A Non-Intrusive Electric Load Component Identification Method Based on Event Detection

The electric load model is very important for the study of power systems. Over the past few years, with the widespread application of power electronics technology, the power electronics interfaced load have been added to the main components of the load model based on the original static load and motor. The power proportion of them is a key factor affecting the accuracy of the load model. Most of the existing noninvasive load identification methods use artificial neural networks to identify specific equipment, which requires a large amount of training data and has complicated algorithms. This paper proposes a non-intrusive electric load component identification method based on event detection. The load connecting to the power grid is monitored by an event detection algorithm, and the characteristics of the active power impact and reactive power change are used to identify the load component. This method only needs to acquire the active power and reactive power when the electric load is in stable operation and has lower requirements for the acquisition equipment. In addition, the identification process does not involve complicated algorithms, is easy to implement at low cost, and possesses good precision.