An Ultrashort-Term Net Load Forecasting Model Based on Phase Space Reconstruction and Deep Neural Network

Mei, Fei; Wu, Qingliang; Shi, Tian; Lu, Jixiang; Pan, Yi; Zheng, Jianhua

doi:10.3390/app9071487

Cited by 27 publications

(11 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…During the heuristics, we gradually increase the number of hidden layers until the predicted value is obviously overfitting, and then gradually reduce the number of hidden layers to minimize the model error, thereby determining the optimal number of hidden layers. In this paper, through multiple experiments, the model performance is best when the number of hidden layers is 4 and the neurons of each hidden layer are 6, 4, 3, and 2, respectively [34]. The activation function of hidden neurons uses the tanh function.…”

Section: ( ) ( ) ( )mentioning

confidence: 99%

Ash-Covered State Evaluation and Alarm of Distributed Photovoltaic System Considering Full Lifetime Degradation

et al. 2020

Self Cite

View full text Add to dashboard Cite

With the widespread attention and research of distributed Photovoltaic (PV) systems, the state evaluation of distributed PV system, especially the evaluation of ash-covered state, has become increasingly prominent. To this end, an ash-covered state evaluation method for the distributed PV systems, which considers the full lifetime degradation, is proposed. First, the PV lifetime degradation rate is calculated based on the historical data of the PV system. Second, the fuzzy C-means (FCM) is used to cluster PV measured AC output power data into four weather types: sunny, cloudless, cloudy, and rainy. According to weather conditions, the Levenberg-Marquardt backpropagation (LMBP) Deep Neural Network (DNN) is adopted to establish four distributed PV AC output power fitting models by using measured meteorological data and measured output power data. Afterward, distributed PV output power fitted value is obtained by reducing the DNN output proportionally according to the PV degradation rate. Moreover, by analyzing the difference between the PV output power fitted value and the measured output power data, the ash-covered state of the system is evaluated, and a state alarm mechanism is established based on the system state. Finally, the validity and rationality of the proposed method are verified by the analysis of examples.

show abstract

Section: ( ) ( ) ( )mentioning

confidence: 99%

Ash-Covered State Evaluation and Alarm of Distributed Photovoltaic System Considering Full Lifetime Degradation

et al. 2020

Self Cite

View full text Add to dashboard Cite

show abstract

“…The results showed that the hybrid forecasting method offers better accuracy and stability than the single prediction methods. Additionally, Mei et al [15] developed an ultrashort-term forecasting model based on the phase space reconstruction and deep neural network (DNN) by considering the characteristics of the net load. The performance of this model was verified using real data, with superior accuracy in forecasting the net load under high PV penetration rates and different weather conditions.…”

Section: Performance Assessmentmentioning

confidence: 99%

Solar Power System Planning and Design

Choi

2020

Applied Sciences

View full text Add to dashboard Cite

show abstract

“…Where ε is the learning rate, which takes the value of 0.8 in this paper, and the superscript k represents the k-th iteration. In this paper, the LM (Levenberg-Marquardt) BP algorithm [31] is used to replace the traditional BP algorithm to fine-tune the DBN. Compared with the traditional BP algorithm, the LMBP algorithm has faster convergence speed and higher convergence reliability, and is more suitable for training neural networks with many hidden layers and neurons.…”

Section: Dbn Based On Levenberg-marquardt Backpropagation (Lmbp) Algomentioning

confidence: 99%

Phase Space Reconstruction Algorithm and Deep Learning-Based Very Short-Term Bus Load Forecasting

Shi

Mei

et al. 2019

Energies

Self Cite

View full text Add to dashboard Cite

With the refinement and intelligence of power system optimal dispatching, the widespread adoption of advanced grid applications that consider the safety and economy of power systems, and the massive access of distributed energy resources, the requirement for bus load prediction accuracy is continuously increasing. Aiming at the volatility brought about by the large-scale access of new energy sources, the adaptability to different forecasting horizons and the time series characteristics of the load, this paper proposes a phase space reconstruction (PSR) and deep belief network (DBN)-based very short-term bus load prediction model. Cross-validation is also employed to optimize the structure of the DBN. The proposed PSR-DBN very short-term bus load forecasting model is verified by applying the real measured load data of a substation. The results prove that, when compared to other alternative models, the PSR-DBN model has higher prediction accuracy and better adaptability for different forecasting horizons in the case of high distributed power penetration and large fluctuation of bus load.

show abstract

An Ultrashort-Term Net Load Forecasting Model Based on Phase Space Reconstruction and Deep Neural Network

Cited by 27 publications

References 25 publications

Ash-Covered State Evaluation and Alarm of Distributed Photovoltaic System Considering Full Lifetime Degradation

Ash-Covered State Evaluation and Alarm of Distributed Photovoltaic System Considering Full Lifetime Degradation

Solar Power System Planning and Design

Phase Space Reconstruction Algorithm and Deep Learning-Based Very Short-Term Bus Load Forecasting

Contact Info

Product

Resources

About