Abstract
Power system is a highly interconnected network which delivers electric power to electricity users. The size of this network varies from small isolated power systems to very large networks like the Power Grid. Sustaining the secure and reliable delivery of electric power requires continuous monitoring of the system. Measuring devices are installed at many locations in the system and the power characteristics including voltage, current, phase angle, and frequency are recorded continuously. To process the large volumes of data obtained from these measuring devices, it is essential to investigate effective data enhancement techniques. Various aspects of data mining including clustering, classification, regression and pattern finding are utilized to effectively manage the data and extract useful information from it.
Neural networks (NNET), support vector machine (SVM), k-means clustering, decision tree and visualization techniques are some of the popular algorithms in data mining which are frequently integrated with the power system studies. Application of an integrating technique combining two or more of these algorithms is also common in power system studies. Data mining offers effective solutions with low computation complexity and high performance to challenging problems in many field of power system including, but not limited to, stability analysis, fault detection, catastrophe prediction, load forecasting, and power system visualization.
Actually, Big Data technology has already been successfully applied as a powerful data-driven tool for solving numerous new challenges in power grid, such as price forecasting [7, 8], load forecasting [9], transient stability assessment [10], outlier detection [11], and fault detection and analysis [12], among others [13, 14]. We also want to get the rectified output from the Big data Technology and work on following areas:
Price Forecasting
Load Forecasting
Transient stability assessment
Outlier detection
Fault detection and analysis