Search

Obtaining accurate and timely crop area information is crucial for crop yield estimates and food security. Because most existing crop mapping models based on remote sensing data have poor generalizability, they cannot be rapidly deployed for crop identification tasks in different regions. Based on a priori knowledge of phenology, we designed an off-center Bayesian deep learning remote sensing crop classification method that can highlight phenological features, combined with an attention mechanism and residual connectivity. In this paper, we first optimize the input image and input features based on a phenology analysis. Then, a convolutional neural network (CNN), recurrent neural network (RNN), and random forest classifier (RFC) were built based on farm data in northeastern Inner Mongolia and applied to perform comparisons with the method proposed here. Then, classification tests were performed on soybean, maize, and rice from four measurement areas in northeastern China to verify the accuracy of the above methods. To further explore the reliability of the method proposed in this paper, an uncertainty analysis was conducted by Bayesian deep learning to analyze the model's learning process and model structure for interpretability. Finally, statistical data collected in Suibin County, Heilongjiang Province, over many years, and Shandong Province in 2020 were used as reference data to verify the applicability of the methods. The experimental results show that the classification accuracy of the three crops reached 90.73% overall and the average F1 and IOU were 89.57% and 81.48%, respectively. Furthermore, the proposed method can be directly applied to crop area estimations in different years in other regions based on its good correlation with official statistics. [ABSTRACT FROM AUTHOR]

Short-term load forecasting plays a significant role in the operation of power systems. Recently, deep learning has been generally employed in short-term load forecasting, primarily in the extraction of the characteristics of digital information in a single dimension without taking into account of the impact of external variables, particularly non-digital elements on load characteristics. In this paper, we propose a joint MFF-SAM-GCN to realize short-term load forecasting. First, we utilize a Bi-directional Long Short-Term Memory (Bi-LSTM) network and One-Dimensional Convolutional Neural Network (1D-CNN) in parallel connection to form a multi-feature fusion (MFF) framework, which can extract spatiotemporal correlation features of the load data. In addition, we introduce a Self-Attention Mechanism (SAM) to further enhance the feature extraction capability of the 1D-CNN network. Then with the deployment of a Graph Convolutional Network (GCN), the external non-digital features such as weather, strength, and direction of wind, etc., are extracted. Moreover, the generated weight matrices are incorporated into the load features to enhance feature recognition ability. Finally, we exploit Bayesian Optimization (BO) to find the optimal hyperparameters of the model to further improve the prediction accuracy. The simulation is taken from our proposed model and six benchmark schemes by using the bus load dataset of the Shandong Open Data Network, China. The results show that the RMSE of our proposed MFF-SAM-GCN model is 0.0284, while the SMAPE is 9.453%,the MBE is 0.025, and R-squared is 0.989, which is better than the selected three traditional machine learning methods and the three deep learning models. [ABSTRACT FROM AUTHOR]

• The vulnerable branches which affect the transient voltage security are identified. • Image-form samples combine grid structure and power flow information visually. • Knowledge visualization makes CNN-based transient voltage stability assessment model more understandable and reliable. • A comprehensive vulnerability indicator integrates the knowledge in CNN, grid topology and branch state. • The proposed method is demonstrated on a real-life provincial grid with renewable energy integration. Because of the increasing proportion of grid-connected converters, the transient voltage problem has increased the risk of the secure and stable operation of power systems. As the extension of transient voltage stability assessment (TVSA), transient voltage vulnerability assessment is the key technique to detect the crucial elements which can cause the insecurity risk of power system transient voltage. Although deep learning-based TVSA models have demonstrated the high accuracy, it is more important to display the knowledge learnt by the models and utilize it to recognize the weakness of transient voltage. This paper presents a transient voltage vulnerability assessment method by knowledge visualization of convolutional neural network (CNN). First, the image-form sample construction method is proposed to characterize the power system operating condition, which prepares for training CNN-based TVSA model. Then, the knowledge in TVSA model is visualized based on Gradient-weighted Class Activation Mapping (Grad-CAM) and Activation Maximization (AM) algorithms, which is analyzed from the perspective of power system stability. Finally, a comprehensive vulnerability indicator is proposed for identify the weakness of transient voltage, which combines the knowledge inside the TVSA model and the indicators with clear physical meaning. Simulation results of the New England 39-bus test system and the real Shandong provincial power grid in China demonstrate that the visualization makes the model more reliable and the proposed method can accurately identify the vulnerable branches. [ABSTRACT FROM AUTHOR]

• A long-term water quality prediction model with higher accuracy. • A graph convolutional network obtains topological information among sequences. • A subsequence cross-fusion network captures local features of a sequence. • A destationarizing attention weakens the distribution shift problem of sequences. Water quality prediction is important for maintaining water stability and control in marine environments. However, water quality parameters are affected by complex environmental factors.Marine time series data exhibit distribution shift and nonstationarity problems, owing to environmental heterogeneity.It is still difficult to obtain the spatial and time dependence of time series data with existing models and the model prediction accuracy cannot be guaranteed. Therefore, this paper proposes a graph-based convolutional neural network model, allowing networks and sequences to interact smoothly. The Time-graph convolutional fusion network(T-GCFN) consists of a graph convolutional fusion network(GCFN) and Time-pyramidal fusion attention(T-PFA). To obtain the spatial dependence of the original sequence, the original sequence is first decomposed into three subsequences by STL, and then the GCFN obtains the original sequence and trinomial subsequence through a multilayer T-GCN.Internal topology information is also obtained. Second, to eliminate the distribution differences between the training and test sets, the T-PFA is weakened using RevIN.To address the distribution shift problem of the sequence and the nonstationarity problem of the destationarized attention attenuation sequence, a higher weight is assigned to the sequence. Finally, depending on the long and short time series of the obtained sequence, the subsequence is downsampled by cross-learning, and the local features are extracted. Experiments on dissolved oxygen, salinity and temperature at six marine ranches on the Shandong peninsula were carried out using the T-GCFN, and the results were compared with those of other deep learning models. The experimental results show that the T-GCFN has better prediction performance and can achieve high-precision predictions of ocean chemistry parameters in the next three days. [ABSTRACT FROM AUTHOR]

The innovation of germplasm resources and the continuous breeding of new varieties of apples (Malus domestica Borkh.) have yielded more than 8000 apple cultivars. The ability to identify apple cultivars with ease and accuracy can solve problems in apple breeding related to property rights protection to promote the healthy development of the global apple industry. However, the existing methods are inconsistent and time-consuming. This paper proposes an efficient and convenient method for the classification of apple cultivars using a deep convolutional neural network with leaf image input, which is the delicate symmetry of a human brain learning. The model was constructed using the TensorFlow framework and trained on a dataset of 12,435 leaf images for the identification of 14 apple cultivars. The proposed method achieved an overall accuracy of 0.9711 and could successfully avoid the over-fitting problem. Tests on an unknown independent testing set resulted in a mean accuracy, mean error, and variance of μ a c c = 0.9685 , μ ε = 0.0315 , and σ 2 = 1.89025 E − 4 , respectively, indicating that the generalization accuracy and stability of the model were very good. Finally, the classification performance for each cultivar was tested. The results show that model had an accuracy of 1.0000 for Ace, Hongrouyouxi, Jazz, and Honey Crisp cultivars, and only one leaf was incorrectly identified for 2001, Ada Red, Jonagold, and Gold Spur cultivars, with accuracies of 0.9787, 0.9800, 0.9773, and 0.9737, respectively. Jingning1 and Pinova cultivars were classified with the lowest accuracies, with 0.8780 and 0.8864, respectively. The results also show that the genetic relationship between cultivars Shoufu 3 and Yanfu 3 is very high, which is mainly because they were both selected from a red mutation of Fuji and bred in Yantai City, Shandong Province, China. Generally, this study indicates that the proposed deep learning model is a novel and improved solution for apple cultivar identification, with high generalization accuracy, stable convergence, and high specificity. [ABSTRACT FROM AUTHOR]