Abstract:With the development of power monitoring technology, more and more real-time data are accumulated in power system. Due to large amount and high dimension of monitoring data, traditional data analysis methods are often unable to discover the hidden rules and complex relationships between monitoring data and fault reasons. In recent years, data mining techniques have been successfully applied in many fields. Compared with traditional statistical methods and commonly used expert system, data-driven based methods also show excellent performance in many fault detection systems. However, most of these methods can only detect the pre-determined faults while ignoring the new emerging ones. And in practice, these new types of power plant failures often result in greater losses. In this paper, we integrate various data mining techniques to analyze the power plant fault and introduce a novelty detection approach. Specifically, Random forest is used to predict power plant fault types and rank important features. Assembling K-means and correlation analysis is applied to further discover the correlation between the faults and features. A new approach RFP, which is proposed in our recent work, is used to discover the new emerging faults. Experiments on a real power plant dataset show that our proposed data-driven approaches can find the important impact factors with plant fault and detect the new emerging fault effectively.

Applications of data mining time series to power systems disturbance analysis

Application of Atomic Sparse Decomposition to Power Systems Disturbance Analysis

A System of Time Series Data Mining Based on Hybrid Model

Application of Hilbert-Huang Transform in Power System Fault Detection

Anomaly Detection Algorithm Based on Electric Equipment

Analysis of short time power quality disturbances based on statistic feature vector symbolic and ensemble empirical mode decomposition

A Fault Diagnosing Method in Power Transmission Systems Based on Time Series Similarity Matching

Wide-Area Time Series Data Mining Based Transient Voltage Stability Assessment

Data-driven Solutions for Power System Fault Analysis and Novelty Detection.

Pattern Mining and Anomaly Detection Based on Power System Synchrophasor Measurements

Outlier Detection Method Based on Compressed Time Series and Voronoi Diagram for Power Data

A NEW METHOD OF DISTURBANCE DETECTION IN POWER SYSTEMS ASED ON HALF-COMPLEX WAVELET TRANSFORM

A Data-Driven Fault Detection Framework Using Mahalanobis Distance Based Dynamic Time Warping

Application of Data Mining Method based on Deep Learning in Power System

Data Mining and Machine Learning for Power System Monitoring, Understanding, and Impact Evaluation

A novel intelligent system for analysis and recognition of power quality disturbance signal

Data Mining And Analysis Of Tree-Caused Faults In Power Distribution Systems

Power system transient security assessment based on multi-channel time series data mining

Application of Data Mining Technology in Traction Power Supply System of Railway

Distinguishing ground faults in different feeders of power distribution networks using fractal information

Time-frequency Analysis for Power Transformer Fault Detection Using Vibration Method