A decision tree-based method for power system fault diagnosis by synchronized Phasor Measurements

This paper provides initial justification for adopting a distributed architecture for phasor measurement unit. The characteristics of synchronized phasor measurement is studied in detail. The benefits and short comings of phasor measurements are highlighted from the point of view of their different applications. The need for extension of phasor measurement unit functionality is brought out through analysis of the application and synchronized phasor measurement characteristic compatibility. The concept of phasor measurement unit is extended to a distributed system with several small functions working independently to provide a complete functionality of a phasor measurement unit. The proposed substation automation works on the framework of IEC 61850 process bus.

As the node power is controllable, the traditional dynamic security region (DSR) is defined on power injection space. The phasor measurement unit (PMU) can measure and transfer the state variables (phase angles) of power system synchronously. How to make full use of the synchronized phasor measurement (SPM) in the security monitoring, protection and control of power system is a challenging research task. In this paper, we presented the concept of practical dynamic security region (PDSR) in phase angle space. A study on PDSR of power systems shows that critical points satisfying transient stability in angle space can still be described by one or a few critical hyper- planes. With the PDSR in the node angle space, a concise expression to calculate the transient stability margin of a current operation point in each direction in node angle space is suggested. Besides, this paper suggested an approach to acquire the PDSR in low-dimensional angle space based on dynamic equivalence of transient stability and a method to select appropriate locations for partially equipping PMU, which provide an innovative way of making use of the PMU information to carry out dynamic security monitoring of power systems.

A relatively perfect system for the fault diagnosis of mechanical and electrical products has been formed through decades of development. Nevertheless, the traditional fault diagnosis methods fail to cope with the gradual huge mechanical and electrical system. As a result, the advantages of fault diagnosis mode driven by data are increasingly prominent. Meanwhile, the effect of fault diagnosis has exceeded the traditional fault diagnosis methods in many fields. Through the use of the deep learning technology based on artificial intelligence, it carries out mapping and fitting. By fully taking advantages of neural network, it can effectively obtain the accurate classification of fault data. A fault diagnosis method based on the fault data of mechanical and electrical system is designed in this thesis. When it comes to the basic process, it is to take data sets for different mechanical and electrical products. Through the use of feature engineering method, it extracts the fault features of data. Through the use of deep learning technology, it carries out the intelligent diagnosis. According to the experimental results, it indicates that the fault diagnosis method based on deep learning technology can distinguish a variety of fault modes in mechanical and electrical system in an effective way. What’s more, good classification results in fault recognition have been achieved by a variety of deep convolutional neural network structures, so the feasibility of the method is further verified.

Voltage instability has caused a great deal of concern in the existing power systems. Despite best efforts, it is still not uncommon a phenomenon and has caused some of the major blackouts and catastrophic failures in the recent past, resulting in huge social and economic losses. The advent of synchrophasor technology has made possible wide-area measurements in real-time, and has found huge applications in power systems. The identification of subregions in power systems that experience a unique voltage instability problem is one of the most important steps of voltage stability analysis. This paper presents a method to identify voltage control area (VCA), based on coherent groups of buses, using system states obtained from synchronized phasor measurements. The coherency is identified by applying hierarchical clustering, a machine learning technique. The coherent buses are identified by applying the machine learning technique on the angles obtained from bus voltage phasors. The results so obtained using the data from Phasor Measurement Units (PMUs) on 10-machine, 39-bus New England power system model are presented.

The electrical power system is becoming more and more complex with the increasing demand of the electricity across the world, as the world is becoming the global village. The electrical power system must be reliable and safe enough to fulfil the energy requirements with the continuous supply of energy. The chances of power system's blackouts and outages are increasing with its increased complexity. So, there is a need of an efficient control system to make the power system more safe, efficient and reliable. Wide Area Monitoring, Protection and Control system (WAMPAC) is becoming an emerging technology for an efficient monitoring, protection and control of power system. Phasor measurement unit (PMU) is an integrated part of wide area monitoring, protection and control (WAMPAC) system, which can be used for monitoring, protection and control of complex electrical power systems. PMU gives the synchronized phasor measurements of voltage and current and can easily monitor and control even small disturbances in power system to protect the power system from any blackouts and outages before any fault occurs. In the proposed paper a PMU model is designed in MATLAB/Simulink and then PMU is used for power system protection and the comparative analysis are done with the conventional protection technique for an interconnected two-area network power system.

The fault diagnosis method based on neural network has many defects, such as complicated network, long training time and slow convergence speed. The particle swarm optimization and neural network integration fault diagnosis methods are proposed to improve the fault diagnosis capability. Firstly, the heuristic global optimization capability of particle swarm optimization is used to optimize the neural network connection weights; then the transformer fault samples are trained and tested by using non-linear processing capacity of neural network. The test results show that such algorithm can effectively avoid unstable neural network, easily falling into local minimum and lower diagnostic accuracy etc. and can effectively increase the convergence speed and fault diagnosis efficiency compared with traditional fault diagnosis method.

In uses, such as stability surveillance of an electric power system, the synchronous voltage current phasor measurement using GPS satellite base timing keeping technology is demanded. In the past, the synchronous phasor measurement technique of a voltage current based in DFT conversion theory was proposed, and it was established as an IEC standard[1]. In the actual electric power system, the frequency of voltage in each place is delicately different, and it is always changing. Since the sampling data in the time window of fixed width cannot take in the integer cycle of a voltage waveform and does not satisfy the signal periodicity conditions of Fourier transform, a spectrum leakage phenomenon occurs. By this, the accuracy of the synchronous phasor measurement by single DFT-BIN frequency falls. This paper proposes the method of realizing recursively DFT conversion with a window function [RWDFT] which can reduce the influence of system voltage frequency change. Moreover, the method of calculating Phasor of true signal frequency is proposed by compounding three or more DFT conversion factors. In simulation verification, the synchronous phasor measuring method which applied the above two methods has checked that a highly precise Phasor measurement result was obtained in system voltage frequency change.

An analytic model for fault diagnosis of power system using optimization technique is expressed as unconstrained 0-1 integer programming problem, and consequently faulty equipment identification can be solved by refined mathematical operation. Considering the configuration of automatic devices in modern power systems, such as protective relays and reclosing relays, an improved analytic model and optimization technique-based method for fault diagnosis of power system is proposed in this paper. The evaluation criteria of the presented model is improved considering the relationship of multiple main protective relays, backup protective relays, malfunctioning protective relays and reclosing relays. Improvements of analytic model for fault diagnosis of electric power system based on optimization techniques are presented firstly. A brief description about the modulars and functions of the online fault diagnosis software which is developed by the authors for Jiangsu Provincial Power Company is given. The adopted EMS data acquisition method and simulated online test results for the power system of Jiangsu Power Company are described.

Rolling bearings are vital components in rotary machines where even minor faults can lead to the collapse of the entire mechanical system, resulting in significant economic losses. The condition monitoring and fault diagnosis of rolling bearings are crucial for safe operation of complex industrial systems. Traditional fault diagnosis methods mainly depend on expert experience, which leads to obvious deficiencies in diagnostic efficiency, especially in terms of accuracy. To address the issues of traditional bearing fault diagnosis relying on expert experience and low efficiency and accuracy of fault diagnosis, a bearing fault diagnosis model combining synchrosqueezed wavelet transform (SWT) and SE-ResNet is proposed. Firstly, the one-dimensional non-stationary bearing vibration signal is converted into a high-frequency two-dimensional time-frequency image through SWT, which serves as the input for the convolutional neural network; Secondly, an attention mechanism is introduced to construct an SE-ResNet model for extracting two-dimensional image features, which enhances the attention of important features and thus enhances the network's representation ability; Finally, the results of fault diagnosis are output through Softmax layer. To validate the capacity of the proposed model, experiments with the bearing dataset collected from Case Western Reserve University (CWRU) were conducted. The results show that the transfer learning accuracy of the CWRU dataset is 86.25% to 99%. This method has good feasibility, good convergence performance, and has high practical application value.

The products of a batch process have high economic value. Meanwhile, a batch process involves complex chemicals and equipment. The variability of its operation leads to a high failure rate. Therefore, early fault diagnosis of batch processes is of great significance. Usually, the available information of the sensor data in batch processing is obscured by its noise. The multistage variation of data results in poor diagnostic performance. This paper constructed a standardized method to enlarge fault information as well as a batch fault diagnosis method based on trend analysis. First, an adaptive standardization based on the time window was created; second, utilizing quadratic fitting, we extracted a data trend under the window; third, a new trend recognition method based on the Euclidean distance calculation principle was composed. The method was verified in penicillin fermentation. We constructed two test datasets: one based on an existing batch, and one based on an unknown batch. The average diagnostic rate of each group was 100% and 87.5%; the mean diagnosis time was the same; 0.2083 h. Compared with traditional fault diagnosis methods, this algorithm has better fault diagnosis ability and feature extraction ability.

To find out the hidden danger in the industrial production process in time, it is necessary to monitor the health condition of the key components of the mechanical system in operation. However, traditional fault diagnosis methods usually adopt manual feature extraction, which not only costs expensively and depends on prior knowledge. Therefore, it is of great significance to study the application of automatic fault identification based on the original vibration signals. Recently, existing studies have shown that most of fault diagnoses are implemented by using deep neural network. Although these methods have achieved satisfactory performances, there are obvious limitations in real applications, that is, the complexity of deep neural network requires a lot of hardware computing resources. This hinders the development of online fault diagnosis tools. To solve this problem, this paper proposes a fault diagnosis model based on lightweight convolutional neural network MobileNet, and realizes an end-to-end intelligent fault classification and diagnosis application. We evaluated the proposed method with the rolling bearing dataset from Western Reserve University. The best average precision, recall and F1 score of ten different bearing health conditions are about 96%, 82% and 88%, respectively. In addition, we also compare the accuracy of the rolling bearing fault diagnosis classification model under the standard ReLU and the improved ReLU. Experimental results show that both obtain good performances, but the improved ReLU reaches the over 96% accuracy more rapidly.

The PMU (Phasor Measurement Unit) is used for accurate measurement of synchronous phasor of voltages and currents. Based on synchronous phasor continuous measurement at the different location of distribution network is possible prevent fault which can cause blackout. Synchronous phasor measurement is described in IEEE C37.118-2005 standard. The IEEE C37.118-2005 is primarily focused on communication with the PMU and precision of the measurement at steady state. PMU manufacturers declare compliance with IEEE C37.1182005, which doesn't mean correct behaviour of the PMU for dynamic changes in the power network. The new standard for the PMU is currently developing. This standard should contain both the issues included in IEEE C37.118-2005 and description of PMU testing in connection with dynamic changes in power network. This article deals with the implementation of tests for testing the dynamic properties of PMU. These tests are conducted by tester, which was designed and implemented based on virtual instrumentation. The results of several tests of commercial PMUs are presented in the article. (4 pages)

The traditional fault diagnosis method only uses the protection and circuit breaker alarm messages to determine the fault sections. This paper put forward a new power grid fault track method based on big data platform. It extends the data source to transformer substation and deal with the fault date by the technology of Spark. This method can solve the problem of mass data processing. This paper analyses fault information by data mining and track the malfunction of protection or circuit breaker by decision tree. It can give the reason of fault and optimize the function of fault diagnosis system.

Power transformers are vital components of electrical power systems, ensuring reliable and efficient energy transfer between high-voltage transmission and low-voltage distribution networks. However, they are prone to various faults, such as insulation breakdowns, winding deformations, partial discharges, and short circuits, which can disrupt electrical service, incur significant economic losses, and pose safety risks. Traditional fault diagnosis methods, including visual inspection, dissolved gas analysis (DGA), and thermal imaging, face challenges such as subjectivity, intermittent data collection, and reliance on expert interpretation. To address these limitations, this paper proposes a novel distributed approach for multi-fault diagnosis of power transformers based on a self-organizing neural network combined with data augmentation and incremental learning techniques. The proposed framework addresses critical challenges, including data quality issues, computational complexity, and the need for real-time adaptability. Data cleaning and preprocessing techniques improve the reliability of input data, while data augmentation generates synthetic samples to mitigate data imbalance and enhance the recognition of rare fault patterns. A two-stage classification model integrates unsupervised and supervised learning, with k-means clustering applied in the first stage for initial fault categorization, followed by a self-organizing neural network in the second stage for refined fault diagnosis. The self-organizing neural network dynamically suppresses inactive nodes and optimizes its training parameter set, reducing computational complexity without sacrificing accuracy. Additionally, incremental learning enables the model to continuously adapt to new fault scenarios without modifying its architecture, ensuring real-time performance and adaptability across diverse operational conditions. Experimental validation demonstrates the effectiveness of the proposed method in achieving accurate, efficient, and adaptive fault diagnosis for power transformers, outperforming traditional and conventional machine learning approaches. This work provides a robust framework for integrating advanced machine learning techniques into power system monitoring, paving the way for automated, real-time, and reliable transformer fault diagnosis systems.

Traditional mechanical fault diagnosis methods for high-voltage circuit breakers (CBs) largely rely on data-driven learning from a substantial amount of labeled fault samples. However, the scarcity of target fault samples in practical engineering applications often limits diagnostic performance, leading to high misdiagnosis rates and poor generalization capabilities. To address these challenges, this study proposes an attribute embedding zero-shot diagnosis (AEZSD) method, designed to overcome the limitations of sample insufficiency. Initially, this paper utilizes phase space reconstruction techniques to thoroughly explore the intrinsic dynamic features of vibrational signals within CBs. Subsequently, by integrating the electromechanical signal characteristics of the CBs, the concept of fault attributes is introduced, and an attribute embedding learning network is constructed. Through this network and statistical rules, the proposed method can effectively identify previously unseen fault types. Experimental results confirm that the AEZSD method can leverage historical fault data to pre-learn fault attribute knowledge and accurately diagnose faults without target fault samples, providing a novel solution for CB fault diagnosis.