研究目的
To develop a reliable protection scheme for PV integrated microgrid that can discriminate between PV array faults and symmetrical line faults using a sparse auto-encoder and deep neural network (SAE-DNN) approach.
研究成果
The proposed SAE-DNN based protection scheme effectively discriminates between PV array faults and symmetrical line faults, achieving high classification accuracy and reliability. It outperforms conventional techniques in terms of dependability and security under both islanding and grid-connected modes. The scheme is validated for practical applications through real-time simulations on OPAL-RT digital simulator.
研究不足
The study focuses on PV integrated microgrids and may not be directly applicable to other types of microgrids or renewable energy sources. The real-time validation is performed on a specific setup, which may not cover all possible field conditions.
1：Experimental Design and Method Selection:
The study employs a SAE-DNN based approach for feature learning and classification of faults in PV integrated microgrids. The methodology involves converting voltage and current signals into grayscale images for input to the SAE.
2：Sample Selection and Data Sources:
Voltage and current signals are retrieved from relaying buses in a 34.5 kV, 60 Hz balanced microgrid system under study.
3：5 kV, 60 Hz balanced microgrid system under study.
List of Experimental Equipment and Materials:
3. List of Experimental Equipment and Materials: The study uses OPAL-RT digital simulator for real-time validation of the proposed protection scheme.
4：Experimental Procedures and Operational Workflow:
The process includes signal conversion, unsupervised feature learning by SAE, stacking of SAEs to form a DNN, and classification using a softmax classifier.
5：Data Analysis Methods:
The performance of the proposed scheme is evaluated through reliability analysis and compared with ANN, SVM, and DT based techniques.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
