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
Concentrating sulfuric acid solution by vacuum membrane distillation with flat PEFE membrane is explored. The effects of sulfuric acid concentration, temperature of the feed, the vacuum degree of the vacuum side on th...Concentrating sulfuric acid solution by vacuum membrane distillation with flat PEFE membrane is explored. The effects of sulfuric acid concentration, temperature of the feed, the vacuum degree of the vacuum side on the flux of membrane distillation and the separation efficiency of acid are investigated. The results illustrate that the flux of the membrane distillation increases with the rise of feed temperature and the vacuum degree of the vacuum side, but it decreases with the rise of the sulfuric acid concentration of the feed. The separation efficiency of acid is correlated with the flux of membrane distillation; the separation efficiency of the acid can amount to 100% in the process, when operative conditions are properly controlled. It can also been obtained from the experiment that, compared with other methods of membrane distillation, the vacuum membrane distillation can obtain greater distillation flux.展开更多
Methyl isobutyl ketone(MIBK) is widely used as extraction agent in hydrometallurgy. As it has a definite solubility in water, so when using MIBK as extraction agent, there will be MIBK in stripping solutions inevitabl...Methyl isobutyl ketone(MIBK) is widely used as extraction agent in hydrometallurgy. As it has a definite solubility in water, so when using MIBK as extraction agent, there will be MIBK in stripping solutions inevitably, which not only pollutes working conditions, but also affects the quality of ultimate product. In order to remove MIBK from aqueous solutions, the means of flat vacuum membrane distillation(VMD) is studied in the paper. The area of the membrane used in the study is 0.02 m 2, the initial volume of feed is 2 L, each experiment was conducted over a time period of 60 120 min. The influences of the factors such as temperature(34.8 55.0 ℃); pressure in the permeate side(10.67 14.67 kPa) and feed flow rate(27.8 69.4 mL/s) were experimentally studied. Increasing the temperature or reducing the pressure in the permeate side results in a faster removal of MIBK; however there is a decrease in removal factor β , increasing the feed flow rate results in a faster removal of MIBK and an increase of removal factor β , especially in the range of lower flow rate. The study indicates that the aim of MIBK removal and recycle from dilute aqueous solutions can be achieved by VMD.展开更多
High purity （99.999% or 5N, mass fraction） indium （In） was obtained through vacuum distillation using a 2N （99%） In as input material under a dynamic vacuum of 5 Pa. The glow discharge mass spectrometry （GDMS）...High purity （99.999% or 5N, mass fraction） indium （In） was obtained through vacuum distillation using a 2N （99%） In as input material under a dynamic vacuum of 5 Pa. The glow discharge mass spectrometry （GDMS） was applied for the analysis of input material and the distilled indium. The results indicate that high-volatile impurities namely Cd, Zn, T1 and Pb can be removed from the indium matrix at the low fraction stage of 1 223 K for 120 min; Low-volatile impurities such as Fe, Ni, Cu, Sn can be reduced at the high fraction stage of 1 323 K for 120 min. The separation coefficient ,8i and activity coefficient Yi of impurities are calculated according to the experiments to fill the inadequate data of the thermodynamics.展开更多
The feasibility of separation of lead anode slime with low silver by vacuum distillation was analyzed theoretically. The volatilization rates and mass fractions of elements, influenced by distillation temperature, hea...The feasibility of separation of lead anode slime with low silver by vacuum distillation was analyzed theoretically. The volatilization rates and mass fractions of elements, influenced by distillation temperature, heat preservation time and material thickness, were investigated under laboratory conditions. The experimental results indicate that almost all of lead and bismuth can be separated from silver-contained multicomponent alloy at 1 223 K for 45 min when the chamber pressure maintains at 10-25 Pa. Silver can be easily enriched in the residue and its mass fraction increases from 3.6% to 27.8% when the distillation temperature is between 1 133 K and 1 373 K. Due to the forming ofintermetallic compounds Cu2Sb, Cul0Sb3 and Ag3Sb, the antimony could not be evaporated completely during the vacuum distillation. EDS analysis indicates that the condensate has a columnar crystal structure.展开更多
concentrating sulfuric acid by membrane distillation in a flat sheet direct contact membrane distillation device with a self made metal PTFE composite membrane has been studied. The effect of sulfuric acid concentrati...concentrating sulfuric acid by membrane distillation in a flat sheet direct contact membrane distillation device with a self made metal PTFE composite membrane has been studied. The effect of sulfuric acid concentration of feed, the flowrate and liquor temperature in high temperature side and low temperature side on the flux and separation efficiency is investigated. The stability of this composite membrane is also inspected. The experimental results show that the metal PTFE composite membrane can be used in membrane distillation, and its properties are very stable. It is feasible to concentrate sulfuric acid by membrane distillation with this membrane.展开更多
In this paper,a novel method of ultra-lightweight convolution neural network(CNN)design based on neural architecture search(NAS)and knowledge distillation(KD)is proposed.It can realize the automatic construction of th...In this paper,a novel method of ultra-lightweight convolution neural network(CNN)design based on neural architecture search(NAS)and knowledge distillation(KD)is proposed.It can realize the automatic construction of the space target inverse synthetic aperture radar(ISAR)image recognition model with ultra-lightweight and high accuracy.This method introduces the NAS method into the radar image recognition for the first time,which solves the time-consuming and labor-consuming problems in the artificial design of the space target ISAR image automatic recognition model(STIIARM).On this basis,the NAS model’s knowledge is transferred to the student model with lower computational complexity by the flow of the solution procedure(FSP)distillation method.Thus,the decline of recognition accuracy caused by the direct compression of model structural parameters can be effectively avoided,and the ultralightweight STIIARM can be obtained.In the method,the Inverted Linear Bottleneck(ILB)and Inverted Residual Block(IRB)are firstly taken as each block’s basic structure in CNN.And the expansion ratio,output filter size,number of IRBs,and convolution kernel size are set as the search parameters to construct a hierarchical decomposition search space.Then,the recognition accuracy and computational complexity are taken as the objective function and constraint conditions,respectively,and the global optimization model of the CNN architecture search is established.Next,the simulated annealing(SA)algorithm is used as the search strategy to search out the lightweight and high accuracy STIIARM directly.After that,based on the three principles of similar block structure,the same corresponding channel number,and the minimum computational complexity,the more lightweight student model is designed,and the FSP matrix pairing between the NAS model and student model is completed.Finally,by minimizing the loss between the FSP matrix pairs of the NAS model and student model,the student model’s weight adjustment is completed.Thus the ultra-lightweight and high accuracy STIIARM is obtained.The proposed method’s effectiveness is verified by the simulation experiments on the ISAR image dataset of five types of space targets.展开更多
The volatile and semi-volatile components in tobacco flavor additives were extracted by both simultaneous distillation extraction and solid-phase micro-extraction. Extraction conditions for solid-phase micro-extractio...The volatile and semi-volatile components in tobacco flavor additives were extracted by both simultaneous distillation extraction and solid-phase micro-extraction. Extraction conditions for solid-phase micro-extraction were optimized with information theory. Then, detection were accomplished by gas chromatography-mass spectrometry. Characteristic of each method was compared. Qualitative analysis and quantitative analysis of 6# tobacco flavor sample were accomplished through both simultaneous distillation extraction and solid-phase micro-extraction. The experimental results show that solid-phase micro-extraction method is the first choice for qualitative analysis and simultaneous distillation extraction is another good selection for quantitative analysis. By means of simultaneous distillation extraction, 20 components are identified, accounting for 92.77% of the total peak areas. Through solid-phase micro-extraction, there are 17 components identified accounting for 91.49% of the total peak areas. The main aromatic components in 6# tobacco flavor sample are propanoic acid, 2-hydroxy-, ethyl ester, menthol and menthyl acetate. The presented method has been successfully used for quality control of tobacco flavor.展开更多
The activity of components of Sn-Zn binary alloy system was predicted based on the molecular interaction volume model （MIVM）. The calculated values are in good agreement with available experimental data of activitie...The activity of components of Sn-Zn binary alloy system was predicted based on the molecular interaction volume model （MIVM）. The calculated values are in good agreement with available experimental data of activities, which indicates that this model is of stability and reliability because the MIVM has a good physical basis. The vapor-liquid phase equilibrium of Sn-Zn alloy system in vacuum distillation was calculated as a function of the activity coefficient. The results show that the content of Sn in vapor phase is 4.2x 10-7 （mass fraction） while in liquid phase it is 90% （mass fraction） at 1 073 K, and the content of Sn in vapor phase increases with increasing the melt temperature and content of Sn in liquid phase. Vacuum distillation experiments were carried out on Sn-Zn alloy for the proper interpretation of the results of the MIVM in the temperature range of 973-1 273 K under pressures of 15-200 Pa. The experimental results show that the content of Sn in vapor phase is 5x 10 6 （mass fraction） while in liquid phase it is 90% （mass fraction） under the operational condition of 1 073 K, 100 rain and 15 Pa. The experimental results are in good agreement with the predicted values of the MIVM for Zn-Sn binary alloy system.展开更多
Soybean leaf morphology is one of the most important morphological and biological characteristics of soybean.The germplasm gene differences of soybeans can lead to different phenotypic traits,among which soybean leaf ...Soybean leaf morphology is one of the most important morphological and biological characteristics of soybean.The germplasm gene differences of soybeans can lead to different phenotypic traits,among which soybean leaf morphology is an important parameter that directly reflects the difference in soybean germplasm.To realize the morphological classification of soybean leaves,a method was proposed based on deep learning to automatically detect soybean leaves and classify leaf morphology.The morphology of soybean leaves included lanceolate,oval,ellipse and round.First,an image collection platform was designed to collect images of soybean leaves.Then,the feature pyramid networks–single shot multibox detector(FPN-SSD)model was proposed to detect the top leaflets of soybean leaves on the collected images.Finally,a classification model based on knowledge distillation was proposed to classify different morphologies of soybean leaves.The obtained results indicated an overall classification accuracy of 0.956 over a private dataset of 3200 soybean leaf images,and the accuracy of classification for each morphology was 1.00,0.97,0.93 and 0.94.The results showed that this method could effectively classify soybean leaf morphology and had great application potential in analyzing other phenotypic traits of soybean.展开更多
An active set truncated-Newton algorithm (ASTNA) is proposed to solve the large-scale bound constrained sub-problems. The global convergence of the algorithm is obtained and two groups of numerical experiments are mad...An active set truncated-Newton algorithm (ASTNA) is proposed to solve the large-scale bound constrained sub-problems. The global convergence of the algorithm is obtained and two groups of numerical experiments are made for the various large-scale problems of varying size. The comparison results between ASTNA and the subspace limited memory quasi-Newton algorithm and between the modified augmented Lagrange multiplier methods combined with ASTNA and the modified barrier function method show the stability and effectiveness of ASTNA for simultaneous optimization of distillation column.展开更多
Base on industrial research and experience, the process of methanol distillation is analyzed,and above all, a new concept of high pressure flowsheet and low pressure flowsheet is defined. The new configuration helps t...Base on industrial research and experience, the process of methanol distillation is analyzed,and above all, a new concept of high pressure flowsheet and low pressure flowsheet is defined. The new configuration helps to handle problems encountered in many factories in China. The inter influence between process and column internal pattern is also pointed out. Recommendation of new column internal designs is given. Finally, industrial examples tell the how the new concept works and the possibility of combining process to give more opens to solve engineering problems.展开更多
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Concentrating sulfuric acid solution by vacuum membrane distillation with flat PEFE membrane is explored. The effects of sulfuric acid concentration, temperature of the feed, the vacuum degree of the vacuum side on the flux of membrane distillation and the separation efficiency of acid are investigated. The results illustrate that the flux of the membrane distillation increases with the rise of feed temperature and the vacuum degree of the vacuum side, but it decreases with the rise of the sulfuric acid concentration of the feed. The separation efficiency of acid is correlated with the flux of membrane distillation; the separation efficiency of the acid can amount to 100% in the process, when operative conditions are properly controlled. It can also been obtained from the experiment that, compared with other methods of membrane distillation, the vacuum membrane distillation can obtain greater distillation flux.
文摘Methyl isobutyl ketone(MIBK) is widely used as extraction agent in hydrometallurgy. As it has a definite solubility in water, so when using MIBK as extraction agent, there will be MIBK in stripping solutions inevitably, which not only pollutes working conditions, but also affects the quality of ultimate product. In order to remove MIBK from aqueous solutions, the means of flat vacuum membrane distillation(VMD) is studied in the paper. The area of the membrane used in the study is 0.02 m 2, the initial volume of feed is 2 L, each experiment was conducted over a time period of 60 120 min. The influences of the factors such as temperature(34.8 55.0 ℃); pressure in the permeate side(10.67 14.67 kPa) and feed flow rate(27.8 69.4 mL/s) were experimentally studied. Increasing the temperature or reducing the pressure in the permeate side results in a faster removal of MIBK; however there is a decrease in removal factor β , increasing the feed flow rate results in a faster removal of MIBK and an increase of removal factor β , especially in the range of lower flow rate. The study indicates that the aim of MIBK removal and recycle from dilute aqueous solutions can be achieved by VMD.
基金Project(2009AA003) supported by Science and Technology Innovation Plan of Yunnan Province, China
文摘High purity （99.999% or 5N, mass fraction） indium （In） was obtained through vacuum distillation using a 2N （99%） In as input material under a dynamic vacuum of 5 Pa. The glow discharge mass spectrometry （GDMS） was applied for the analysis of input material and the distilled indium. The results indicate that high-volatile impurities namely Cd, Zn, T1 and Pb can be removed from the indium matrix at the low fraction stage of 1 223 K for 120 min; Low-volatile impurities such as Fe, Ni, Cu, Sn can be reduced at the high fraction stage of 1 323 K for 120 min. The separation coefficient ,8i and activity coefficient Yi of impurities are calculated according to the experiments to fill the inadequate data of the thermodynamics.
基金Project(2010CI009) supported by Science Foundation of Yunnan Province,ChinaProjects(U1202271) supported by the National Natural Science Foundation of China
文摘The feasibility of separation of lead anode slime with low silver by vacuum distillation was analyzed theoretically. The volatilization rates and mass fractions of elements, influenced by distillation temperature, heat preservation time and material thickness, were investigated under laboratory conditions. The experimental results indicate that almost all of lead and bismuth can be separated from silver-contained multicomponent alloy at 1 223 K for 45 min when the chamber pressure maintains at 10-25 Pa. Silver can be easily enriched in the residue and its mass fraction increases from 3.6% to 27.8% when the distillation temperature is between 1 133 K and 1 373 K. Due to the forming ofintermetallic compounds Cu2Sb, Cul0Sb3 and Ag3Sb, the antimony could not be evaporated completely during the vacuum distillation. EDS analysis indicates that the condensate has a columnar crystal structure.
文摘concentrating sulfuric acid by membrane distillation in a flat sheet direct contact membrane distillation device with a self made metal PTFE composite membrane has been studied. The effect of sulfuric acid concentration of feed, the flowrate and liquor temperature in high temperature side and low temperature side on the flux and separation efficiency is investigated. The stability of this composite membrane is also inspected. The experimental results show that the metal PTFE composite membrane can be used in membrane distillation, and its properties are very stable. It is feasible to concentrate sulfuric acid by membrane distillation with this membrane.
文摘In this paper,a novel method of ultra-lightweight convolution neural network(CNN)design based on neural architecture search(NAS)and knowledge distillation(KD)is proposed.It can realize the automatic construction of the space target inverse synthetic aperture radar(ISAR)image recognition model with ultra-lightweight and high accuracy.This method introduces the NAS method into the radar image recognition for the first time,which solves the time-consuming and labor-consuming problems in the artificial design of the space target ISAR image automatic recognition model(STIIARM).On this basis,the NAS model’s knowledge is transferred to the student model with lower computational complexity by the flow of the solution procedure(FSP)distillation method.Thus,the decline of recognition accuracy caused by the direct compression of model structural parameters can be effectively avoided,and the ultralightweight STIIARM can be obtained.In the method,the Inverted Linear Bottleneck(ILB)and Inverted Residual Block(IRB)are firstly taken as each block’s basic structure in CNN.And the expansion ratio,output filter size,number of IRBs,and convolution kernel size are set as the search parameters to construct a hierarchical decomposition search space.Then,the recognition accuracy and computational complexity are taken as the objective function and constraint conditions,respectively,and the global optimization model of the CNN architecture search is established.Next,the simulated annealing(SA)algorithm is used as the search strategy to search out the lightweight and high accuracy STIIARM directly.After that,based on the three principles of similar block structure,the same corresponding channel number,and the minimum computational complexity,the more lightweight student model is designed,and the FSP matrix pairing between the NAS model and student model is completed.Finally,by minimizing the loss between the FSP matrix pairs of the NAS model and student model,the student model’s weight adjustment is completed.Thus the ultra-lightweight and high accuracy STIIARM is obtained.The proposed method’s effectiveness is verified by the simulation experiments on the ISAR image dataset of five types of space targets.
文摘The volatile and semi-volatile components in tobacco flavor additives were extracted by both simultaneous distillation extraction and solid-phase micro-extraction. Extraction conditions for solid-phase micro-extraction were optimized with information theory. Then, detection were accomplished by gas chromatography-mass spectrometry. Characteristic of each method was compared. Qualitative analysis and quantitative analysis of 6# tobacco flavor sample were accomplished through both simultaneous distillation extraction and solid-phase micro-extraction. The experimental results show that solid-phase micro-extraction method is the first choice for qualitative analysis and simultaneous distillation extraction is another good selection for quantitative analysis. By means of simultaneous distillation extraction, 20 components are identified, accounting for 92.77% of the total peak areas. Through solid-phase micro-extraction, there are 17 components identified accounting for 91.49% of the total peak areas. The main aromatic components in 6# tobacco flavor sample are propanoic acid, 2-hydroxy-, ethyl ester, menthol and menthyl acetate. The presented method has been successfully used for quality control of tobacco flavor.
基金Project(2012CB722803) supported by the Key Project of National Basic Research and Development Program of ChinaProject(2011FA008)supported by the Key Project of Science and Technology Program of Yunnan Province,China
文摘The activity of components of Sn-Zn binary alloy system was predicted based on the molecular interaction volume model （MIVM）. The calculated values are in good agreement with available experimental data of activities, which indicates that this model is of stability and reliability because the MIVM has a good physical basis. The vapor-liquid phase equilibrium of Sn-Zn alloy system in vacuum distillation was calculated as a function of the activity coefficient. The results show that the content of Sn in vapor phase is 4.2x 10-7 （mass fraction） while in liquid phase it is 90% （mass fraction） at 1 073 K, and the content of Sn in vapor phase increases with increasing the melt temperature and content of Sn in liquid phase. Vacuum distillation experiments were carried out on Sn-Zn alloy for the proper interpretation of the results of the MIVM in the temperature range of 973-1 273 K under pressures of 15-200 Pa. The experimental results show that the content of Sn in vapor phase is 5x 10 6 （mass fraction） while in liquid phase it is 90% （mass fraction） under the operational condition of 1 073 K, 100 rain and 15 Pa. The experimental results are in good agreement with the predicted values of the MIVM for Zn-Sn binary alloy system.
基金Supported by Heilongjiang Province Philosophy and Social Science Research Planning Project(17TQB059)。
文摘Soybean leaf morphology is one of the most important morphological and biological characteristics of soybean.The germplasm gene differences of soybeans can lead to different phenotypic traits,among which soybean leaf morphology is an important parameter that directly reflects the difference in soybean germplasm.To realize the morphological classification of soybean leaves,a method was proposed based on deep learning to automatically detect soybean leaves and classify leaf morphology.The morphology of soybean leaves included lanceolate,oval,ellipse and round.First,an image collection platform was designed to collect images of soybean leaves.Then,the feature pyramid networks–single shot multibox detector(FPN-SSD)model was proposed to detect the top leaflets of soybean leaves on the collected images.Finally,a classification model based on knowledge distillation was proposed to classify different morphologies of soybean leaves.The obtained results indicated an overall classification accuracy of 0.956 over a private dataset of 3200 soybean leaf images,and the accuracy of classification for each morphology was 1.00,0.97,0.93 and 0.94.The results showed that this method could effectively classify soybean leaf morphology and had great application potential in analyzing other phenotypic traits of soybean.
基金Project (2002CB312200) supported by the National Key Basic Research and Development Program of China Project(03JJY3109) supported by the Natural Science Foundation of Hunan Province
文摘An active set truncated-Newton algorithm (ASTNA) is proposed to solve the large-scale bound constrained sub-problems. The global convergence of the algorithm is obtained and two groups of numerical experiments are made for the various large-scale problems of varying size. The comparison results between ASTNA and the subspace limited memory quasi-Newton algorithm and between the modified augmented Lagrange multiplier methods combined with ASTNA and the modified barrier function method show the stability and effectiveness of ASTNA for simultaneous optimization of distillation column.
文摘Base on industrial research and experience, the process of methanol distillation is analyzed,and above all, a new concept of high pressure flowsheet and low pressure flowsheet is defined. The new configuration helps to handle problems encountered in many factories in China. The inter influence between process and column internal pattern is also pointed out. Recommendation of new column internal designs is given. Finally, industrial examples tell the how the new concept works and the possibility of combining process to give more opens to solve engineering problems.
