Machine learning models for predicting thermal desorption remediation of soils contaminated with polycyclic aromatic hydrocarbons.

Among various remediation methods for organic-contaminated soil, thermal desorption stands out due to its broad treatment range and high efficiency. Nonetheless, analyzing the contribution of factors in complex soil remediation systems and deducing the results under multiple conditions are challenging, given the complexities arising from diverse soil properties, heating conditions, and contaminant types. Machine learning (ML) methods serve as a powerful analytical tool that can extract meaningful insights from datasets and reveal hidden relationships. Due to insufficient research on soil thermal desorption for remediation of organic sites using ML methods, this study took organic pollutants represented by polycyclic aromatic hydrocarbons (PAHs) as the research object and sorted out a comprehensive data set containing >700 data points on the thermal desorption of soil contaminated with PAHs from published literature. Several ML models, including artificial neural network (ANN), random forest (RF), and support vector regression (SVR), were applied. Model optimization and regression fitting centered on soil remediation efficiency, with feature importance analysis conducted on soil and contaminant properties and heating conditions. This approach enabled the quantitative evaluation and prediction of thermal desorption remediation effects on soil contaminated with PAHs. Results indicated that ML models, particularly the RF model (R2 = 0.90), exhibited high accuracy in predicting remediation efficiency. The hierarchical significance of the features within the RF model is elucidated as follows: heating conditions account for 52 %, contaminant properties for 28 %, and soil properties for 20 % of the model's predictive power. A comprehensive analysis suggests that practical applications should emphasize heating conditions for efficient soil remediation. This research provides a crucial reference for optimizing and implementing thermal desorption in the quest for more efficient and reliable soil remediation strategies.

Toward a comprehensive understanding of alicyclic compounds: Bio-effects perspective and deep learning approach.

The escalating use of alicyclic compounds in modern industrial production has led to a rapid increase of these substances in the environment, posing significant health hazards. Addressing this challenge necessitates a comprehensive understanding of these compounds, which can be achieved through the deep learning approach. Graph neural networks (GNN) known for its' extraordinary ability to process graph data with rich relationships, have been employed in various molecular prediction tasks. In this study, alicyclic molecules screened from PCBA, Toxcast and Tox21 are made as general bioactivity and biological targets' activity prediction datasets. GNN-based models are trained on the two datasets, while the Attentive FP and PAGTN achieve best performance individually. In addition, alicyclic carbon atoms make the greatest contribution to biological activity, which indicate that the alicycle structures have significant impact on the carbon atoms' contribution. Moreover, there are terrific number of active molecules in other public datasets, indicates that alicyclic compounds deserve more attention in POPs control. This study uncovered deeper structural-activity relationships within these compounds, offering new perspectives and methodologies for academic research in the field.

Enhancement of anaerobic digestion from food waste via inert substances based on metagenomic analysis: Oxidative phosphorylation and metabolism.

The application of anaerobic digestion (AD) in the treatment of food waste (FW) has become widespread. However, the presence of inert substances, such as bones, ceramics, and shells, within FW introduces a degree of uncertainty into the AD process. To clarify this intricate issue, this study conducted an in-depth investigation into the influence of inert substances on AD. The results revealed that when inert substances were present at a concentration of 0.08 g/g VSS, methane productivity in the AD process was significantly augmented by 86%. Subsequent investigations suggested that this positive effect was primarily evident in various biochemical processes, including solubilization, hydrolysis acidification, methanogenesis, and the accumulation of extracellular polymeric substances. Metagenomic analysis showed that inert substances enhance the relative abundance of hydrolytic bacteria and have a pronounced impact on the relative abundance of hydrogenotrophic methanogens (Methanosarcina) and acetotrophic methanogens (Methanobacterium). Additionally, inert substances significantly increased the relative abundance of functional genes in oxidative phosphorylation, a pivotal pathway for ATP synthesis. Furthermore, inert substances had a substantial effect on the functional genes related to the metabolic pathways associated with methanogenesis (both hydrogenotrophic and acetotrophic). This comprehensive study shed light on the substantial impact of inert substances on the AD of food waste, contributing to an enhanced understanding of the underlying mechanisms of anaerobic fermentation.

Data-driven models applying in household hazardous waste: Amount prediction and classification in Shanghai.

Precisely predicting the amount of household hazardous waste (HHW) and classifying it intelligently is crucial for effective city management. Although data-driven models have the potential to address these problems, there have been few studies utilizing this approach for HHW prediction and classification due to the scarcity of available data. To address this, the current study employed the prophet model to forecast HHW quantities based on the Integration of Two Networks systems in Shanghai. HHW classification was performed using HVGGNet structures, which were based on VGG and transfer learning. To expedite the process of finding the optimal global learning rate, the method of cyclical learning rate was adopted, thus avoiding the need for repeated testing. Results showed that the average rate of HHW generation was 0.1 g/person/day, with the most significant waste categories being fluorescent lamps (30.6 %), paint barrels (26.1 %), medicine (26.2 %), battery (15.8 %), thermometer (0.03 %), and others (1.22 %). Recovering rare earth element (18.85 kg), Cd (3064.10 kg), Hg (15643.43 kg), Zn (14239.07 kg), Ag (11805.81 kg), Ni (4956.64 kg) and Li (1081.45 kg) from HHW can help avoid groundwater pollution, soil contamination and air pollution. HVGGNet-11 demonstrated 90.5 % precision and was deemed most suitable for HHW sorting. Furthermore, the prophet model predicted that HHW in Shanghai would increase from 794.43 t in 2020 to 2049.67 t in 2025.

Transformer-based enhanced model for accurate prediction and comprehensive analysis of hazardous waste generation in Shanghai: Implications for sustainable waste management strategies.

The escalating generation of hazardous waste (HW) has become a pressing concern worldwide, straining waste management systems and posing significant health hazards. Addressing this challenge necessitates an accurate understanding of HW generation, which can be achieved through the application of advanced models. The Transformer model, known for its ability to capture complex nonlinear processes, proves invaluable in extracting essential features and making precise HW generation predictions. To enhance comprehension of the key factors influencing HW generation, visualization techniques such as SHapley Additive exPlanations (SHAP) provide insightful explanations. In this study, a novel approach combining classical deep learning algorithms with the Transformer model is proposed, yielding impressive results with an R2 value of 0.953 and an RMSE of 7.284 for HW prediction. Notably, among the five key fields considered-demographics, socio-economics, industrial production, environmental governance, and medical health-industrial production emerges as the primary contributor, accounting for over 50% of HW generation. Moreover, a high rate of industrial development is anticipated to further accelerate this process.

Hi-MGT: A hybrid molecule graph transformer for toxicity identification.

Conventional toxicity testing methods that rely on animal experimentation are resource-intensive, time-consuming, and ethically controversial. Therefore, the development of alternative non-animal testing approaches is crucial. This study proposes a novel hybrid graph transformer architecture, termed Hi-MGT, for the toxicity identification. An innovative aggregation strategy, referred to as GNN-GT combination, enables Hi-MGT to simultaneously and comprehensively aggregate local and global structural information of molecules, thus elucidating more informative toxicity information hidden in molecule graphs. The results show that the state-of-the-art model outperforms current baseline CML and DL models on a diverse range of toxicity endpoints and is even comparable to large-scale pretrained GNNs with geometry enhancement. Additionally, the impact of hyperparameters on model performance is investigated, and a systematic ablation study is conducted to demonstrate the effectiveness of the GNN-GT combination. Moreover, this study provides valuable insights into the learning process on molecules and proposes a novel similarity-based method for toxic site detection, which could potentially facilitate toxicity identification and analysis. Overall, the Hi-MGT model represents a significant advancement in the development of alternative non-animal testing approaches for toxicity identification, with promising implications for enhancing human safety in the use of chemical compounds.

MSWNet: A visual deep machine learning method adopting transfer learning based upon ResNet 50 for municipal solid waste sorting.

An intelligent and efficient methodology is needed owning to the continuous increase of global municipal solid waste (MSW). This is because the common methods of manual and semi-mechanical screenings not only consume large amount of manpower and material resources but also accelerate virus community transmission. As the categories of MSW are diverse considering their compositions, chemical reactions, and processing procedures, etc., resulting in low efficiencies in MSW sorting using the traditional methods. Deep machine learning can help MSW sorting becoming into a smarter and more efficient mode. This study for the first time applied MSWNet in MSW sorting, a ResNet-50 with transfer learning. The method of cyclical learning rate was taken to avoid blind finding, and tests were repeated until accidentally encountering a good value. Measures of visualization were also considered to make the MSWNet model more transparent and accountable. Results showed transfer learning enhanced the efficiency of training time (from 741 s to 598.5 s), and improved the accuracy of recognition performance (from 88.50% to 93.50%); MSWNet showed a better performance in MSW classsification in terms of sensitivity (93.50%), precision (93.40%), F1-score (93.40%), accuracy (93.50%) and AUC (92.00%). The findings of this study can be taken as a reference for building the model MSW classification by deep learning, quantifying a suitable learning rate, and changing the data from high dimensions to two dimensions. Electronic Supplementary material: Supplementary material is available in the online version of this article at 10.1007/s11783-023-1677-1 and is accessible for authorized users.

Agglomeration-influenced transformation of heavy metals in gas-solid phases during simulated sewage sludge co-incineration: Effects of phosphorus and operating temperature.

Phosphorus and operating temperature not only affect the agglomeration behavior but also the transformation and migration of heavy metals. Accordingly, this study examined the effect of temperature and phosphorus in a fluidized bed combustion process to understand the emission and distribution of heavy metals by both experimental and thermodynamic calculations. The experimental results indicated that the sodium-phosphate reactions occur before the sodium-silicate reaction in the solid phase when the ratio of P/Na was 1/2. A low-melting-point sodium phosphate component, such as NaPO3, leads to easier particle agglomeration than Na2O-SiO2. In terms of the emissions of heavy metals, Pb and Cd show a similar trend: both the amount of emission smaller than that without adding phosphorus and the amount of emission share an upward trend with the operating time increased during MSS fluidized bed combustion. However, with the presence of phosphorus, the emission of Cr shows slightly decreased, and then sharply dropped, after that, increasing with operating time increased. Generally speaking, the maximum amount of Pb and Cd emitted was at 900 °C, followed by 800 °C and 700 °C. The higher temperature would promote the volatilization of Pb and Cd to emit. On the other hand, Cr emitted at the beginning tended to increase but later decreases when the temperatures were 700 and 900 °C, which may be due to the emission of Cr being influenced by the different affinities of both Al and Cr, reacting with Na in a fluidized bed incinerator. As for the distribution of heavy metals in the solid phase, a higher concentration of heavy metals was found in both the coarsest and finest particles during the process of agglomeration/defluidization.

Deep learning hybrid predictions for the amount of municipal solid waste: A case study in Shanghai.

It is crucial to precisely estimate the municipal solid waste (MSW) amount for its sustainable management. Owing to learning complicated and abstract features between the factors and target, deep learning has recently emerged as one of the useful tools with potential to predict the MSW amount. Therefore, this study aimed to design an MSW amount predicted system in Shanghai, consisting of Attention (A), one-dimensional convolutional neural network (C), and long short-term memory (L), to investigate the relationship between exogenous series (24 socioeconomics factors and past MSW amount) and target (MSW amount). The role of Attention, 1D-CNN, LSTM played on the MSW predicted amount also have investigated. The results show that attention is crucial for decoding the encoding information, which would improve performance between predicted and known MSW amount (R2 in A-L-C, L-A-C, L-C-A was 89.45%, 90.77%, and 95.31%, respectively.). CNN modules appear to be positioned similarly across the MSW predicted system. Finally, R2 in L-A-C, A-L-C, and A-C-L was 85.44%, 91.61%, and 89.45%, which suggested that LSTM as an intermediary between CNN and Attention modules seems a wise measure to predict the MSW amount based on the correlation efficiency. In addition, some socioeconomic factors including the average number of people in households and budget revenue may be chosen for the decision-making of MSW management in Shanghai city in the future, according to the weight of neurons in fully connected layers by the visual technology.

Applying a deep residual network coupling with transfer learning for recyclable waste sorting.

Recyclable waste sorting has become a key step for promoting the development of a circular economy with the gradual realization of carbon neutrality around the world. This study aims to develop an intelligent and efficient method for recyclable waste sorting by the method of deep learning. Thus, RWNet models, which refers to various ResNet structures (ResNet-18, ResNet-34, ResNet-50, ResNet-101, and ResNet-152) based on transfer learning, were proposed to classify different types of recyclable waste. Cyclical learning rate and data augmentation were taken to improve the performance of RWNet models. In addition, accuracy, precision, recall, F1 score, and ROC were taken to evaluate the performance of RWNet models. Results showed that the accuracy of various RWNet models is almost at 88%, and the best accuracy is 88.8% in RWNet-152. The highest precision, recall, and F1 score in terms of weighted average value appeared in RWNet-101 (89.9%), RWNet-152 (88.8%), and RWNet-152 (88.9%), respectively. The area under the ROC curve (AUC) is higher than 0.9, except for the AUC value of plastic (0.85), which indicated that most of the recyclable waste can be well sorted by RWNet models. This study demonstrates the good performance of RWNet models that can be used to automatically sort most of the recyclable waste, which paves the way for better recyclable waste management.

Deep convolutional neural networks for construction and demolition waste classification: VGGNet structures, cyclical learning rate, and knowledge transfer.

The sorting of Construction and Demolition (C&D) waste is a critical step to linking the recycling system and to the macro prediction, which helps to promote the development of the circular economy. Moreover, the effective classification and automated separation process will also help to stop the spreading of pathogenic organisms, such as virus and bacteria, by minimizing human intervention in the sorting process, while also helping to prevent further contamination by COVID-19 virus. This study aims to develop an efficient method to sort C&D waste through deep learning combined with knowledge transfer approach. In this paper, CVGGNet models, that is four VGG structures (VGGNet-11, VGGNet-13, VGGNet-16, and VGGNet-19), based on knowledge transfer combined with the technology of data augmentation and cyclical learning rate, are proposed to classify ten types of C&D waste images. Results show that 2.5 × 10-4, 1.8 × 10-4, 0.8 × 10-4, and 1.0 × 10-4 are the optimum learning rate for CVGGNet-11, CVGGNet-13, CVGGNet-16, and CVGGNet-19, respectively. Knowledge transfer helped shorten the training time from 1039.45 s to 991.05 s, and while it improved the performance of the CVGGNet-11 model in training, validation, and test datasets. The average training time increases as the number of the layers in the CVGGNet architecture rises: CVGGNet-11 (991.05 s) ˂ CVGGNet-13 (1025.76 s) ˂ CVGGNet-16 (1090.48 s) ˂ CVGGNet-19 (1337.81 s). Compared to other CVGGNet models, CVGGNet-16 showed an excellent performance in various C&D waste types, in terms of accuracy (76.6%), weighted average precision (76.8%), weighted average recall (76.6%), weighted average F1-score (76.6%) and micro average ROC (87.0%). In addition, the t-distributed Stochastic Neighbor Embedding (t-SNE) approach can reduce the dataset to a lower dimension and distinctly separate each type of C&D waste. This study demonstrates the good performance of CVGGNet models that can be used to automatically sort most of the C&D waste, paving the way for better C&D waste management.

Pyrolytic characteristics of fine materials from municipal solid waste using TG-FTIR, Py-GC/MS, and deep learning approach: Kinetics, thermodynamics, and gaseous products distribution.

Fine materials (FM) from municipal solid waste (MSW) classification require disposal, and pyrolysis is a feasible method for the treatments. Hence, the behavior, kinetics, and products of FM pyrolysis were investigated in this study. A deep learning algorithm was firstly employed to predict and verify the TG data during the process of FM pyrolysis. The results showed that FM pyrolysis could be divided into drying (<138 °C), de-volatilization (138-570 °C), and decomposition stage (≥570 °C above). The de-volatilization can further be divided into stage 2 and stage 3, with values of activation energy estimated by Flynn-Wall-Ozawa and Kissinger-Akahira-Sunose methods as 123.35 and 172.95 kJ/mol, respectively. The gas products like H2O, CO2, CH4, and CO, as well as functional groups like phenols and carbonyl (CO), were all detected during the process of FM pyrolysis by thermogravimetric-fourier transform infrared spectrometry at a heating rate of 10 °C/min. The main species detected by pyrolysis-gas chromatography-mass spectrometry analyzer included acid (41.98%) and aliphatic hydrocarbon (22.44%). Finally, the 1D-CNN-LSTM algorithm demonstrated an outstanding generalization capability to predict the relationship between FM composition and temperature, with R2 reaching 93.91%. In sum, this study provided a reference for the treatment of FM from MSW classification as well as the feasibility and practicability of deep learning applied in pyrolysis.

Comprehensive understanding the transition behaviors and mechanisms of chlorine and metal ions in municipal solid waste incineration fly ash during thermal treatment.

Municipal solid waste incineration fly ash is classified as the hazardous waste because of its high levels of heavy metals alkali chlorides, and polychlorinated dibenzo-p-dioxins. Thermal treatment is widely used for fly ash treatment because of its advantages of reduction and harmless. The transformation behaviors of chlorine and metal ions during the thermal treatment of fly ash has a significant impact on the harmless and resource of fly ash. At present, the migration behaviors of chlorine and metal ions during thermal treatment of fly ash is not clearly demonstrated. In this manuscript, the phase compositions, transformation behaviors, the variation of mass and content of chlorine and various metal ions were analyzed through diverse characterization methods under different sintering temperatures to understand the migration behaviors of chlorine and metal ions during thermal treatment. Roasting experiments showed that the migration behaviors of heavy metals and chlorides were consistent. The chlorine, sodium, potassium and heavy metal ions can be removed sharply while the calcium, aluminum, magnesium and iron were decreased slightly when the roasting temperature was above 750 °C. The findings also suggested that removed chlorides were soluble chlorides and unstable crystals in municipal solid waste incineration fly ash were inclined to formed steady structure under high temperature. The structure of roasted fly ash became denser and generated ceramic-like particle due to thermal agglomeration and chemical reactions.

Estimation of municipal solid waste amount based on one-dimension convolutional neural network and long short-term memory with attention mechanism model: A case study of Shanghai.

Municipal solid waste (MSW) amount has direct influence on MSW management, policy-decision making, and MSW treatment methods. Machine learning has great potential for prediction, but few studies apply the approaches of deep learning to forecast the quantity of MSW. Therefore, the aim of this study is to evaluate the feasibility and practicability of employing the methods of supervised learning, including Attention, one-dimension Convolutional Neural Network (1D-CNN) and Long Short-Term Memory (LSTM) to predict the MSW Amount in Shanghai. Integrated 1D-CNN and LSTM with Attention model, the new structure model (1D-CNN-LSTM-Attention, 1D-CLA), is designed to forecast MSW amount. In addition, the influence of socioeconomic factors on MSW amount, the structure and layers distribution of Attention, 1D-CNN, LSTM and 1D-CLA are also discussed. The results indicate that the correlation coefficients of Attention, one-dimension CNN, LSTM, and proposed 1D-CLA model to predict the MSW in Shanghai are 78%, 86.6%, 90%, and 95.3%, respectively, suggesting the feasible and practicable. The values of 24, 0.01, 50 and 25 for the number of neurons, dropout, the value of epoch number and Batch size best fit 1D-CLA to predict the amount of MSW in Shanghai. Furthermore, the performance of 1D-CLA is better than any single model or two model's combination (R2 is 95.3%) and the mechanism of 1D-CLA is contributed by three former models following the order: LSTM>CNN>Attention.

Sequential extraction for heavy metal distribution of bottom ash from fluidized bed co-combusted phosphorus-rich sludge under the agglomeration/defluidization process.

Agglomeration that occurs during municipal sewage sludge (MSS) fluidized bed co-combustion might affect heavy metal distribution and the transformation of bottom ash. A study on the mobility and speciation of heavy metals that accompanies agglomeration behavior and phosphorus addition should be examined during MSS co-combustion. Meanwhile, the aim of this study was to evaluate the total content and speciation of heavy metals during the MSS fluidized bed co-combustion by the chemical sequential extraction procedure (SEP). The risk assessment code (RAC) and individual contamination factor (ICF) are calculated to evaluate the mobility of heavy metals and their environmental risks in agglomerates. Moreover, identification of agglomerates is established by both characterization (scanning electron microscopy/energy-dispersive spectroscopy, X-ray photoelectron spectroscopy) and thermodynamic simulation (HSC chemistry software). The experimental results indicated that P and Na would form the lower melting-point compounds such as NaPO3 and Na2O in the bottom ash, which promoted agglomeration during MSS fluidized bed co-combustion. According to the simulation, Na and P have a stronger affinity than Si and Cr, and this reaction is not only influenced by particle agglomeration, but also by heavy metal distribution during modified MSS co-combustion. Nevertheless, the results of ICFs and RACs obtained from the SEP indicated that for heavy metals trapped in agglomerates, a weaker binding such as physical covering by eutectics might be considered as the dominant reaction compared with chemical binding to form a metal complex.