Rui Xu

Data-driven aeration optimization is an effective approach for reducing energy consumption in wastewater treatment plants (WWTPs). However, in information-limited scenarios, newly established or emerging-market WWTPs often lack historical labels for aeration actions, making it difficult to construct high-precision surrogate models. Conventional cross-plant model deployments face severe data distribution shifts, and standard multi-objective optimization algorithms are prone to generating non-physical extrapolation errors, such as achieving compliance with “zero aeration” under low-concentration conditions. To break through inter-plant data barriers, this study proposes an intelligent aeration decision-making framework that integrates cross-domain transfer learning with physics-informed constraints. First, this study designs an adversarial network incorporating a state-action decoupling bypass. By employing a gradient reversal layer (GRL) to extract domain-invariant representations while the decoupling bypass preserves the physical sensitivity of control commands, this network achieves robust cross-plant knowledge transfer. Second, this study proposes a physics-informed multi-objective particle swarm optimization (PI-MOPSO) algorithm, which embeds the theoretical oxygen demand as a physical penalty into the fitness function, ensuring the physical reliability of the optimization decisions. Experiments demonstrate that the surrogate model restricts the prediction errors for effluent chemical oxygen demand (COD) and effluent ammonium nitrogen removal rates to within 1%. Validated by statistical tests, the improved algorithm effectively circumvents non-physical prediction biases. Its Pareto front achieves a spacing metric of 0.0027, outperforming baseline algorithms in hypervolume stability. This framework provides reliable aeration scheduling references conforming to biochemical dynamics for target WWTPs lacking historical action labels, offering a promising theoretical foundation for future practical engineering applications.

Data-driven aeration optimization is an effective approach for reducing energy consumption in wastewater treatment plants (WWTPs). However, newly established or emerging-market WWTPs often lack historical aeration logs, making it difficult to construct high-precision surrogate models. Conventional cross-plant model deployments face severe data distribution shifts, and standard multi-objective optimization algorithms are prone to generating non-physical extrapolation errors, such as achieving compliance with "zero aeration" under low-concentration conditions. To break through inter-plant data barriers, this study proposes an intelligent aeration decision-making framework that integrates cross-domain transfer learning with physics-informed constraints. First, this study designs an adversarial network based on air-to-water ratio and removal rate features. By employing a gradient reversal layer (GRL) to extract domain-invariant representations, this network achieves cross-plant knowledge transfer. Second, this study proposes a physics-informed multi-objective particle swarm optimization (PI-MOPSO) algorithm, which embeds the theoretical oxygen demand as a physical penalty into the fitness function, ensuring the physical reliability of the optimization decisions. Experiments demonstrate that the surrogate model restricts the prediction errors for effluent chemical oxygen demand (COD) and total nitrogen (TN) removal rates to within 1%. Validated by statistical tests, the improved algorithm effectively circumvents non-physical prediction biases. Its Pareto front achieves a spacing metric of 0.0027, outperforming baseline algorithms in hypervolume stability. This framework provides optimal aeration scheduling strategies conforming to biochemical dynamics for target WWTPs lacking aeration action labels, demonstrating substantial practical engineering value.

Third-view industrial assembly videos exhibit long transitional periods, weak hand–tool cues under occlusion, and strict efficiency requirements, which jointly challenge existing video action recognition models. This study proposes an efficiency-oriented video transformer framework that explicitly targets temporal redundancy while preserving fine-grained interaction evidence. A motion-guided PreFocusSelector performs input-level frame selection and outputs an effective subsequence for backbone inference, thereby reducing redundant spatiotemporal token processing. To compensate for subtle local dynamics, an AxisSplit Adapter is introduced to enhance short-range spatiotemporal modeling with lightweight overhead. In addition, a temporal attention pooling module replaces uniform temporal averaging by learning discriminative temporal weights, improving video-level evidence integration under sparse signals. Experiments are conducted on a third-view assembly action dataset filtered to 1759 clips with a consistent train/validation/test split. Under a fixed evaluation protocol and 10 independent runs, the proposed method achieves 96.41% ± 0.22% Top-1 accuracy and 94.83% ± 0.26% Macro-F1, while reducing computational cost from 88.1 to 37.6 G FLOPs with modest parameter increase. A common frame subsampling baseline is included to contextualize efficiency comparisons, and comprehensive implementation details and ablation studies are provided to support reproducibility.

DO is an important index to characterize environmental water quality. The time series fluctuation of DO can be analyzed via frequency band decomposition, which is very valuable for water quality simulations. In this paper, DO in the Chengdu area of China was studied using variational mode decomposition with daily meteorological data and water quality data from 2020 to 2022. After variable decomposition, the DO data were first decomposed into different frequency band signals named IMF1, IMF2, IMF3, IMF4, and IMF5. IMF1 represented the low-frequency signal with long-term trend characteristics of the data. IMF2 to IMF5 represented the high-frequency signal with short-term mutation characteristics of the data. By combining the variable decomposition results with the correlation analysis, it was found that the long-term trend characteristics of DO are affected by the superposition of meteorological factors, hydrological factors, and water pollution factors but have a weak correlation with any single determining factor. The air temperature, water temperature, phosphorus, air pressure, pH value, chemical oxygen demand, and nitrogen were relatively strongly correlated with the long-term trend characteristics of DO. The short-term mutation characteristics of DO were mainly determined using the characteristics of the water body itself, while the influence of the meteorological factors could basically be ignored. The water temperature, pH value, and eutrophication were the biggest influencing factors. Then, a predictive framework combining frequency division with a deep learning model or a machine learning model was constructed to predict DO. The predicted results of GRU, random forest, and XGBoost with and without the framework were compared. It was shown that, after removing the interference factors with correlations less than 0.3, the predicted value of DO was much closer to the actual value. The XGBoost and random forest models with decomposed signals had a high degree of simulation fitting and could be used to predict DO in the Chengdu area. The above research approach can be applied to further explore the prediction of various pollution factors in different areas of China.

In environmental hydrodynamics, a research topic that has gained popularity is the transmission and diffusion of water pollutants. Various types of change processes in hydrological and water quality are directly related to meteorological changes. If these changing characteristics are classified effectively, this will be conducive to the application of deep learning theory in water pollution simulation. When periodically monitoring water quality, data were represented with a candlestick chart, and different classification features were displayed. The water quality data from the research area from 2012 to 2019 generated 24 classification results in line with the physics laws. Therefore, a deep learning water pollution prediction method was proposed to classify the changing process of pollution to improve the prediction accuracy of water quality, based on candlestick theory, visual geometry group, and gate recurrent unit (CT-VGG-GRU). In this method, after the periodic changes of water quality were represented by candlestick graphically, the features were extracted by the VGG network based on its advantages in graphic feature extraction. Then, this feature and other scenario parameters were fused as the input of the time series network model, and the pollutant concentration sequence at the predicted station constituted the output of the model. Finally, a hybrid model combining graphical and time series features was formed, and this model used continuous time series data from multiple stations on the Lijiang River watershed to train and validate the model. Experimental results indicated that, compared with other comparison models, such as the back propagation neural network (BPNN), support vector regression (SVR), GRU, and VGG-GRU, the proposed model had the highest prediction accuracy, especially for the prediction of extreme values. Additionally, the change trend of water pollution was closer to the real situation, which indicated that the process change information of water pollution could be fully extracted by the CT-VGG-GRU model based on candlestick theory. For the water quality indicators DO, COD_Mn, and NH₃-N, the mean absolute errors (MAE) were 0.284, 0.113, and 0.014, the root mean square errors (RMSE) were 0.315, 0.122, and 0.016, and the symmetric mean absolute percentage errors (SMAPE) were 0.022, 0.108, and 0.127, respectively. The established CT-VGG-GRU model achieved superior computational performance. Using the proposed model, the classification information of the river pollution process could be obtained effectively and the time series information could also be retained, which made the application of the deep learning model to the transmission and diffusion process of river water pollution more explanatory. The proposed model can provide a new method for water quality prediction.

Deep learning models have been widely used in time-series numerical prediction of atmospheric environmental quality. The fundamental feature of this application is to discover the correlation between influencing factors and target parameters through a deep network structure. These relationships in original data are affected by several different frequency factors. If the deep network is adopted without guidance, these correlations may be masked by entangled multifrequency data, which will cause the problem of insufficient correlation feature extraction and difficult model interpretation. Because the wavelet transform has the ability to separate these entangled multifrequency data, and these correlations can be extracted by deep learning methods, a hybrid model combining wavelet transform and transformer-like (WTformer) was designed to extract time–frequency domain features and prediction of air quality. The 2018–2021 hourly data in Guilin was used as the benchmark training dataset. Pollutants and meteorological variables in the local dataset are decomposed into five frequency bands by wavelet. The analysis of the WTformer model showed that particulate matter (PM_2.5 and PM₁₀) had an obvious correlation in the low-frequency band and a low correlation in the high-frequency band. PM_2.5 and temperature had a negative correlation in the high-frequency band and an obvious positive correlation in the low-frequency band. PM_2.5 and wind speed had a low correlation in the high-frequency band and an obvious negative correlation in the low-frequency band. These results showed that the laws of variables in the time–frequency domain could be found by the model, which made it possible to explain the model. The experimental results show that the prediction performance of the established model was better than that of multilayer perceptron (MLP), one-dimensional convolutional neural network (1D-CNN), gate recurrent unit (GRU), long short-term memory (LSTM) and Transformer, in all time steps (1, 4, 8, 24 and 48 h).

Currently, the continuous change prediction of PM_2.5 concentration is an air pollution research hotspot. Combining physical methods and deep learning models to divide the pollution process of PM_2.5 into effective multiple types is necessary to achieve a reliable prediction of the PM_2.5 value. Therefore, a candlestick chart sample generator was designed to generate the candlestick chart from the online PM_2.5 continuous monitoring data of the Guilin monitoring station site. After these generated candlestick charts were analyzed through the Gaussian diffusion model, it was found that the characteristics of the physical transmission process of PM_2.5 pollutants can be reflected. Based on a set three-day period, using the time linear convolution method, 2188 sets of candlestick chart data were obtained from the 2013–2018 PM_2.5 concentration data. There existed 16 categories generated by unsupervised classification that met the established classification judgment standards. After the statistical analysis, it was found that the accuracy rate of the change trend of these classifications reached 99.68% during the next period. Using the candlestick chart data as the training dataset, the Visual Geometry Group (VGG) model, an improved convolutional neural network model, was used for the classification. The experimental results showed that the overall accuracy (OA) value of the candlestick chart combination classification was 96.19%, and the Kappa coefficient was 0.960. IN the VGG model, the overall accuracy was improved by 1.93%, on average, compared with the support vector machines (SVM), LeNet, and AlexNet models. According to the experimental results, using the VGG classification method to classify continuous pollution data in the form of candlestick charts can more comprehensively retain the characteristics of the physical pollution process and provide a classification basis for accurately predicting PM_2.5 values. At the same time, the statistical feasibility of this method has been proved.