Research progress in water quality prediction based on deep learning technology: a review.

Water, an invaluable and non-renewable resource, plays an indispensable role in human survival and societal development. Accurate forecasting of water quality involves early identification of future pollutant concentrations and water quality indices, enabling evidence-based decision-making and targeted environmental interventions. The emergence of advanced computational technologies, particularly deep learning, has garnered considerable interest among researchers for applications in water quality prediction because of its robust data analytics capabilities. This article comprehensively reviews the deployment of deep learning methodologies in water quality forecasting, encompassing single-model and mixed-model approaches. Additionally, we delineate optimization strategies, data fusion techniques, and other factors influencing the efficacy of deep learning-based water quality prediction models, because understanding and mastering these factors are crucial for accurate water quality prediction. Although challenges such as data scarcity, long-term prediction accuracy, and limited deployments of large-scale models persist, future research aims to address these limitations by refining prediction algorithms, leveraging high-dimensional datasets, evaluating model performance, and broadening large-scale model application. These efforts contribute to precise water resource management and environmental conservation.

Quarterly electricity consumption prediction based on time series decomposition method and gray model.

Accurately predicting electricity consumption is crucial for reducing power waste and maintaining power system stability. To address the non-linear and seasonal fluctuations of electricity consumption, this paper proposes a seasonal prediction method based on Seasonal and Trend decomposition using Loess (STL) algorithm and gray model by introducing time series decomposition method. The STL decomposition algorithm decomposes fluctuating electricity data into three components: trend, seasonal, and remainder. Then reasonable methods are used to predict components with different data characteristics. The novel model is employed to analyze the quarterly electricity consumption in Zhejiang province of China from 2014Q4 to 2022Q3. The experimental results show that the prediction accuracy of this model is superior to the state-of-the art models; the MAPE and RMSPE values are 1.77% and 2.37%, respectively. Our model that can effectively identify seasonal fluctuations in data sequences provides a new method for predicting seasonal fluctuation data and optimizing seasonal electricity supply schemes.

Carbon price prediction based on multiple decomposition and XGBoost algorithm.

Carbon trading is an effective way to limit global carbon dioxide emissions. The carbon pricing mechanisms play an essential role in the decision of the market participants and policymakers. This study proposes a carbon price prediction model, multi-decomposition-XGBOOST, which is based on sample entropy and a new multiple decomposition algorithm. The main steps of the proposed model are as follows: (1) decompose the price series into multiple intrinsic mode functions (IMFs) by using complete ensemble empirical mode decomposition with adaptive noise (CEEMDAN); (2) decompose the IMF with the highest sample entropy by variational mode decomposition (VMD); (3) select and recombine some IMFs based on their sample entropy, and then perform another round of decomposition via CEEMDAN; (4) predict IMFs by XGBoost model and sum up the prediction results. The model has exhibited reliable predictive performance in both the highly fluctuating Beijing carbon market and the comparatively stable Hubei carbon market. The proposed model in Beijing carbon market achieves improvements of 30.437%, 44.543%, and 42.895% in RMSE, MAE, and MAPE, when compared to the single XGBoost models. Similarly, in Hubei carbon market, the RMSE, MAE, and MAPE based on multi-decomposition-XGBOOST model decreased by 28.504%, 39.356%, and 39.394%. The findings indicate that the proposed model has better predictive performance for both volatile and stable carbon prices.

A PCA-EEMD-CNN-Attention-GRU-Encoder-Decoder Accurate Prediction Model for Key Parameters of Seawater Quality in Zhanjiang Bay.

In order to effectively solve the problem of low accuracy of seawater water quality prediction, an optimized water quality parameter prediction model is constructed in this paper. The model first screened the key factors of water quality data with the principal component analysis (PCA) algorithm, then realized the de-noising of the key factors of water quality data with an ensemble empirical mode decomposition (EEMD) algorithm, and the data were input into the two-dimensional convolutional neural network (2D-CNN) module to extract features, which were used for training and learning by attention, gated recurrent unit, and an encoder-decoder (attention-GRU-encoder-decoder, attention-GED) integrated module. The trained prediction model was used to predict the content of key parameters of water quality. In this paper, the water quality data of six typical online monitoring stations from 2017 to 2021 were used to verify the proposed model. The experimental results show that, based on short-term series prediction, the root mean square error (RMSE), mean absolute percentage error (MAPE), and decision coefficient (R2) were 0.246, 0.307, and 97.80%, respectively. Based on the long-term series prediction, RMSE, MAPE, and R2 were 0.878, 0.594, and 92.23%, respectively, which were all better than the prediction model based on an enhanced clustering algorithm and adam with a radial basis function neural network (ECA-Adam-RBFNN), a prediction model based on a softplus extreme learning machine method with partial least squares and particle swarm optimization (PSO-SELM-PLS), and a wavelet transform-depth Bi-S-SRU (Bi-directional Stacked Simple Recurrent Unit) prediction model. The PCA-EEMD-CNN-attention-GED prediction model not only has high prediction accuracy but can also provide a decision-making basis for the water quality control and management of aquaculture in the waters around Zhanjiang Bay.

Bioinformatics Analysis of the Periodicity in Proteins with Coiled-Coil Structure-Enumerating All Decompositions of Sequence Periods.

A coiled coil is a structural motif in proteins that consists of at least two α-helices wound around each other. For structural stabilization, these α-helices form interhelical contacts via their amino acid side chains. However, there are restrictions as to the distances along the amino acid sequence at which those contacts occur. As the spatial period of the α-helix is 3.6, the most frequent distances between hydrophobic contacts are 3, 4, and 7. Up to now, the multitude of possible decompositions of α-helices participating in coiled coils at these distances has not been explored systematically. Here, we present an algorithm that computes all non-redundant decompositions of sequence periods of hydrophobic amino acids into distances of 3, 4, and 7. Further, we examine which decompositions can be found in nature by analyzing the available data and taking a closer look at correlations between the properties of the coiled coil and its decomposition. We find that the availability of decompositions allowing for coiled-coil formation without putting too much strain on the α-helix geometry follows an oscillatory pattern in respect of period length. Our algorithm supplies the basis for exploring the possible decompositions of coiled coils of any period length.

Lamb Wave-Based Damage Localization and Quantification in Composites Using Probabilistic Imaging Algorithm and Statistical Method.

Quantitatively and accurately monitoring the damage to composites is critical for estimating the remaining life of structures and determining whether maintenance is essential. This paper proposed an active sensing method for damage localization and quantification in composite plates. The probabilistic imaging algorithm and the statistical method were introduced to reduce the impact of composite anisotropy on the accuracy of damage detection. The matching pursuit decomposition (MPD) algorithm was utilized to extract the precise TOF for damage detection. The damage localization was realized by comprehensively evaluating the damage probability evaluation results of all sensing paths in the monitoring area. Meanwhile, the scattering source was recognized on the elliptical trajectory obtained through the TOF of each sensing path to estimate the damage size. Damage size was characterized by the Gaussian kernel probability density distribution of scattering sources. The algorithm was validated by through-thickness hole damages of various locations and sizes in composite plates. The experimental results demonstrated that the localization and quantification absolute error are within 11 mm and 2.2 mm, respectively, with a sensor spacing of 100 mm. The algorithm proposed in this paper can accurately locate and quantify damage in composite plate-like structures.

Analysis of active compounds and geographical origin discrimination of Atractylodes macrocephala Koidz. by using high performance liquid chromatography-diode array detection fingerprints combined with chemometrics.

In this work, a simple and effective strategy for the determination of 12 active compounds of Atractylodes macrocephala Koidz. (AM) was proposed by using high performance liquid chromatography-diode array detection (HPLC-DAD) combined with alternating trilinear decomposition (ATLD) algorithm. Utilizing the "second-order advantage", three common problems in HPLC could be resolved, namely baseline drifts, peak overlaps, and unknown interferences. 12 compounds were rapidly eluted within 12.5 min, and the average spiked recoveries were 80.8-109.9%. The figures of merit reflected the feasibility of the proposed method. Compared with the results of the traditional univariate calibration method based on HPLC-UV technique, the proposed strategy further verified the reliability and simplicity of the mathematical separation. On this basis, partial least squares-discriminant analysis (PLS-DA) was applied to discriminate 113 AM samples from different geographical origins, and variable importance in projection (VIP) was used to further screen the main differential components that affect the regional division of AM. A series of results show that the AM samples from the three regions have obviously different clustering trends. Overall, the strategy is expected to provide a scientific basis for the modern research of medicinal materials, and it is also conducive to the clinical use and market supervision of AM.

A novel short-term carbon emission prediction model based on secondary decomposition method and long short-term memory network.

Grasping the dynamics of carbon emission in time plays a key role in formulating carbon emission reduction policies. In order to provide more accurate carbon emission prediction results for planners, a novel short-term carbon emission prediction model is proposed. In this paper, the secondary decomposition technology combining ensemble empirical mode decomposition (EEMD) and variational mode decomposition (VMD) is used to process the original data, and the partial autocorrelation function (PACF) is applied to select the optimal model input. Then, the long short-term memory network (LSTM) is chosen for prediction. The secondary decomposition algorithm is innovatively introduced into the field of carbon emission prediction, and the empirical results illustrate that the secondary decomposition technology can further improve the prediction accuracy. Combined with the secondary decomposition, the R2, MAPE, and RMSE of the model are improved by 2.20%, 43.08%, and 36.92% on average. And the proposed model shows excellent prediction accuracy (R2 = 0.9983, MAPE = 0.0031, RMSE = 118.1610) compared with other 12 comparison models. Therefore, this model not only has potential value in the formulation of carbon emission reduction plans, but also provides a valuable reference for future carbon emission forecasting research.

A novel methodology concentrating on risk propagation to conduct a risk analysis based on a directed complex network.

A novel methodology is proposed in the present study to describe the risk propagation process by quantitatively evaluating the criticality and sensitivity of risk events according to complex network theory, based on which risk matrices are developed to interrupt the risk propagation process by setting up safety barriers. The applicability and accuracy of the improved k-shell decomposition algorithm and risk flow model for calculating the criticality proposed in this study are verified by the susceptible-infected-recovered (SIR) simulation, which is widely regarded as a benchmark for complex networks (CN) issues. The results confirm the advantages of the proposed methodologies considering comprehensively various comparison indicators. The sensitivity of the nodes is quantified by running an SIR simulation with a variable infection rate and recovery rate. Finally, the criticality and sensitivity of risk events contribute to the development of risk matrices with three different risk scenarios, based on which the applicability and effectiveness of safety barriers are qualitatively analyzed to interrupt the risk propagation process. The framework and methodologies proposed in this study could well present the risk propagation process within CNs and are proven to have a great potential for studies on safety barriers.

Modeling and optimizing an agro-supply chain considering different quality grades and storage systems for fresh products: a Benders decomposition solution approach.

This paper proposes a mathematical model in the context of agro-supply chain management, considering specific characteristics of agro-products to assist purchase, storage, and transportation decisions. In addition, a new method for determining the required quality score of different types of products is proposed based on their loss factors and purchasing costs. The model aims to minimize total cost imposed by purchasing fresh products, opening warehouses, holding inventories, operational activities, and transportation. Two sets of examples, including small and medium-sized problems, are implemented by general algebraic modeling language (GAMS) software to evaluate the model. Then, Benders decomposition (BD) algorithm is applied to tackle the complexity of solving large-sized instances. The results of both GAMS and BD are compared in terms of objective function values and computational time to demonstrate the efficiency of the BD algorithm. Finally, the model is applied in a real case study involving an apple supply chain to obtain managerial insights.