change different states. Then short time energy entropy ratio calculated, which can enhance impact feature event. occurrence end are extracted as vector by double threshold method....

A new fault diagnosis method based on variational mode decomposition (VMD) is proposed in order to realize the identification of high voltage switches. Firstly, after preprocessing vibration signal, feature extraction VMD energy entropy analyzed. Subsequently, SVM utilized calculate type by utilizing extracted features. Parameter optimization solved grid search algorithm improve performance SVM. Verification scientific nature this for provided field data, which was obtained through an...

Abstract The paper proposes a novel approach to diagnose circuit breaker faults by integrating OCSSA, VMD, CNN, and BiLSTM algorithms. Firstly, OCSSA optimizes the parameters of VMD reduce noise interference. Then, decomposes operating signals into characteristic signals. Secondly, CNN captures mechanical fault characteristics from decomposed Finally, applies CNN’s results for feature classification. Experimental show that combining improved with CNN-BiLSTM model achieves classification...

Closing spring fatigue faults of high voltage circuit breakers affect the timing parameters in closing operations and reduce performance breaker. Traditional tests parameter based on travel curve cannot be applied online, sensor installation is complicated. In this paper, a new method to extract key breaker from vibration signal under fault proposed. First, simulated Automatic Dynamic Analysis Mechanical Systems (ADAMS). Results indicate that time intervals between points can used as...