Search

Abstract Agricultural Energy Internet (AEI), representing a key evolutionary direction in the integrated energy landscape of rural regions, holds a vital position in advancing the electrification of agricultural sectors. However, the disjointed control between agricultural loads and grid operations hinders the collaborative development of agriculture and energy. Addressing these issues, this paper investigates the current applications of artificial intelligence in the fields of agriculture and energy. The authors examine the evolutionary path of AEI, particularly emphasizing the critical technologies emerging from the intersection of agriculture, energy, and digital networks. Furthermore, the authors examine the critical technologies of reinforcement learning in the context of smart grid applications. In response to the challenges posed by low energy efficiency in rural areas, a reinforcement learning framework is proposed for coordinating fisheries, agriculture, livestock farming, and rural distribution networks. This framework provides a clear pathway for the application of reinforcement learning in AEI. This research acts as a conduit, merging agricultural and energy domains to promote a cohesive progression that markedly aids in the enhancement of rural electrification and the adoption of sustainable energy methodologies through reinforcement learning.

Abstract The weather has a significant impact on power load and power system planning. Stochastic weather simulation is important in the field of power systems. However, due to factors such as long recording years, observation technology, and so on, the historical weather data often have the problem of missing or insufficient. Meteorological data are characterized by changeable, rapid change, and high dimensions. Therefore, it is a challenging task to accurately grasp the law of weather data. This article presents a random weather simulation model based on gate recurrent unit (GRU) and generative adversarial networks (GAN). GRU selectively learns or forgets what was in the previous moment during training; it can learn the previous and current data of the time series data. When combined with the GAN, it will produce data with the same distribution as the original weather data. The proposed method was evaluated on a real weather dataset, and the results show that the proposed method outperforms the other contrast algorithms.

Abstract Oracle bone script, one of the earliest known forms of ancient Chinese writing, presents invaluable research materials for scholars studying the humanities and geography of the Shang Dynasty, dating back 3,000 years. The immense historical and cultural significance of these writings cannot be overstated. However, the passage of time has obscured much of their meaning, presenting a significant challenge in deciphering these ancient texts. With the advent of Artificial Intelligence (AI), employing AI to assist in deciphering Oracle Bone Characters (OBCs) has become a feasible option. Yet, progress in this area has been hindered by a lack of high-quality datasets. To address this issue, this paper details the creation of the HUST-OBC dataset. This dataset encompasses 77,064 images of 1,588 individual deciphered characters and 62,989 images of 9,411 undeciphered characters, with a total of 140,053 images, compiled from diverse sources. The hope is that this dataset could inspire and assist future research in deciphering those unknown OBCs. All the codes and datasets are available at https://github.com/Pengjie-W/HUST-OBC .

Abstract Solid‐state Li metal battery has attracted increasing interests for its potentially high energy density and excellent safety assurance, which is a promising candidate for next generation battery system. However, the low ionic conductivity and Li+ transport number of solid‐state polymer electrolytes limit their practical application. Herein, a composite polymer electrolyte with self‐inserted structure is proposed using the layered double hydroxides (LDHs) as dopant to achieve a fast Li+ transport channel in poly(vinylidene‐co‐trifluoroethylene) [P(VDF‐TrFE)] based polymer electrolyte. In such a composite electrolyte, P(VDF‐TrFE) polymer has an all‐trans conformation, in which all fluorine atoms locate on one side of the polymer chain, providing fast Li+ transport highways. Meanwhile, the LDH can immobilize the anions of Li salts based on the electrostatic interactions, promoting the dissociation of Li salts, thereby enhancing the ionic conductivity (6.4 × 10−4 S cm−1) and Li+ transference number (0.76). The anion immobilization effect can realize uniform electric field distribution at the anode surface and suppress the dendritic Li growth. Moreover, the hydrogen bonding interaction between LDH and polymer chains also endows the composite electrolyte with strong mechanical properties. Thus, at room temperature, the Li || Li symmetric cells can be stably cycled over 1000 h at a current density of 0.2 mA cm−2, and the full cells with LiFePO4 cathode deliver a high capacity retention (>95%) after 200 cycles. This work offers a promising route to construct solid‐state polymer electrolytes with fast Li+ transport.

Abstract In the context of modern agricultural production mode and domestic energy consumption, profound changes have taken place in agricultural and rural energy consumption, resulting in the demand for new technology development in various sectors of source, network, and load in rural energy systems. Agricultural energy internet (AEI) has promoted the development of renewable energy and agricultural electrification in villages. The construction of the AEI is crucial for achieving the synergistic development of agriculture, energy, and environment. The basic theory and key technologies of AEI are investigated and the prospects for the direction of agricultural energy technology are conducted. The research investigation shows that the AEI framework proposed by China Agricultural University is of great significance for realizing agricultural electrification and reducing agricultural carbon emissions. The article describes in detail the key data and characteristics of agricultural electrification loads. Digital twin technology and virtual power plant technology will be crucial to the construction of advanced AEI.

The global energy demand is increasing due to climate changes and carbon usages. Accumulating evidences showed energy sources using offshore wind from the sea can be added to increase our consumption capacity in long term. In addition, building offshore wind farms can also be environmentally advantageous compared to onshore farms. The assessment of wind energy resources is crucial for the site selection of wind farms. Currently, short-term wind forecast models have been developed to predict the wind power generation. However, methods are needed to improve the forecasting accuracy for ever-changing weather data. So, we try to use deep learning methods to predict long-term wind energy for identifying potential offshore wind farms. The experimental results indicate that PredRNN++ prediction model designed from the spatiotemporal perspective is feasible to evaluate long-term wind energy resources and has better performance than traditional LSTM.

In order to solve the high latency of traditional cloud computing and the processing capacity limitation of Internet of Things (IoT) users, Multi-access Edge Computing (MEC) migrates computing and storage capabilities from the remote data center to the edge of network, providing users with computation services quickly and directly. In this paper, we investigate the impact of the randomness caused by the movement of the IoT user on decision-making for offloading, where the connection between the IoT user and the MEC servers is uncertain. This uncertainty would be the main obstacle to assign the task accurately. Consequently, if the assigned task cannot match well with the real connection time, a migration (connection time is not enough to process) would be caused. In order to address the impact of this uncertainty, we formulate the offloading decision as an optimization problem considering the transmission, computation and migration. With the help of Stochastic Programming(SP), we use the posteriori recourse to compensate for inaccurate predictions. Meanwhile, in heterogeneous networks, considering multiple candidate MEC servers could be selected simultaneously due to overlapping, we also introduce the Multi-Arm Bandit (MAB) theory for MEC selection. The extensive simulations validate the improvement and effectiveness of the proposed SP-based Multi-arm bandit Method (SMM) for offloading in terms of reward, cost, energy consumption and delay. The results show that SMM can achieve about 20% improvement compared with the traditional offloading method that does not consider the randomness, and it also outperforms the existing SP/MAB based method for offloading.

Contingency analysis (CA) is fundamental to power system planning and operation. The compensation method and linear power flow (LPF) models are frequently used in CA to boost efficiency. However, the conventional compensation scheme fails to adapt to modern LPF models with novel structures. To bridge this gap, this paper proposes a generalized warm-start compensation algorithm (GWSCA) for LPFs. To efficiently calculate the inversion of a perturbed coefficient matrix in LPF models, a generalized decomposition method is devised based on the matrix inverse lemma and full rank decomposition. A warm-start strategy is developed to improve the computational accuracy by using the precontingency operating point of a power grid. The proposed GWSCA is incorporated with an LPF model based on logarithmic transformation of voltage magnitudes to implement fast and accurate CA. Simulation results show that GWSCA outperforms the conventional algorithms in terms of computational efficiency and accuracy.

Summary: Image-based AI has thrived as a potentially revolutionary tool for predicting molecular biomarker statuses, which aids in categorizing patients for appropriate medical treatments. However, many methods using hematoxylin and eosin-stained (H&E) whole-slide images (WSIs) have been found to be inefficient because of the presence of numerous uninformative or irrelevant image patches. In this study, we introduced the region of biomarker relevance (ROB) concept to identify the morphological areas most closely associated with biomarkers for accurate status prediction. We actualized this concept within a framework called saliency ROB search (SRS) to enable efficient and effective predictions. By evaluating various lung adenocarcinoma (LUAD) biomarkers, we showcased the superior performance of SRS compared to current state-of-the-art AI approaches. These findings suggest that AI tools, built on the ROB concept, can achieve enhanced molecular biomarker prediction accuracy from pathological images.

As a strong interference source for the all-time optical imaging surveillance of maritime targets, sea surface glare is difficult to mitigate accurately because of its time-varying characteristics due to lighting conditions and seawater fluctuations. In this paper, we propose an adaptive suppression approach to sea surface glare, which establishes a geographic polarization suppression model based on real-time information regarding geographic positioning and the orientation information of the floating platform, and also combines dynamic polarization control and pixel normalization to achieve adaptive suppression of sea surface glare. Experimental results show that this approach can mitigate the influence of rapidly changing glare effectively, and the SSIM indexes between the images without glare and those with glare suppression of the same scenes exceed 0.8, which is suitable for all-time glare suppression on the sea surface under natural lighting conditions.

Investigating the response of alien invasive plants to water level gradient in wetlands is important for developing strategies to prevent invasions by alien plants in these ecosystems. Controlled experiments were conducted to investigate the response of root morphology in the alien invasive plant Amaranthus retroflexus to water level gradient. The plants were planted in three water levels from low to high that resulted in drought, wet and flooded conditions. The results showed that biomass and allocation of A. retroflexus under flooded conditions were significantly lower than the other two treatments (p

In the spring of 2022, a new round of epidemic broke out in Shanghai, causing a shock to the Shanghai carbon trading market. Against this background, this paper studied the impact of the new epidemic on the price of Shanghai carbon emission rights and tried to explore the prediction model under the unexpected event. First, because a model based on point value data cannot capture the information hidden in inter-day price fluctuation, based on the interval price of Shanghai carbon emission rights (SHEA) and its influencing factors, an autoregressive conditional interval model with jumping and exogenous variables (ACIXJ) was established to explore the influence of the Russian–Ukrainian conflict and COVID-19 on the interval price of SHEA, respectively. The empirical results show that the conflict between Russia and Ukraine has no obvious influence on the price of SHEA, but COVID-19 led to a decline in the price trend of SHEA over four days before the city was closed, and the volatility changed significantly on the day before the city was closed. The price fluctuation was the strongest within 3 days after the city was closed; In addition, in order to accurately predict the interval data of SHEA against the background of COVID-19, based on the interval data decomposition algorithm (BEMD), a hybrid forecasting model of NDGM-ACIXJ/CNN-LSTM was constructed, in which the discrete gray model of approximate nonhomogeneous exponential series (NDGM) combined with the ACIXJ model is used to predict the high-frequency sub-interval, and the convolution neural network long-term and short-term memory model (CNN-LSTM) is used to predict the low-frequency sub-interval. The empirical results show that the prediction model proposed in this article has higher prediction precision than the reference models (ACIX, ACIXJ, NDGM-ACIXJ, BEMD-ACIX/CNN-LSTM, BEMD-ACIXJ/CNN-LSTM).

An electronic nose (e‐nose) mimics the mammalian olfactory system in identifying odors and expands human olfaction boundaries by tracing toxins and explosives. However, existing feature‐based odor recognition algorithms rely on domain‐specific expertise, which may limit the performance due to information loss during the feature extraction process. Inspired by human olfaction, a smart electronic nose enabled by an all‐feature olfactory algorithm (AFOA) is proposed, whereby all features in a gas sensing cycle of semiconductor gas sensors, including the response, equilibrium, and recovery processes are utilized. Specifically, our method combines 1D convolutional and recurrent neural networks with channel and temporal attention modules to fully utilize complementary global and dynamic information. It is further demonstrated that a novel data augmentation method can transform the raw data into a suitable representation for feature extraction. Results show that the e‐nose simply comprising of six semiconductor gas sensors achieves superior performances to state‐of‐the‐art methods on the Chinese liquor data. Ablation studies reveal the contribution of each sensor in odor recognition. Therefore, a deep‐learning‐enabled codesign of sensor arrays and recognition algorithms can reduce the heavy demand for a huge amount of highly specialized gas sensors and provide interpretable insights into odor recognition dynamics in an iterative way.

To promote the commercialization of yellow pea flour (YPF) due to its nutritional benefits. Four biscuits with different YPF ratio (10%–50%) were conducted to explore the optimal addition percentage. The effects of YPF on the rheological and baking performance of biscuits were performed. The results showed that the substitution ratio of YPF and milling methods had a critical impact on the rheological properties of dough. The dough stability decreased gradually while a softening degree increased with YPF ratio increased. In a term of biscuits, the dimensions of length (L), width (W), thickness (T) and color (L*) of biscuits reduced as YPF addition ratio increased, while colors (a* and b*) and hardness apparently increased. In addition, milling methods had a great influence on the texture and sensory evaluation of four biscuits. The dimensions and color parameters of biscuits from fine flours were larger than that from coarse flours, whereas hardness from fine flours was relatively softer, indicating flour with fine particle size could accelerate the extension and expansion of dough network, and improve Maillard reaction during baking. The highest sensory score for short and tough biscuits was obtained given at YPF ratio of 30% without compromising the qualities of biscuits. Keywords: Yellow pea flour, Rheological properties, Biscuits, Sensory evaluation

OBJECTIVE: This study aimed to identify the most useful ultrasound (US) features associated with definite neonatal necrotizing enterocolitis (NEC) and their prognostic values, particularly the calculated markers combined with important features. METHODS: A total of 213 suspected NEC cases were collected from the neonatal department of our hospital from January 2015 to August 2017. Each infant received both X-ray and US examinations. RESULTS: No differences were found in sex composition and delivery modes between groups. NEC-positive neonates had poorer prognosis compared to negative ones. The NEC group showed a higher frequency of abnormal signals. US showed higher NEC-related frequencies in different parameters. A variable (named predictor in US [PUS]) with five features was constructed. For NEC diagnosis, this variable provided a much higher area under the curve Q2 (AUC) (0.965) than other parameters. In this model, PUS had a cutoff value of 0.376 with a 0.900 sensitivity and 0.922 specificity. In prognosis, the closest factors were selected to draw a receiver operating characteristic curve, as well as a novel calculated variable US prognostic (USPro) marker. USPro had a much higher AUC (0.86) than other single features and showed a cutoff value of 0.18145, with 0.75 sensitivity and 0.84 specificity. This variable had a weaker power in prognosis when compared with PUS in diagnosis. CONCLUSIONS: The application of abdominal color Doppler US can provide high accuracy and sensitivity in NEC diagnosis and also contribute to its prognosis, without induction of radiation. Suspected neonates should be examined using this technique as early as possible.

Accurate transient stability assessment (TSA) is a fundamental requirement for ensuring secure and stable operation of power systems. Tremendous efforts have been made to apply artificial intelligence approaches for TSA with phasor measurement unit data. However, many previous approaches may be failed to provide favorable accuracy due to the shallow architectures and error-prone hand-crafting features. This paper proposed a model for TSA, which is termed multi-branch stacked denoising autoencoder (MSDAE). This model is a unified framework integrating multiple stacked denoising autoencoders (SDAEs), one fusion layer, and one logistic regression (LR) layer. Initially, the SDAEs at the bottom of MSDAE extract features from multiple kinds of measurements respectively. Then, the extracted features are encoded into unified fusion features by the fusion layer. Finally, the LR layer performs TSA by using the fusion features. The depth of the architecture contributes to the remarkable ability for feature learning, while the width of the architecture (i.e., the multiple branches) enables MSDAE to deal with different kinds of measurements by a reasonable mechanism. In this way, MSDAE achieves feature extraction and classification intrinsically and simultaneously, namely, achieves TSA in an end-to-end manner. The results of experiments on IEEE 50-machine system demonstrate the superiority of the proposed model over the prior methods.

Text in natural images contains rich semantics that is often highly relevant to objects or scene. In this paper, we focus on the problem of fully exploiting scene text for visual understanding. The main idea is combining word representations and deep visual features in a globally trainable deep convolutional neural network. First, the recognized words are obtained by a scene text reading system. Next, we combine the word embedding of the recognized words and the deep visual features into a single representation that is optimized by a convolutional neural network for fine-grained image classification. In our framework, the attention mechanism is adopted to compute the relevance between each recognized word and the given image, which further enhances the recognition performance. We have performed experiments on two datasets: con-text dataset and drink bottle dataset, which are proposed for fine-grained classification of business places and drink bottles, respectively. The experimental results consistently demonstrate that the proposed method of combining textual and visual cues significantly outperforms classification with only visual representation. Moreover, we have shown that the learned representation improves the retrieval performance on the drink bottle images by a large margin, making it potentially powerful in product search.

It is meaningful to investigate the impact of alien invasive plants on wetland sediment. Field experiments were conducted to study the impact of the alien invasive plant, Amaranthus retroflexus, on wetland sediment properties under two growth stages. The results showed that growth of A. retroflexus tended to increase moisture content, porosity, and could significantly decrease total nitrogen (TN) and nitrate (N-NO3ˉ) concentrations in sediment and ammonium (N-NH4+) concentration in interstitial water (p < 0.05). However, no obvious impact was observed on total phosphorus (TP) nor on N-NH4+ concentrations in sediment. There was a difference in impact of A. retroflexus on sediment properties under two growth stages. The plants with a longer growth stage significantly decreased loss on ignition in sediment and P-PO43ˉ concentration in interstitial water. There were also significant differences in N-NH4+ concentration in both sediment and interstitial water over the vertical profile of the plant. In the plant treatment with a longer growth stage, P-PO43ˉ concentration also showed significant differences over the vertical profile. Therefore, the growth of A. retroflexus can improve sediment physical and chemical properties, reducing N-NH4+ and P-PO43ˉ release from sediment to overlying water to some extent, thereby decreasing sediment nutrient loading.