Adversarial distillation(AD)has emerged as a potential solution to tackle the challenging optimization problem of loss with hard labels in adversarial training.However,fixed sample-agnostic and student-egocentric atta...Adversarial distillation(AD)has emerged as a potential solution to tackle the challenging optimization problem of loss with hard labels in adversarial training.However,fixed sample-agnostic and student-egocentric attack strategies are unsuitable for distillation.Additionally,the reliability of guidance from static teachers diminishes as target models become more robust.This paper proposes an AD method called Learnable Distillation Attack Strategies and Evolvable Teachers Adversarial Distillation(LDAS&ET-AD).Firstly,a learnable distillation attack strategies generating mechanism is developed to automatically generate sample-dependent attack strategies tailored for distillation.A strategy model is introduced to produce attack strategies that enable adversarial examples(AEs)to be created in areas where the target model significantly diverges from the teachers by competing with the target model in minimizing or maximizing the AD loss.Secondly,a teacher evolution strategy is introduced to enhance the reliability and effectiveness of knowledge in improving the generalization performance of the target model.By calculating the experimentally updated target model’s validation performance on both clean samples and AEs,the impact of distillation from each training sample and AE on the target model’s generalization and robustness abilities is assessed to serve as feedback to fine-tune standard and robust teachers accordingly.Experiments evaluate the performance of LDAS&ET-AD against different adversarial attacks on the CIFAR-10 and CIFAR-100 datasets.The experimental results demonstrate that the proposed method achieves a robust precision of 45.39%and 42.63%against AutoAttack(AA)on the CIFAR-10 dataset for ResNet-18 and MobileNet-V2,respectively,marking an improvement of 2.31%and 3.49%over the baseline method.In comparison to state-of-the-art adversarial defense techniques,our method surpasses Introspective Adversarial Distillation,the top-performing method in terms of robustness under AA attack for the CIFAR-10 dataset,with enhancements of 1.40%and 1.43%for ResNet-18 and MobileNet-V2,respectively.These findings demonstrate the effectiveness of our proposed method in enhancing the robustness of deep learning networks(DNNs)against prevalent adversarial attacks when compared to other competing methods.In conclusion,LDAS&ET-AD provides reliable and informative soft labels to one of the most promising defense methods,AT,alleviating the limitations of untrusted teachers and unsuitable AEs in existing AD techniques.We hope this paper promotes the development of DNNs in real-world trust-sensitive fields and helps ensure a more secure and dependable future for artificial intelligence systems.展开更多
In this work,the ternary azeotrope of tert-butyl alcohol/ethyl acetate/water is separated by extractive distillation(ED)to recover the available constituents and protect the environment.Based on the conductor like shi...In this work,the ternary azeotrope of tert-butyl alcohol/ethyl acetate/water is separated by extractive distillation(ED)to recover the available constituents and protect the environment.Based on the conductor like shielding model and relative volatility method,ethylene glycol was selected as the extractant in the separation process.In addition,in view of the characteristic that the relative volatility between components changes with pressure,the multi-objective optimization method based on nondominated sorting genetic algorithm II optimizes the pressure and the amount of solvent cooperatively to avoid falling into the optimal local solution.Based on the optimal process parameters,the proposed heat-integrated process can reduce the gas emissions by 29.30%.The heat-integrated ED,further coupled with the pervaporation process,can reduce gas emission by 42.36%and has the highest exergy efficiency of 47.56%.In addition,based on the heat-integrated process,the proposed two heat pump assisted heat-integrated ED processes show good economic and environmental performance.The double heat pump assisted heat-integrated ED can reduce the total annual cost by 28.78%and the gas emissions by 55.83%compared with the basis process,which has a good application prospect.This work provides a feasible approach for the separation of ternary azeotropes.展开更多
Neural networks are often viewed as pure‘black box’models,lacking interpretability and extrapolation capabilities of pure mechanistic models.This work proposes a new approach that,with the help of neural networks,im...Neural networks are often viewed as pure‘black box’models,lacking interpretability and extrapolation capabilities of pure mechanistic models.This work proposes a new approach that,with the help of neural networks,improves the conformity of the first-principal model to the actual plant.The final result is still a first-principal model rather than a hybrid model,which maintains the advantage of the high interpretability of first-principal model.This work better simulates industrial batch distillation which separates four components:water,ethylene glycol,diethylene glycol,and triethylene glycol.GRU(gated recurrent neural network)and LSTM(long short-term memory)were used to obtain empirical parameters of mechanistic model that are difficult to measure directly.These were used to improve the empirical processes in mechanistic model,thus correcting unreasonable model assumptions and achieving better predictability for batch distillation.The proposed method was verified using a case study from one industrial plant case,and the results show its advancement in improving model predictions and the potential to extend to other similar systems.展开更多
A huge amount of energy is always consumed to separate the ternary azeotropic mixtures by distillations.The heterogeneous azeotropic distillation and the pressure-swing distillation are two kinds of effective technolo...A huge amount of energy is always consumed to separate the ternary azeotropic mixtures by distillations.The heterogeneous azeotropic distillation and the pressure-swing distillation are two kinds of effective technologies to separate heterogeneous azeotropes without entrainer addition.To give better play to the synergistic energy-saving effect of these two processes,a novel pressure-swing-assisted ternary heterogeneous azeotropic distillation(THAD)process is proposed firstly.In this process,the ternary heterogeneous azeotrope is decanted into two liquid phases before being refluxed into the azeotropic distillation column to avoid the aqueous phase remixing,and three columns'pressures are modified to decrease the flowrates of the recycle streams.Then the dividing wall column and heat integration technologies are introduced to further reduce its energy consumption,and the pressureswing-assisted ternary heterogeneous azeotropic dividing-wall column and its heat integration structure are achieved.A genetic algorithm procedure is used to optimize the proposed processes.The design results show that the proposed processes have higher energy efficiencies and lower CO_(2)emissions than the published THAD process.展开更多
Here we demonstrate the proof-of-concept for microchannel reactive distillation for alcohol-to-jet application:combining ethanol/water separation and ethanol dehydration in one unit operation.Ethanol is first distille...Here we demonstrate the proof-of-concept for microchannel reactive distillation for alcohol-to-jet application:combining ethanol/water separation and ethanol dehydration in one unit operation.Ethanol is first distilled into the vapor phase,converted to ethylene and water,and then the water co-product is condensed to shift the reaction equilibrium.Process intensification is achieved through rapid mass transfer-ethanol stripping from thin wicks using novel microchannel architectures-leading to lower residence time and improved separation efficiency.Energy savings are realized with integration of unit operations.For example,heat of condensing water can offset vaporizing ethanol.Furthermore,the dehydration reaction equilibrium shifts towards completion by immediate removal of the water byproduct upon formation while maintaining aqueous feedstock in the condensed phase.For aqueous ethanol feedstock(40%_w),71% ethanol conversion with 91% selectivity to ethylene was demonstrated at 220℃,600psig,and 0.28 h^(-1) wt hour space velocity.2.7 stages of separation were also demonstrated,under these conditions,using a device length of 8.3 cm.This provides a height equivalent of a theoretical plate(HETP),a measure of separation efficiency,of ^(3).3 cm.By comparison,conventional distillation packing provides an HETP of ^(3)0 cm.Thus,9,1 × reduction in HETP was demonstrated over conventional technology,providing a means for significant energy savings and an example of process intensification.Finally,preliminary process economic analysis indicates that by using microchannel reactive distillation technology,the operating and capital costs for the ethanol separation and dehydration portion of an envisioned alcoholto-jet process could be reduced by at least 35% and 55%,respectively,relative to the incumbent technology,provided future improvements to microchannel reactive distillation design and operability are made.展开更多
This study explored the synergistic interaction of sewage sludge(SS)and distillation residue(DR)during co-pyrolysis for the optimized treatment of sewage sludge in cement kiln systems,utilizing thermogravimetric analy...This study explored the synergistic interaction of sewage sludge(SS)and distillation residue(DR)during co-pyrolysis for the optimized treatment of sewage sludge in cement kiln systems,utilizing thermogravimetric analysis(TGA)and thermogravimetric analysis with mass spectrometry(TGA-MS).The results reveal the coexisting synergistic and antagonistic effects in the co-pyrolysis of SS/DR.The synergistic effect arises from hydrogen free radicals in SS and catalytic components in ash fractions,while the antagonistic effect is mainly due to the melting of DR on the surface of SS particles during pyrolysis and the reaction of SS ash with alkali metals to form inert substances.SS/DR co-pyrolysis reduces the yielding of coke and gas while increasing tar production.This study will promote the reduction,recycling,and harmless treatment of hazardous solid waste.展开更多
Time-frequency analysis is a successfully used tool for analyzing the local features of seismic data.However,it suffers from several inevitable limitations,such as the restricted time-frequency resolution,the difficul...Time-frequency analysis is a successfully used tool for analyzing the local features of seismic data.However,it suffers from several inevitable limitations,such as the restricted time-frequency resolution,the difficulty in selecting parameters,and the low computational efficiency.Inspired by deep learning,we suggest a deep learning-based workflow for seismic time-frequency analysis.The sparse S transform network(SSTNet)is first built to map the relationship between synthetic traces and sparse S transform spectra,which can be easily pre-trained by using synthetic traces and training labels.Next,we introduce knowledge distillation(KD)based transfer learning to re-train SSTNet by using a field data set without training labels,which is named the sparse S transform network with knowledge distillation(KD-SSTNet).In this way,we can effectively calculate the sparse time-frequency spectra of field data and avoid the use of field training labels.To test the availability of the suggested KD-SSTNet,we apply it to field data to estimate seismic attenuation for reservoir characterization and make detailed comparisons with the traditional time-frequency analysis methods.展开更多
Research on panicle detection is one of the most important aspects of paddy phenotypic analysis.A phenotyping method that uses unmanned aerial vehicles can be an excellent alternative to field-based methods.Neverthele...Research on panicle detection is one of the most important aspects of paddy phenotypic analysis.A phenotyping method that uses unmanned aerial vehicles can be an excellent alternative to field-based methods.Nevertheless,it entails many other challenges,including different illuminations,panicle sizes,shape distortions,partial occlusions,and complex backgrounds.Object detection algorithms are directly affected by these factors.This work proposes a model for detecting panicles called Border Sensitive Knowledge Distillation(BSKD).It is designed to prioritize the preservation of knowledge in border areas through the use of feature distillation.Our feature-based knowledge distillation method allows us to compress the model without sacrificing its effectiveness.An imitation mask is used to distinguish panicle-related foreground features from irrelevant background features.A significant improvement in Unmanned Aerial Vehicle(UAV)images is achieved when students imitate the teacher’s features.On the UAV rice imagery dataset,the proposed BSKD model shows superior performance with 76.3%mAP,88.3%precision,90.1%recall and 92.6%F1 score.展开更多
Knowledge distillation(KD) enhances student network generalization by transferring dark knowledge from a complex teacher network. To optimize computational expenditure and memory utilization, self-knowledge distillati...Knowledge distillation(KD) enhances student network generalization by transferring dark knowledge from a complex teacher network. To optimize computational expenditure and memory utilization, self-knowledge distillation(SKD) extracts dark knowledge from the model itself rather than an external teacher network. However, previous SKD methods performed distillation indiscriminately on full datasets, overlooking the analysis of representative samples. In this work, we present a novel two-stage approach to providing targeted knowledge on specific samples, named two-stage approach self-knowledge distillation(TOAST). We first soften the hard targets using class medoids generated based on logit vectors per class. Then, we iteratively distill the under-trained data with past predictions of half the batch size. The two-stage knowledge is linearly combined, efficiently enhancing model performance. Extensive experiments conducted on five backbone architectures show our method is model-agnostic and achieves the best generalization performance.Besides, TOAST is strongly compatible with existing augmentation-based regularization methods. Our method also obtains a speedup of up to 2.95x compared with a recent state-of-the-art method.展开更多
Knowledge distillation,as a pivotal technique in the field of model compression,has been widely applied across various domains.However,the problem of student model performance being limited due to inherent biases in t...Knowledge distillation,as a pivotal technique in the field of model compression,has been widely applied across various domains.However,the problem of student model performance being limited due to inherent biases in the teacher model during the distillation process still persists.To address the inherent biases in knowledge distillation,we propose a de-biased knowledge distillation framework tailored for binary classification tasks.For the pre-trained teacher model,biases in the soft labels are mitigated through knowledge infusion and label de-biasing techniques.Based on this,a de-biased distillation loss is introduced,allowing the de-biased labels to replace the soft labels as the fitting target for the student model.This approach enables the student model to learn from the corrected model information,achieving high-performance deployment on lightweight student models.Experiments conducted on multiple real-world datasets demonstrate that deep learning models compressed under the de-biased knowledge distillation framework significantly outperform traditional response-based and feature-based knowledge distillation models across various evaluation metrics,highlighting the effectiveness and superiority of the de-biased knowledge distillation framework in model compression.展开更多
Unmanned aerial vehicles(UAVs) have gained significant attention in practical applications, especially the low-altitude aerial(LAA) object detection imposes stringent requirements on recognition accuracy and computati...Unmanned aerial vehicles(UAVs) have gained significant attention in practical applications, especially the low-altitude aerial(LAA) object detection imposes stringent requirements on recognition accuracy and computational resources. In this paper, the LAA images-oriented tensor decomposition and knowledge distillation-based network(TDKD-Net) is proposed,where the TT-format TD(tensor decomposition) and equalweighted response-based KD(knowledge distillation) methods are designed to minimize redundant parameters while ensuring comparable performance. Moreover, some robust network structures are developed, including the small object detection head and the dual-domain attention mechanism, which enable the model to leverage the learned knowledge from small-scale targets and selectively focus on salient features. Considering the imbalance of bounding box regression samples and the inaccuracy of regression geometric factors, the focal and efficient IoU(intersection of union) loss with optimal transport assignment(F-EIoU-OTA)mechanism is proposed to improve the detection accuracy. The proposed TDKD-Net is comprehensively evaluated through extensive experiments, and the results have demonstrated the effectiveness and superiority of the developed methods in comparison to other advanced detection algorithms, which also present high generalization and strong robustness. As a resource-efficient precise network, the complex detection of small and occluded LAA objects is also well addressed by TDKD-Net, which provides useful insights on handling imbalanced issues and realizing domain adaptation.展开更多
Strabismus significantly impacts human health as a prevalent ophthalmic condition.Early detection of strabismus is crucial for effective treatment and prognosis.Traditional deep learning models for strabismus detectio...Strabismus significantly impacts human health as a prevalent ophthalmic condition.Early detection of strabismus is crucial for effective treatment and prognosis.Traditional deep learning models for strabismus detection often fail to estimate prediction certainty precisely.This paper employed a Bayesian deep learning algorithm with knowledge distillation,improving the model's performance and uncertainty estimation ability.Trained on 6807 images from two tertiary hospitals,the model showed significantly higher diagnostic accuracy than traditional deep-learning models.Experimental results revealed that knowledge distillation enhanced the Bayesian model’s performance and uncertainty estimation ability.These findings underscore the combined benefits of using Bayesian deep learning algorithms and knowledge distillation,which improve the reliability and accuracy of strabismus diagnostic predictions.展开更多
In recent years,anomaly detection has attracted much attention in industrial production.As traditional anomaly detection methods usually rely on direct comparison of samples,they often ignore the intrinsic relationshi...In recent years,anomaly detection has attracted much attention in industrial production.As traditional anomaly detection methods usually rely on direct comparison of samples,they often ignore the intrinsic relationship between samples,resulting in poor accuracy in recognizing anomalous samples.To address this problem,a knowledge distillation anomaly detection method based on feature reconstruction was proposed in this study.Knowledge distillation was performed after inverting the structure of the teacher-student network to avoid the teacher-student network sharing the same inputs and similar structure.Representability was improved by using feature splicing to unify features at different levels,and the merged features were processed and reconstructed using an improved Transformer.The experimental results show that the proposed method achieves better performance on the MVTec dataset,verifying its effectiveness and feasibility in anomaly detection tasks.This study provides a new idea to improve the accuracy and efficiency of anomaly detection.展开更多
In this study,the effect of yeast strains(X16 and RMS2),fruit seed,and pectinase on the quality of apple distilled spirits were investigated with apple as material.The results showed that pectinase shortened the ferme...In this study,the effect of yeast strains(X16 and RMS2),fruit seed,and pectinase on the quality of apple distilled spirits were investigated with apple as material.The results showed that pectinase shortened the fermentation period.The strain X16,fruit seed remaining,and pectinase addition groups had higher yields of crude distilled spirits than the strain RMS2,fruit seed removal,and without pectinase groups,respectively.Regarding the first-grade distilled spirits quality,the X16 group had higher content of total acids,total esters,higher alcohols,and alcohol content than the RMS2 group;the group with fruit seeds had higher total acids but lower alcohol content and total esters than the group with fruit seed removal;the group with pectinase addition had higher total acids and alcohol content than the group without pectinase addition.Regarding the second-grade distilled spirits quality,the X16 group had higher total acids,total esters,and alcohol content than the RMS2 group;the group without fruit seeds had higher alcohol content than the group with fruit seeds;the group with pectinase addition showcased higher total acids but lower alcohol content than the group without pectinase addition.In summary,yeast strains and pectinase affected the quality of apple distilled spirits,and strain X16 was more suitable for brewing apple distilled spirits.Pectinase affected the quality of apple distilled spirits by affecting fermentation rate and temperature.展开更多
The experimental and analytical investigation was conducted on a solar-powered single-effect distillation (SED). The evaporator was designed to be an electrical evaporator as opposed to the steam evaporator that exist...The experimental and analytical investigation was conducted on a solar-powered single-effect distillation (SED). The evaporator was designed to be an electrical evaporator as opposed to the steam evaporator that existed previously. Using sun-tracking solar panels, the electrical evaporator in the designed distillation unit was powered by solar energy. Approximately 20 kWh was utilized by the small-scale distillation apparatus. This type of design is mobile, so remote areas and countries with fragile economies can utilize it on a small or large scale. Utilizing the principles of energy and mass conservation, the amount of distillate water and power required for a single unit was determined, at the low salinity (2200 PPM) with fixed boiling point temperature (Tb = 75˚C), the unit performance is approx. 98.4%. The experimental results and those derived from a mathematical model were compared, and both showed strong accord. Using engineering equation solver (EES) software, a computer program was developed for this research scenario.展开更多
This paper delves into the comparative study of tray and packed column pressure swing distillation systems, focusing on the separation of a ternary mixture containing ethanol, tetrahydrofuran (THF), and water. The stu...This paper delves into the comparative study of tray and packed column pressure swing distillation systems, focusing on the separation of a ternary mixture containing ethanol, tetrahydrofuran (THF), and water. The study particularly emphasizes the production of 99.5 w/w% tetrahydrofuran from the downstream product of 1,4-butanediol synthesis via diethyl maleate. Pro/II simulation software is utilized to explore various system configurations, including sieve trays, valve trays, and packed columns. Material and energy balances are performed to ascertain stream compositions and energy demands. The investigation encompasses the effects of column operating pressure on condenser and reboiler temperatures, as well as the implications of utility streams. A rigorous distillation model is employed to compare valve tray, sieve tray, and random packing (utilizing Norton Super Intalox) column designs by varying the number of trays, reflux ratio, and second distillation column pressure. Heat exchangers are integrated into the model, and their areas and utility flow rates are computed and integrated into the economic assessment. Economic analysis, guided by Net Present Value (NPV) calculations over a 20-year span, drives the selection of the most cost-effective design. Results demonstrate that while all designs are energy-efficient, the packed column system emerges as the most economical choice, offering a comprehensive framework for the separation process. Furthermore, optimal design configurations and operating conditions for both tray and packed column systems are outlined, providing valuable insights for industrial applications.展开更多
Extractive distillation(ED)and solvent-assisted pressure-swing distillation(SA-PSD)are both special distillation processes that perform good at separating pressure-insensitive azeotropes.However,few reported studies h...Extractive distillation(ED)and solvent-assisted pressure-swing distillation(SA-PSD)are both special distillation processes that perform good at separating pressure-insensitive azeotropes.However,few reported studies have compared the performance of the two processes.In this paper,ED processes with N-methylpyrrolidone(NMP)and dimethlac-etamide(DMCA)as entrainer,SA-PSD process with isopropyl-alcohol(IPA)as solvent and SA-PSD process with partial heat integration(PHI-PSD)are proposed to achieve high purity separation of a mixture of cyclohexane/2-butanol system.The optimal operating conditions of the processes are obtained after optimizing with NSGA-Ⅱ algorithm when total annual cost(TAC)and the entropy production of process are set as objectives.The optimal results show that the optimal PHI-PSD process has lower TAC by 28.7% and the lower entropy production by 39.5% than the optimal SA-PSD process while the ED process with NMP as entrainer has lower TAC by 50.9% and the lower entropy production by 56.1% than the optimal SA-PSD process.The optimal results show that the ED process with NMP as entrainer has the best economic and thermodynamic efficiency among the four proposed processes in this paper.展开更多
This wok proposed the extraction distillation coupled pervaporation(ED+PV) technology process using two different solvents to separate isopropanol(IPA) and diisopropyl ether(DIPE) from DIPE/IPA/H_(2)O ternary heteroge...This wok proposed the extraction distillation coupled pervaporation(ED+PV) technology process using two different solvents to separate isopropanol(IPA) and diisopropyl ether(DIPE) from DIPE/IPA/H_(2)O ternary heterogeneous azeotropes in industrial wastewater from the synthesis of isopropanol in this study.Based on strict design specifications, simulation and sequential iteration methods are used for process design and optimization. Compared to the ethylene glycol(EG)-EG+H_(2)O process and the 1,3-propanediol(PDO)-IPA+H_(2)O process, the total annual cost(TAC) of the EG-IPA+H_(2)O process decreased by 20.76% and 7.86%(PDO). Compared to the EG-EG+H_(2)O process, the TAC of the PDO-IPA+H_(2)O process reduced 14%, but the global warming potential(GWP) and human toxicity of the PDO-IPA+H_(2)O process increased 11.3% and 4.07% respectively. Compared to the PDO-IPA+H_(2)O process, the EG-IPA+H_(2)O process saves 7.86%(TAC), 9.78%(GWP) and 9.85%(human toxicity). The ED+PV process with EG is superior to PDO in factors of TAC, energy consumption, human toxicity and environment. The EG-IPA+H_(2)O process changed the separation order of the products of the multi-azeotropic system, reduced the cost and energy conservation of the system, and enhanced the environmental protection evaluation of the process, is the best process through life cycle assessment for analyzing the economy, energy conservation, environmental assessment and human toxicity, designing cleaner products, controlling waste discharge, and promoting the chemical purification industry. This work provides a new process design and optimized separation ideas, will have a good guiding significance for the research and application separation of multi-azeotropic mixture with mixed solvents in organic wastewater from the cleaner chemical production, has been up to standard wastewater discharge process, and realized the development goal of carbon peak and carbon neutrality in the sustainable development of chemical clean industry.展开更多
The conventional distillation is hard to accomplish the separation of acetonitrile/ethyl acetate/n-hexane mixture. Herein, a heterogeneous azeotropic distillation(HAD) without adding entrainer is proposed to separate ...The conventional distillation is hard to accomplish the separation of acetonitrile/ethyl acetate/n-hexane mixture. Herein, a heterogeneous azeotropic distillation(HAD) without adding entrainer is proposed to separate ternary mixture. The proposed scheme is optimized via the simulated annealing algorithm and minimum total annual cost(TAC) is used as objective functions. To minimize energy consumption,heat pump is added on the basis of optimal heterogeneous azeotropic distillation and heat integration technology is used to further improve the energy recovery. The TAC, gas emission, energy consumption and exergy destruction are used to discuss the economy and environmental protection of processes.Among all the processes, the heat pump with higher preheating temperature(HPT) assisted HAD process by combining with heat integration(HAD-HPT-HI) has best performances on economic, environment,energy and exergy. Compared with conventional HAD process, the HAD-HPT-HI achieves the reductions of 52.17%, 68.86%, 65.87% and 65.46% on TAC, total energy consumption, gas emissions and exergy destruction, respectively.展开更多
A green and effective electrolytic process was developed to produce high-purity Mg metal using primary and secondary resources containing Mg O as a feedstock. The electrolysis of various Mg O resources was conducted u...A green and effective electrolytic process was developed to produce high-purity Mg metal using primary and secondary resources containing Mg O as a feedstock. The electrolysis of various Mg O resources was conducted using a Cu cathode in MgF2– LiF – KCl molten salt at 1043 K by applying an average current of 1.44 A for 12.5 h. The electrolysis of calcined North Korean magnesite and seawater Mg O clinker yielded Mg alloys of MgCu2and(Cu) phases with current efficiencies of 89.6–92.4%. The electrolysis of oxidized Mg O-C refractory brick, aged ferronickel slag, and ferronickel slag yielded Mg alloys of MgCu2and(Cu) phases with current efficiencies of 59.3–92.3%. The vacuum distillation of Mg alloys obtained was conducted at 1300 K for 10 h to produce high-purity Mg metal. After vacuum distillation, Mg metal with a purity of above 99.994% was obtained. Therefore, this study demonstrates the feasibility of the production of high-purity Mg metal from various Mg O resources using a novel electrolytic process with a Cu cathode, followed by vacuum distillation.展开更多
基金the National Key Research and Development Program of China(2021YFB1006200)Major Science and Technology Project of Henan Province in China(221100211200).Grant was received by S.Li.
文摘Adversarial distillation(AD)has emerged as a potential solution to tackle the challenging optimization problem of loss with hard labels in adversarial training.However,fixed sample-agnostic and student-egocentric attack strategies are unsuitable for distillation.Additionally,the reliability of guidance from static teachers diminishes as target models become more robust.This paper proposes an AD method called Learnable Distillation Attack Strategies and Evolvable Teachers Adversarial Distillation(LDAS&ET-AD).Firstly,a learnable distillation attack strategies generating mechanism is developed to automatically generate sample-dependent attack strategies tailored for distillation.A strategy model is introduced to produce attack strategies that enable adversarial examples(AEs)to be created in areas where the target model significantly diverges from the teachers by competing with the target model in minimizing or maximizing the AD loss.Secondly,a teacher evolution strategy is introduced to enhance the reliability and effectiveness of knowledge in improving the generalization performance of the target model.By calculating the experimentally updated target model’s validation performance on both clean samples and AEs,the impact of distillation from each training sample and AE on the target model’s generalization and robustness abilities is assessed to serve as feedback to fine-tune standard and robust teachers accordingly.Experiments evaluate the performance of LDAS&ET-AD against different adversarial attacks on the CIFAR-10 and CIFAR-100 datasets.The experimental results demonstrate that the proposed method achieves a robust precision of 45.39%and 42.63%against AutoAttack(AA)on the CIFAR-10 dataset for ResNet-18 and MobileNet-V2,respectively,marking an improvement of 2.31%and 3.49%over the baseline method.In comparison to state-of-the-art adversarial defense techniques,our method surpasses Introspective Adversarial Distillation,the top-performing method in terms of robustness under AA attack for the CIFAR-10 dataset,with enhancements of 1.40%and 1.43%for ResNet-18 and MobileNet-V2,respectively.These findings demonstrate the effectiveness of our proposed method in enhancing the robustness of deep learning networks(DNNs)against prevalent adversarial attacks when compared to other competing methods.In conclusion,LDAS&ET-AD provides reliable and informative soft labels to one of the most promising defense methods,AT,alleviating the limitations of untrusted teachers and unsuitable AEs in existing AD techniques.We hope this paper promotes the development of DNNs in real-world trust-sensitive fields and helps ensure a more secure and dependable future for artificial intelligence systems.
基金supported by the National Natural Science Foundation of China(22178188).
文摘In this work,the ternary azeotrope of tert-butyl alcohol/ethyl acetate/water is separated by extractive distillation(ED)to recover the available constituents and protect the environment.Based on the conductor like shielding model and relative volatility method,ethylene glycol was selected as the extractant in the separation process.In addition,in view of the characteristic that the relative volatility between components changes with pressure,the multi-objective optimization method based on nondominated sorting genetic algorithm II optimizes the pressure and the amount of solvent cooperatively to avoid falling into the optimal local solution.Based on the optimal process parameters,the proposed heat-integrated process can reduce the gas emissions by 29.30%.The heat-integrated ED,further coupled with the pervaporation process,can reduce gas emission by 42.36%and has the highest exergy efficiency of 47.56%.In addition,based on the heat-integrated process,the proposed two heat pump assisted heat-integrated ED processes show good economic and environmental performance.The double heat pump assisted heat-integrated ED can reduce the total annual cost by 28.78%and the gas emissions by 55.83%compared with the basis process,which has a good application prospect.This work provides a feasible approach for the separation of ternary azeotropes.
基金supported by Beijing Natural Science Foundation(2222037)by the Fundamental Research Funds for the Central Universities.
文摘Neural networks are often viewed as pure‘black box’models,lacking interpretability and extrapolation capabilities of pure mechanistic models.This work proposes a new approach that,with the help of neural networks,improves the conformity of the first-principal model to the actual plant.The final result is still a first-principal model rather than a hybrid model,which maintains the advantage of the high interpretability of first-principal model.This work better simulates industrial batch distillation which separates four components:water,ethylene glycol,diethylene glycol,and triethylene glycol.GRU(gated recurrent neural network)and LSTM(long short-term memory)were used to obtain empirical parameters of mechanistic model that are difficult to measure directly.These were used to improve the empirical processes in mechanistic model,thus correcting unreasonable model assumptions and achieving better predictability for batch distillation.The proposed method was verified using a case study from one industrial plant case,and the results show its advancement in improving model predictions and the potential to extend to other similar systems.
基金supported by the National Natural Science Foundation of China(21406170)the State Key Laboratory of Chemical Engineering(SKL-ChE-22B02).
文摘A huge amount of energy is always consumed to separate the ternary azeotropic mixtures by distillations.The heterogeneous azeotropic distillation and the pressure-swing distillation are two kinds of effective technologies to separate heterogeneous azeotropes without entrainer addition.To give better play to the synergistic energy-saving effect of these two processes,a novel pressure-swing-assisted ternary heterogeneous azeotropic distillation(THAD)process is proposed firstly.In this process,the ternary heterogeneous azeotrope is decanted into two liquid phases before being refluxed into the azeotropic distillation column to avoid the aqueous phase remixing,and three columns'pressures are modified to decrease the flowrates of the recycle streams.Then the dividing wall column and heat integration technologies are introduced to further reduce its energy consumption,and the pressureswing-assisted ternary heterogeneous azeotropic dividing-wall column and its heat integration structure are achieved.A genetic algorithm procedure is used to optimize the proposed processes.The design results show that the proposed processes have higher energy efficiencies and lower CO_(2)emissions than the published THAD process.
基金financially U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, Bioenergy Technologies Office, and the Office of Technology Transitions Technology Commercialization FundFinancial support also was provided by Lanza Tech through a Cooperative Research and Development Agreement。
文摘Here we demonstrate the proof-of-concept for microchannel reactive distillation for alcohol-to-jet application:combining ethanol/water separation and ethanol dehydration in one unit operation.Ethanol is first distilled into the vapor phase,converted to ethylene and water,and then the water co-product is condensed to shift the reaction equilibrium.Process intensification is achieved through rapid mass transfer-ethanol stripping from thin wicks using novel microchannel architectures-leading to lower residence time and improved separation efficiency.Energy savings are realized with integration of unit operations.For example,heat of condensing water can offset vaporizing ethanol.Furthermore,the dehydration reaction equilibrium shifts towards completion by immediate removal of the water byproduct upon formation while maintaining aqueous feedstock in the condensed phase.For aqueous ethanol feedstock(40%_w),71% ethanol conversion with 91% selectivity to ethylene was demonstrated at 220℃,600psig,and 0.28 h^(-1) wt hour space velocity.2.7 stages of separation were also demonstrated,under these conditions,using a device length of 8.3 cm.This provides a height equivalent of a theoretical plate(HETP),a measure of separation efficiency,of ^(3).3 cm.By comparison,conventional distillation packing provides an HETP of ^(3)0 cm.Thus,9,1 × reduction in HETP was demonstrated over conventional technology,providing a means for significant energy savings and an example of process intensification.Finally,preliminary process economic analysis indicates that by using microchannel reactive distillation technology,the operating and capital costs for the ethanol separation and dehydration portion of an envisioned alcoholto-jet process could be reduced by at least 35% and 55%,respectively,relative to the incumbent technology,provided future improvements to microchannel reactive distillation design and operability are made.
基金Funded by National College Student Innovation and Entrepreneurship Training Program Project(No.CY202036)。
文摘This study explored the synergistic interaction of sewage sludge(SS)and distillation residue(DR)during co-pyrolysis for the optimized treatment of sewage sludge in cement kiln systems,utilizing thermogravimetric analysis(TGA)and thermogravimetric analysis with mass spectrometry(TGA-MS).The results reveal the coexisting synergistic and antagonistic effects in the co-pyrolysis of SS/DR.The synergistic effect arises from hydrogen free radicals in SS and catalytic components in ash fractions,while the antagonistic effect is mainly due to the melting of DR on the surface of SS particles during pyrolysis and the reaction of SS ash with alkali metals to form inert substances.SS/DR co-pyrolysis reduces the yielding of coke and gas while increasing tar production.This study will promote the reduction,recycling,and harmless treatment of hazardous solid waste.
基金supported by the National Natural Science Foundation of China (42274144,42304122,and 41974155)the Key Research and Development Program of Shaanxi (2023-YBGY-076)+1 种基金the National Key R&D Program of China (2020YFA0713404)the China Uranium Industry and East China University of Technology Joint Innovation Fund (NRE202107)。
文摘Time-frequency analysis is a successfully used tool for analyzing the local features of seismic data.However,it suffers from several inevitable limitations,such as the restricted time-frequency resolution,the difficulty in selecting parameters,and the low computational efficiency.Inspired by deep learning,we suggest a deep learning-based workflow for seismic time-frequency analysis.The sparse S transform network(SSTNet)is first built to map the relationship between synthetic traces and sparse S transform spectra,which can be easily pre-trained by using synthetic traces and training labels.Next,we introduce knowledge distillation(KD)based transfer learning to re-train SSTNet by using a field data set without training labels,which is named the sparse S transform network with knowledge distillation(KD-SSTNet).In this way,we can effectively calculate the sparse time-frequency spectra of field data and avoid the use of field training labels.To test the availability of the suggested KD-SSTNet,we apply it to field data to estimate seismic attenuation for reservoir characterization and make detailed comparisons with the traditional time-frequency analysis methods.
文摘Research on panicle detection is one of the most important aspects of paddy phenotypic analysis.A phenotyping method that uses unmanned aerial vehicles can be an excellent alternative to field-based methods.Nevertheless,it entails many other challenges,including different illuminations,panicle sizes,shape distortions,partial occlusions,and complex backgrounds.Object detection algorithms are directly affected by these factors.This work proposes a model for detecting panicles called Border Sensitive Knowledge Distillation(BSKD).It is designed to prioritize the preservation of knowledge in border areas through the use of feature distillation.Our feature-based knowledge distillation method allows us to compress the model without sacrificing its effectiveness.An imitation mask is used to distinguish panicle-related foreground features from irrelevant background features.A significant improvement in Unmanned Aerial Vehicle(UAV)images is achieved when students imitate the teacher’s features.On the UAV rice imagery dataset,the proposed BSKD model shows superior performance with 76.3%mAP,88.3%precision,90.1%recall and 92.6%F1 score.
基金supported by the National Natural Science Foundation of China (62176061)。
文摘Knowledge distillation(KD) enhances student network generalization by transferring dark knowledge from a complex teacher network. To optimize computational expenditure and memory utilization, self-knowledge distillation(SKD) extracts dark knowledge from the model itself rather than an external teacher network. However, previous SKD methods performed distillation indiscriminately on full datasets, overlooking the analysis of representative samples. In this work, we present a novel two-stage approach to providing targeted knowledge on specific samples, named two-stage approach self-knowledge distillation(TOAST). We first soften the hard targets using class medoids generated based on logit vectors per class. Then, we iteratively distill the under-trained data with past predictions of half the batch size. The two-stage knowledge is linearly combined, efficiently enhancing model performance. Extensive experiments conducted on five backbone architectures show our method is model-agnostic and achieves the best generalization performance.Besides, TOAST is strongly compatible with existing augmentation-based regularization methods. Our method also obtains a speedup of up to 2.95x compared with a recent state-of-the-art method.
基金supported by the National Natural Science Foundation of China under Grant No.62172056Young Elite Scientists Sponsorship Program by CAST under Grant No.2022QNRC001.
文摘Knowledge distillation,as a pivotal technique in the field of model compression,has been widely applied across various domains.However,the problem of student model performance being limited due to inherent biases in the teacher model during the distillation process still persists.To address the inherent biases in knowledge distillation,we propose a de-biased knowledge distillation framework tailored for binary classification tasks.For the pre-trained teacher model,biases in the soft labels are mitigated through knowledge infusion and label de-biasing techniques.Based on this,a de-biased distillation loss is introduced,allowing the de-biased labels to replace the soft labels as the fitting target for the student model.This approach enables the student model to learn from the corrected model information,achieving high-performance deployment on lightweight student models.Experiments conducted on multiple real-world datasets demonstrate that deep learning models compressed under the de-biased knowledge distillation framework significantly outperform traditional response-based and feature-based knowledge distillation models across various evaluation metrics,highlighting the effectiveness and superiority of the de-biased knowledge distillation framework in model compression.
基金supported in part by the National Natural Science Foundation of China (62073271)the Natural Science Foundation for Distinguished Young Scholars of the Fujian Province of China (2023J06010)the Fundamental Research Funds for the Central Universities of China(20720220076)。
文摘Unmanned aerial vehicles(UAVs) have gained significant attention in practical applications, especially the low-altitude aerial(LAA) object detection imposes stringent requirements on recognition accuracy and computational resources. In this paper, the LAA images-oriented tensor decomposition and knowledge distillation-based network(TDKD-Net) is proposed,where the TT-format TD(tensor decomposition) and equalweighted response-based KD(knowledge distillation) methods are designed to minimize redundant parameters while ensuring comparable performance. Moreover, some robust network structures are developed, including the small object detection head and the dual-domain attention mechanism, which enable the model to leverage the learned knowledge from small-scale targets and selectively focus on salient features. Considering the imbalance of bounding box regression samples and the inaccuracy of regression geometric factors, the focal and efficient IoU(intersection of union) loss with optimal transport assignment(F-EIoU-OTA)mechanism is proposed to improve the detection accuracy. The proposed TDKD-Net is comprehensively evaluated through extensive experiments, and the results have demonstrated the effectiveness and superiority of the developed methods in comparison to other advanced detection algorithms, which also present high generalization and strong robustness. As a resource-efficient precise network, the complex detection of small and occluded LAA objects is also well addressed by TDKD-Net, which provides useful insights on handling imbalanced issues and realizing domain adaptation.
基金supported in part by the Guangdong Natu-ral Science Foundation(No.2022A1515011396)in part by the National Key R and D Program of China(No.2021ZD0111502)in part by the Science Research Startup Foundation of Shantou University(No.NTF20021)。
文摘Strabismus significantly impacts human health as a prevalent ophthalmic condition.Early detection of strabismus is crucial for effective treatment and prognosis.Traditional deep learning models for strabismus detection often fail to estimate prediction certainty precisely.This paper employed a Bayesian deep learning algorithm with knowledge distillation,improving the model's performance and uncertainty estimation ability.Trained on 6807 images from two tertiary hospitals,the model showed significantly higher diagnostic accuracy than traditional deep-learning models.Experimental results revealed that knowledge distillation enhanced the Bayesian model’s performance and uncertainty estimation ability.These findings underscore the combined benefits of using Bayesian deep learning algorithms and knowledge distillation,which improve the reliability and accuracy of strabismus diagnostic predictions.
文摘In recent years,anomaly detection has attracted much attention in industrial production.As traditional anomaly detection methods usually rely on direct comparison of samples,they often ignore the intrinsic relationship between samples,resulting in poor accuracy in recognizing anomalous samples.To address this problem,a knowledge distillation anomaly detection method based on feature reconstruction was proposed in this study.Knowledge distillation was performed after inverting the structure of the teacher-student network to avoid the teacher-student network sharing the same inputs and similar structure.Representability was improved by using feature splicing to unify features at different levels,and the merged features were processed and reconstructed using an improved Transformer.The experimental results show that the proposed method achieves better performance on the MVTec dataset,verifying its effectiveness and feasibility in anomaly detection tasks.This study provides a new idea to improve the accuracy and efficiency of anomaly detection.
文摘In this study,the effect of yeast strains(X16 and RMS2),fruit seed,and pectinase on the quality of apple distilled spirits were investigated with apple as material.The results showed that pectinase shortened the fermentation period.The strain X16,fruit seed remaining,and pectinase addition groups had higher yields of crude distilled spirits than the strain RMS2,fruit seed removal,and without pectinase groups,respectively.Regarding the first-grade distilled spirits quality,the X16 group had higher content of total acids,total esters,higher alcohols,and alcohol content than the RMS2 group;the group with fruit seeds had higher total acids but lower alcohol content and total esters than the group with fruit seed removal;the group with pectinase addition had higher total acids and alcohol content than the group without pectinase addition.Regarding the second-grade distilled spirits quality,the X16 group had higher total acids,total esters,and alcohol content than the RMS2 group;the group without fruit seeds had higher alcohol content than the group with fruit seeds;the group with pectinase addition showcased higher total acids but lower alcohol content than the group without pectinase addition.In summary,yeast strains and pectinase affected the quality of apple distilled spirits,and strain X16 was more suitable for brewing apple distilled spirits.Pectinase affected the quality of apple distilled spirits by affecting fermentation rate and temperature.
文摘The experimental and analytical investigation was conducted on a solar-powered single-effect distillation (SED). The evaporator was designed to be an electrical evaporator as opposed to the steam evaporator that existed previously. Using sun-tracking solar panels, the electrical evaporator in the designed distillation unit was powered by solar energy. Approximately 20 kWh was utilized by the small-scale distillation apparatus. This type of design is mobile, so remote areas and countries with fragile economies can utilize it on a small or large scale. Utilizing the principles of energy and mass conservation, the amount of distillate water and power required for a single unit was determined, at the low salinity (2200 PPM) with fixed boiling point temperature (Tb = 75˚C), the unit performance is approx. 98.4%. The experimental results and those derived from a mathematical model were compared, and both showed strong accord. Using engineering equation solver (EES) software, a computer program was developed for this research scenario.
文摘This paper delves into the comparative study of tray and packed column pressure swing distillation systems, focusing on the separation of a ternary mixture containing ethanol, tetrahydrofuran (THF), and water. The study particularly emphasizes the production of 99.5 w/w% tetrahydrofuran from the downstream product of 1,4-butanediol synthesis via diethyl maleate. Pro/II simulation software is utilized to explore various system configurations, including sieve trays, valve trays, and packed columns. Material and energy balances are performed to ascertain stream compositions and energy demands. The investigation encompasses the effects of column operating pressure on condenser and reboiler temperatures, as well as the implications of utility streams. A rigorous distillation model is employed to compare valve tray, sieve tray, and random packing (utilizing Norton Super Intalox) column designs by varying the number of trays, reflux ratio, and second distillation column pressure. Heat exchangers are integrated into the model, and their areas and utility flow rates are computed and integrated into the economic assessment. Economic analysis, guided by Net Present Value (NPV) calculations over a 20-year span, drives the selection of the most cost-effective design. Results demonstrate that while all designs are energy-efficient, the packed column system emerges as the most economical choice, offering a comprehensive framework for the separation process. Furthermore, optimal design configurations and operating conditions for both tray and packed column systems are outlined, providing valuable insights for industrial applications.
基金supported by the National Natural Science Foundation of China(22178030,21878025,22078026)。
文摘Extractive distillation(ED)and solvent-assisted pressure-swing distillation(SA-PSD)are both special distillation processes that perform good at separating pressure-insensitive azeotropes.However,few reported studies have compared the performance of the two processes.In this paper,ED processes with N-methylpyrrolidone(NMP)and dimethlac-etamide(DMCA)as entrainer,SA-PSD process with isopropyl-alcohol(IPA)as solvent and SA-PSD process with partial heat integration(PHI-PSD)are proposed to achieve high purity separation of a mixture of cyclohexane/2-butanol system.The optimal operating conditions of the processes are obtained after optimizing with NSGA-Ⅱ algorithm when total annual cost(TAC)and the entropy production of process are set as objectives.The optimal results show that the optimal PHI-PSD process has lower TAC by 28.7% and the lower entropy production by 39.5% than the optimal SA-PSD process while the ED process with NMP as entrainer has lower TAC by 50.9% and the lower entropy production by 56.1% than the optimal SA-PSD process.The optimal results show that the ED process with NMP as entrainer has the best economic and thermodynamic efficiency among the four proposed processes in this paper.
基金supported by the National Natural Science Foundation of China (21776145 and 21808117)。
文摘This wok proposed the extraction distillation coupled pervaporation(ED+PV) technology process using two different solvents to separate isopropanol(IPA) and diisopropyl ether(DIPE) from DIPE/IPA/H_(2)O ternary heterogeneous azeotropes in industrial wastewater from the synthesis of isopropanol in this study.Based on strict design specifications, simulation and sequential iteration methods are used for process design and optimization. Compared to the ethylene glycol(EG)-EG+H_(2)O process and the 1,3-propanediol(PDO)-IPA+H_(2)O process, the total annual cost(TAC) of the EG-IPA+H_(2)O process decreased by 20.76% and 7.86%(PDO). Compared to the EG-EG+H_(2)O process, the TAC of the PDO-IPA+H_(2)O process reduced 14%, but the global warming potential(GWP) and human toxicity of the PDO-IPA+H_(2)O process increased 11.3% and 4.07% respectively. Compared to the PDO-IPA+H_(2)O process, the EG-IPA+H_(2)O process saves 7.86%(TAC), 9.78%(GWP) and 9.85%(human toxicity). The ED+PV process with EG is superior to PDO in factors of TAC, energy consumption, human toxicity and environment. The EG-IPA+H_(2)O process changed the separation order of the products of the multi-azeotropic system, reduced the cost and energy conservation of the system, and enhanced the environmental protection evaluation of the process, is the best process through life cycle assessment for analyzing the economy, energy conservation, environmental assessment and human toxicity, designing cleaner products, controlling waste discharge, and promoting the chemical purification industry. This work provides a new process design and optimized separation ideas, will have a good guiding significance for the research and application separation of multi-azeotropic mixture with mixed solvents in organic wastewater from the cleaner chemical production, has been up to standard wastewater discharge process, and realized the development goal of carbon peak and carbon neutrality in the sustainable development of chemical clean industry.
基金financial support provided by the National Natural Science Foundation of China (22178030, 21878025, and 22078026)。
文摘The conventional distillation is hard to accomplish the separation of acetonitrile/ethyl acetate/n-hexane mixture. Herein, a heterogeneous azeotropic distillation(HAD) without adding entrainer is proposed to separate ternary mixture. The proposed scheme is optimized via the simulated annealing algorithm and minimum total annual cost(TAC) is used as objective functions. To minimize energy consumption,heat pump is added on the basis of optimal heterogeneous azeotropic distillation and heat integration technology is used to further improve the energy recovery. The TAC, gas emission, energy consumption and exergy destruction are used to discuss the economy and environmental protection of processes.Among all the processes, the heat pump with higher preheating temperature(HPT) assisted HAD process by combining with heat integration(HAD-HPT-HI) has best performances on economic, environment,energy and exergy. Compared with conventional HAD process, the HAD-HPT-HI achieves the reductions of 52.17%, 68.86%, 65.87% and 65.46% on TAC, total energy consumption, gas emissions and exergy destruction, respectively.
基金supported by the Korea Evaluation Institute of Industrial Technology funded by the Korean Ministry of Industry in Korea (Project No.:20000970, 20–9805)Basic Research Project (22–3803) of the Korea Institute of Geoscience and Mineral Resources (KIGAM) funded by the Ministry of Science and ICT of Korea。
文摘A green and effective electrolytic process was developed to produce high-purity Mg metal using primary and secondary resources containing Mg O as a feedstock. The electrolysis of various Mg O resources was conducted using a Cu cathode in MgF2– LiF – KCl molten salt at 1043 K by applying an average current of 1.44 A for 12.5 h. The electrolysis of calcined North Korean magnesite and seawater Mg O clinker yielded Mg alloys of MgCu2and(Cu) phases with current efficiencies of 89.6–92.4%. The electrolysis of oxidized Mg O-C refractory brick, aged ferronickel slag, and ferronickel slag yielded Mg alloys of MgCu2and(Cu) phases with current efficiencies of 59.3–92.3%. The vacuum distillation of Mg alloys obtained was conducted at 1300 K for 10 h to produce high-purity Mg metal. After vacuum distillation, Mg metal with a purity of above 99.994% was obtained. Therefore, this study demonstrates the feasibility of the production of high-purity Mg metal from various Mg O resources using a novel electrolytic process with a Cu cathode, followed by vacuum distillation.
