With the integration of photovoltaic (PV) power into an electrical network, the complexity of the grid management is increasing because of intermittent and fluctuation nature of solar energy. Solar irradiance forecasting is essential to facilitate planning and managing electricity generation and distribution in smart grid cyber-physical system (CPS). The performance of existing short-term forecasting methods is far from satisfactory due to a lack of reliable and fast time-frequency model for continuous-time solar irradiance data. To address this problem, this paper proposes a new method, Elman Neural Network (ENN) driven Wavelet Transform (WT-ENN), for hourly solar irradiance forecasting. Firstly, the solar irradiance series was decomposed into a set of constitutive series using wavelet transform. Secondly, the new wavelet coefficients were predicted by ENNs in every sub-series with the best network structure and parameters. Thirdly, Wavelet reconstruction will predict next hour solar irradiance through the aggregation of outputs of the ensemble of ENNs. Finally, the forecasting performance was evaluated using two large real-world solar irradiance datasets. Experiment results show that the new WT–ENN model outperforms a large number of alternative methods and an average forecast skill of 0.7590 over the persistence model. Thus, it is concluded that the proposed approach can significantly improve the forecasting accuracy and reliability.

Emotion recognition research has attracted great interest in various research fields, and electroencephalography (EEG) is considered a promising tool for extracting emotion-related information. However, traditional EEG-based emotion recognition methods ignore the spatial correlation between electrodes. To address this problem, this paper proposes an EEG-based emotion recognition method combining differential entropy feature matrix (DEFM) and 2D-CNN-LSTM. In this work, first, the one-dimensional EEG vector sequence is converted into a two-dimensional grid matrix sequence, which corresponds to the distribution of brain regions of the EEG electrode positions, and can better characterize the spatial correlation between the EEG signals of multiple adjacent electrodes. Then, the EEG signal is divided into equal time windows, and the differential entropy (DE) of each electrode in this time window is calculated, it is combined with a two-dimensional grid matrix and differential entropy to obtain a new data representation that can capture the spatiotemporal correlation of the EEG signal, which is called DEFM. Secondly, we use 2D-CNN-LSTM to accurately identify the emotional categories contained in the EEG signals and finally classify them through the fully connected layer. Experiments are conducted on the widely used DEAP dataset. Experimental results show that the method achieves an average classification accuracy of 91.92% and 92.31% for valence and arousal, respectively. The method performs outstandingly in emotion recognition. This method effectively combines the temporal and spatial correlation of EEG signals, improves the accuracy and robustness of EEG emotion recognition, and has broad application prospects in the field of emotion classification and recognition based on EEG signals.

TiO₂ was prepared through a sol–gel method as the thermo-catalyst for oxidative desulfurization (ODS), and H₂O₂ was selected as the oxidant.

In the solar power industry, irradiance forecasts are needed for planning, scheduling, and managing of photovoltaic power plants and grid-combined generating systems. A widely used method is artificial intelligence (AI), in particular, artificial neural networks, which can be trained over both historical values of irradiance and meteorological variables such as temperature, humidity, wind speed, pressure, and precipitation. In this paper, a novel version of the gated recurrent unit (GRU) method is combined with weather forecasts in order to predict solar irradiance. This method is used to forecast irradiance over a horizon of 24 h. Experiments show that the proposed method is able to outperform other AI methods. In particular, GRU using weather forecast data reduces the root mean squared error by 23.3% relative to a backpropagation neural network and 11.9% relative to a recurrent neural network. Compared to long short-term memory, the training time is reduced by 36.6%. Compared to persistence, the improvement in the forecast skill of the GRU is 42.0%. In summary, GRU is a promising technology which can be used effectively in irradiance forecasting.

The intermittence and fluctuation character of solar irradiance places severe limitations on most of its applications. The precise forecast of solar irradiance is the critical factor in predicting the output power of a photovoltaic power generation system. In the present study, Model I-A and Model II-B based on traditional long short-term memory (LSTM) are discussed, and the effects of different parameters are investigated; meanwhile, Model II-AC, Model II-AD, Model II-BC, and Model II-BD based on a novel LSTM-MLP structure with two-branch input are proposed for hour-ahead solar irradiance prediction. Different lagging time parameters and different main input and auxiliary input parameters have been discussed and analyzed. The proposed method is verified on real data over 5 years. The experimental results demonstrate that Model II-BD shows the best performance because it considers the weather information of the next moment, the root mean square error (RMSE) is 62.1618 W/m², the normalized root mean square error (nRMSE) is 32.2702%, and the forecast skill (FS) is 0.4477. The proposed algorithm is 19.19% more accurate than the backpropagation neural network (BPNN) in terms of RMSE.