Li Xiong, Kai Kang, Yunlong Zhang, Wen Zhang, Yunlong Zhang, and Chuansheng Cao
Convolutional neural network, temporal convolutional network(TCN), Kalman filter, scenario generation, copula function
The global shift towards low-carbon actions has highlighted the importance of promoting new energy sources, especially the great potential for development of wind and solar power generation. Given the instability and high volatility of wind and solar power generation, this article proposes a temporal convolutional network (TCN) based on Kalman filter to predict the wind and photovoltaic output coefficients of Ordos Zero-Carbon Industrial Park. It utilises the TCN to extract useful features from the original time series data, and the extracted features can be used as the input of Kalman filter. The output of Kalman filter can be used as the feedback of the TCN to adjust and optimise the network parameters. The collaborative work of the two can achieve more accurate and reliable time series data analysis and prediction. The prediction results show that this method can effectively predict. In addition, this article also uses Copula function to generate power generation scenarios, and optimises them through Monte Carlo model. The refined output scenarios retain the correlation and provide a benchmark for power system reliability assessment.
Important Links:
Go Back
