A Predictive Method for the Frequency Nadir Based on Convolutional Neural Network
- Resource Type
- Conference
- Authors
- Lin, Jintian; Chen, Longyu; Zhang, Yichao; Chen, Qingyue; Wang, Xiaoru
- Source
- 2021 IEEE 2nd China International Youth Conference on Electrical Engineering (CIYCEE) Electrical Engineering (CIYCEE), 2021 IEEE 2nd China International Youth Conference on. :1-8 Dec, 2021
- Subject
- Power, Energy and Industry Applications
Robotics and Control Systems
Transportation
Tensors
Correlation
Training data
Prediction methods
Prediction algorithms
Spatial databases
Power grids
Frequency nadir
convolutional neural network
deep learning
dynamic frequency prediction
power system
- Language
Severe disturbance may make the frequency fall below allowable value and make power system unable to maintain a steady frequency. In this paper, a predictive method for frequency nadir is proposed based on convolutional neural network (CNN). The measured operation data before and immediately after the disturbance is used as the input of CNN, with the frequency nadir predictive value as the output. The CNN input tensor is constructed on a 2-D plane that is able to reflect spatial distribution characteristics of nodes operation data. The electrical distance is used to describe the spatial correlation of power system nodes, and the t-SNE dimensionality reduction algorithm is presented to map the high-dimensional distance information of nodes to the 2-D plane. The CNN with deep network structure is adopted and the network parameters are trained by power system training data. The case study results on modified IEEE 39-node system and an actual power grid in USA shows that the proposed method can predict the frequency nadir after the disturbance accurately.