Distillation Technology Research Articles

Optimizing Federated Learning Efficiency: A Comparative Analysis of Model Compression Techniques for Communication Reduction

Abstract. The rapidly growing number of connected devices continuously produces massive amounts of heterogeneous data, posing challenges to data privacy and security when leveraging them to train the high-quality big model. Federated learning, which enables numerous users to train a global model without cooperatively exchanging data, has emerged as a viable alternative. Though achieving significant progress, existing federated learning methods still struggle with large communication volumes, especially when the local devices only have limited computing and communication capabilities. To alleviate the efficiency issue, this paper compares the effect of three compression methods in promoting the training of federated learning models, including pruning, quantization, and knowledge distillation. The findings reveal that these methods reduce resource consumption while maintaining high model performance. Combining the pruning, quantization, and knowledge distillation technology through sequential application and parameter aggregation helps balance the model size and performance. Our cascading lightweight strategy, which preserves each method's unique edge while promoting deeper collaboration between them, has been shown beneficial through extensive testing.

Feasible synthesis of anti-fouling amphiphobic composite membrane by surface modification with SiO2 nanoparticles for effective membrane distillation

Membrane distillation (MD) technology has shown unique advantages in treating saline organic wastewaters, however, improving membrane materials of anti-wetting/fouling properties is still agap for the large-scale application. In this study, a non-fluorinated composite membrane showing both good amphiphobic properties and high flux was synthesized by feasible electrostatic attraction. Characterizations indicated that SiO2nanoparticles were successfully loaded on the PVDF membrane (SiO2/PVDF) surface to form a multi-level rough amphiphobic structure, with contact angles of water and ethylene glycol as 139° and 128°, respectively. In MD tests, the SiO2/PVDF membrane exhibited a stable flux of 15 LMH with 3.5 wt% NaCl and 19.5 LMH with 50 mg/L kerosene. Notably, after a 24-hour continuous test, the flux remained above 9 LMH for NaCl and 11 LMH for kerosene, with salt rejection rates exceeding 98 %, indicating excellent anti-fouling characteristics. This study is expected to provide a feasible membrane modification for MD process in treating high saline organic wastewater.

Loss tolerance quantum key distribution based on the signal extraction model and advantage distillation technology

Compared with traditional networks, quantum key distribution (QKD) offers the ultimate resources, allowing two remote users to share secret symmetric keys regardless of the capabilities of eavesdroppers. However, the widespread application of commercial QKD is still challenging due to the low photon detection efficiency and the extremely high transmission loss. Here we demonstrate a fully commercial phase-encoding QKD system using a signal extraction model and advantage distillation technology to suppress detector noise and perform real-time pre-error correction. 1.89 × 10−10 in the asymptotic case and 7.43 × 10−12 in the nonasymptotic case secret key bits per pulse are achieved with a total loss of 70.05 dB. This method not only increases the transmission loss tolerance but also provides a more realistic deployment of quantum communication.

Deconstructing the ‘Gandhāra still’: a new challenge to the accepted trajectory of early distillation technology

Deconstructing the ‘Gandhāra still’: a new challenge to the accepted trajectory of early distillation technology

Construction of PTFE/PI-PI/PANI-PA composite nanofibrous membrane for efficient photothermal membrane distillation and VOCs interception

Water-soluble volatile organic compounds (VOCs) are one of the difficult substances in seawater and industrial wastewater treatment. Photothermal membrane distillation (PMD) technology has great advantages in solving the problems of low energy conversion efficiency and serious pollution of traditional membrane distillation (MD) technology, but it still faces the challenge that VOCs cannot be effectively removed.To solve this problem, we prepared photothermal Polytetrafluoroethylene/Polyimide-Polyimide/Polyaniline-Polyamide (PTFE/PI-PI/PANI-PA) composite membranes with VOCs interception. The composite membrane was fabricated by applying the previously studied PTFE/PI-PI/PANI/PANI membrane by interfacially polymerising an ultrathin PA layer with controllable pore sizes. Among them, the PTFE/PI-PI/PANI base membrane in the composite membrane not only provides strong support and durability as a backbone structure, but the PANI photothermal layer also provides photothermal conversion for effective PMD operation. In addition, the ultra-thin PA layer acts as a molecular sieve separator to effectively intercept VOCs in the feed solution. When the feed solution was 35 g⋅L−1 NaCl, the average permeate flux of the #M-PA-0.1 membrane was 1.44 ± 0.02 L⋅m−2⋅h−1 with a salt interception rate of 99.99 % or more after 80 h of PMD testing. Meanwhile, when 50 mg⋅L−1 of VOCs was added to the feed solution, the VOCs retention rate >99.35 %, and the concentration of VOCs on the permeate side was much lower than the corresponding concentration in the drinking water standards recommended by the World Health Organisation (WHO) and the US Environmental Protection Agency (EPA). This study effectively combines the molecular sieve effect with the solar-powered evaporation process, which provides a research basis for the preparation of high-performance PMD membranes that can effectively remove volatile organic compounds.

A low-carbon dual-functional evaporator with photothermal and catalytic properties for enhanced VOCs removal during water evaporation

Solar interface distillation technologies are widely used for seawater desalination, wastewater treatment, and clean water production. However, during evaporation, volatile organic compounds (VOCs) in the water often evaporate coincidently with the water vapor, presenting a major pollutant challenge. Combining solar evaporation with advanced oxidation processes (AOP) is a potential solution for simultaneous water purification and VOC degradation. Herein, a dual-functional sponge photothermal evaporator (MBC-SA/MS) was designed to assess the efficacy of this approach. A simple porous sponge embedded with a composite catalyst, comprising MoO3−x, BiOCl, and CNT, was able to demonstrate simultaneous photothermal evaporation and catalytic degradation. MBC-SA/MS exhibited exceptional photothermal conversion efficiency and robust catalytic activity, achieving an evaporation rate of 2.1 kg m−2 h−1 under 1.0 sun intensity. Effective mitigating of phenol emissions was observed through persulfate activation resulting in 90 % phenol degradation. Life cycle assessment (LCA) confirmed this evaporator system had low environmental impact and was sustainable. This research presents a new compelling strategy for tackling VOC pollution and simultaneously enhancing water purification via solar interfacial evaporation.

The potential of integrating solar-powered membrane distillation with a humidification-dehumidification system to recover potable water from textile wastewater

Textile production is energy- and water-intensive, and membrane distillation (MD) has shown potential for reclaiming potable water from textile wastewater. However, in the current state, the energy and water recovery potential of MD is lower compared to other conventional distillation technologies. To address these issues, this study proposes and assesses a novel solar-powered hybrid Sweeping Gas MD (SGMD) unit integrated with a humidification dehumidification (HDH) system. Real textile wastewater from bleaching and dyeing processes is treated in the hybrid SGMD-HDH system to recover freshwater. An optical-thermal sub-model for the parabolic trough collector and the heat and mass transfer model for the dehumidifier are developed and validated using experimental measurements. The results reveal that the hybrid SGMD-HDH can exhibit nearly 20% higher water flux and 50% higher gain output ratio (GOR) compared to the standalone MD system. The highest water production and average GOR for bleaching wastewater feed solution reached 11.72 kg/day and 0.64, respectively. The higher concentration of chemical contaminants in dyeing wastewater decreased water flux and GOR by up to 14% and 10%, respectively, compared to bleaching wastewater. Elemental analysis showed increased carbon concentrations on the polytetrafluoroethylene membrane surface arising from organic fouling.

Enhanced solar desalination with photothermal hydrophobic carbon nanotube-infused PVDF membranes in air-gap membrane distillation

This work aims to increase the efficiency of the solar powered-air gap membrane distillation (SP-AGMD) process; a desalination method driven by solar energy, providing an eco-friendly and sustainable approach to addressing global water shortages. The innovation lies in integrating photothermal hydrophobic multiwalled carbon nanotubes (h-MWCNTs), in varying weight percentages from 5 % to 60 %, into polyvinylidene fluoride (PVDF) membranes using the phase inversion membrane fabrication technique. The h-MWCNTs were synthesized through oxidation and functionalization with oleylamine (OL) to improve their photothermal properties. Their successful integration was confirmed via scanning electron microscopy (SEM), thermogravimetric analysis (TGA), Fourier transform infrared spectroscopy (FTIR), and Raman spectroscopy. The h-MWCNTs-based SP-AGMD membranes were further evaluated for wettability, liquid entry pressure (LEP), surface temperature, and solar absorbance, demonstrating significant solar light absorption and localized surface heating. This generated the necessary driving force for the AGMD process. Performance metrics such as vapor flux, salt rejection, specific thermal energy consumption, photothermal efficiency, and temperature polarization (TP) coefficient were significantly improved, especially with a 20 % h-MWCNTs addition, which tripled the solar-energy-driven flux and increased photothermal efficiency by 326 % under standard solar conditions, compared to unmodified membranes. All h-MWCNTs-based SP-AGMD membranes achieved over 99 % salt rejection. Lastly, the membranes were tested with real seawater to confirm their applicability for desalination. This photothermal approach offers a scalable, sustainable solution for water purification, making a significant advancement in membrane distillation technology.

Springtail-inspired omniphobic slippery membrane with nano-concave re-entrant structures for membrane distillation

Omniphobic membranes, due to their exceptional properties, have drawn significant attention for overcoming the bottleneck in membrane distillation (MD) technology. This study demonstrates an innovative method for fabricating an omniphobic membrane that is simple and facile compared to other methods such as wet/dry etching and photolithography. The surface morphology of springtails was imitated using electrospraying technique to coat a polyvinylidene fluoride substrate with concave-shaped polystyrene beads that were successfully developed by controlling the electrical traction (voltage) and air resistance (humidity). Then, the lipid coating of springtail surfaces was mimicked by dip-coating the membrane in a low-toxicity short-chain perfluoropolyether lubricant. The concave structure’s tiny air pockets increased membrane hydrophobicity significantly, indicated by the fact that the first round of water bouncing took only 16.3 ms. Finally, in MD treatment of seawater containing 1.0 mM sodium dodecyl sulfate, the optimized omniphobic membrane maintained a stable 99.9% salt rejection rate.

Functional of thin-film composite Janus membrane with sacrificial layer for inorganic scaling control in membrane distillation

Membrane distillation technology, utilized for treating hypersaline wastewater from seawater desalination, often encounters challenges related to inorganic scaling, adversely affecting membrane performance. Herein, we introduce a innovative approach employing a sacrificial layer on the surface of Thin Film Composite (TFC) membranes to concurrently enhance inorganic scaling resistance and facilitate membrane reusability. The sacrificial layer (Fe3+-TA) consisted of tannic acid (TA) complexed with iron ions (Fe3+) and could be removed and regenerated in situ. The results demonstrated that the Fe3+-TA layer significantly improved the membrane's surface smoothness and densification, maintaining superior anti-scaling performance. The modified membrane exhibited remarkable durability, sustaining six reuse cycles with a flux recovery exceeding 97 % in gypsum scaling tests. Furthermore, the formation of new complexes during gypsum scaling tests confirmed the membrane's augmented scaling retardation capabilities. Thus, integrating of a sacrificial layer into TFC membranes presents a promising strategy for advancing membrane distillation processes in hypersaline wastewater treatment.

An adversarial network reinforcement method for power data time series generation based on knowledge transfer and Rayleigh differential privacy

Abstract In order to solve the problem of sensitive information leakage of power data in machine learning and artificial intelligence processing, this paper proposes a method of civic-text data reinforcement based on time series generation adversarial network, which combines knowledge transfer and Rényi differential privacy protection framework to balance privacy and availability, while improving the accuracy of the model. First, the generative adversarial network generates training data that is similar to the real data but secure through adversarial learning, ensuring that even if the data is stolen, the original information cannot be inferred. Secondly, through the integration of recurrent neural network training and autoencoder, the dynamic and static characteristics of the power load data can be effectively captured, and the high dimension of the learning space can be reduced. In addition, knowledge distillation technology is used to guide the generator of training student models through a powerful teacher network of knowledge transfer, improving the diversity and quality of the generated data. Finally, from the four aspects of privacy, data similarity and diversity, and practicability, the practical application value and potential of the framework for electricity consumption data are comprehensively evaluated.

Pilot-scale testing of a multi-tube type falling film distillation column equipped with a biphasic thermosyphon as a new alternative for the desalination of brackish water and seawater

This paper proposes an innovative technology for desalinating brackish water and seawater using a multi-tube type falling film distillation column integrated with a biphasic thermosyphon. Based on the literature survey, this proposal has not been previously explored. In this study, the viability of the pilot-scale application of this technology for desalination was tested, and the process performance was evaluated in terms of distillate flow rate, salinity removal, and energy consumption, considering different experimental conditions. Synthetic solutions containing 10.0 and 35.0 g/L of sodium chloride were used to simulate brackish water and seawater salinities, respectively. The thermal desalination pilot plant integrating a compact falling film distillation column and a biphasic thermosyphon demonstrated high effectiveness, consistently producing desalinated water with a conductivity below 10 μS cm−1. Considering both concentrations, the optimal condition for desalinated water production was a feed temperature of 85 °C, a vapor chamber temperature of 121 °C, and an energy consumption of 16 kW. This new technological option’s energy consumption is approximately 33 % lower than that of a simulated flash distillation column operating under similar conditions. In conclusion, this study presents promising results, establishing falling film distillation technology as a viable alternative for desalinating brackish water and seawater.

Identification of internal voids in pavement based on improved knowledge distillation technology

Investigating methods for the detection of internal voids within road structures is a critical measure to ensure the safety and integrity of roadway operations. The purpose of this research is to investigate on the identification method of internal voids in pavement based on improved knowledge distillation technology. Ground penetrating radar data in three dimensions were extensively collected to capture the internal voids present within roadways, and this data was subsequently validated through in-situ verification. The echo characteristics of ground penetrating radar for areas with road voids were analyzed, and a dataset containing 1700 images of these internal voids was established. A YOLOv8 model improvement method was proposed, and a model for the detection of internal road voids was constructed based on the improved YOLO v8 framework. To further refine the model's performance, a knowledge distillation method based on multiple guidance from teacher assistants was developed. A stochastic learning approach was integrated, resulting in the establishment of a model optimized by this stochastic learning scheme for the identification of internal road voids. The results demonstrate that the presence of overfitting during the training phase of the void identification model can restrict its performance within a certain domain. The proposed stochastic learning-optimized, multi-teacher assistant guided knowledge distillation model, adeptly harnesses the performance benefits of both the teacher and assistant models by means of knowledge transfer, consequently achieving a significant improvement in the detection of internal road voids.

Automated hybrid distillation sequence synthesis framework of mixed azeotropes and nonazeotropes

Distillation technology is a crucial, energy-intensive component of energy-supply systems, thereby motivating the improved design of distillation systems. This study investigated an automated framework for synthesizing hybrid distillation sequences, providing a systematic approach for the design and optimization of distillation sequences involving azeotropes. The proposed method separated mixed azeotropes and nonazeotropes by replacing the simple columns in the superstructure of a conventional distillation sequence with extractive distillation columns. The automated calculation of the superstructures was implemented by calling the external interface of the process simulator. The results of these calculations were used to formulate a mixed-integer linear programming problem that was optimized with the objective function of the total annualized cost (TAC). Finally, the optimal solution and all feasible alternatives were obtained. Moreover, the reliability and efficiency of the proposed framework were demonstrated through two case studies. The results show that the separation position of the azeotrope in the sequence and the presence or absence of heat integration significantly impact the TAC of the system. However, using an existing component versus an external component as an entrainer makes very little difference in the TAC for the optimal sequences.

Sustainable process design and multi-objective optimization of efficient and energy-saving separation of xylene isomers via extractive distillation based on double extractants

The focus of the chemical industry on environmental protection and energy conservation has driven the development of green distillation technology. Xylene isomers are extensively utilized industrial raw materials, whose separation is challenging and energy-intensive due to their closely similar boiling points. Therefore, it is urgent to develop energy-saving and environmental protection technology. In this paper, an energy-saving and environmental protection process for separating the xylene isomers by extractive distillation was designed combing the vapor recompression process with feed preheating (VRAPFP). The parameters of extractive distillation processes were synthetically optimized by multi-objective genetic algorithm (MOGA) with total annual cost (TAC), CO2 emission (ECO2) and separation efficiency (Eext) as the objective evaluation indexes. The economical optimal process was used further for VRAPFP design through heat integration to reduce energy consumption, and all the designed candidate processes were synthetically analyzed, compared and evaluated with performance indexes. Results show that the VRAPFP process has superior economic efficiency and ecological sustainability compared with the existing distillation process, wherein the TAC and ECO2 are reduced by 14.2 % and 30.2 %, respectively. This work provides a new reference and idea for separating xylene isomers with the economy and developing new processes for energy conservation and environmental protection.

Performance of capacitive deionizing membrane distillation (CDIMD) process for the organic sewage treatment and how operating parameters can affect it

To solve the fouling issue of traditional membrane distillation (MD) technique, a newly-designed hybrid purification system integrating capacitive deionization (CDI) with MD technology was established for the organic sewage treatment in this study, namely capacitive deionizing membrane distillation (CDIMD) technology. Complexes formation via cross-linking among protein, divalent cations, and polysaccharide, an archcriminal inducing MD fouling, was remarkably alleviated in the CDIMD system considering that negatively-charged organics were clearly divided away from those divalent cations in an electric field circumstance. The CDIMD performance was deeply explored in this study focusing on the impact mechanisms of core operational parameters. An appropriate increment of feed temperature (60–65 °C) and feed flow velocity (15.72–17.82 mm/s) contributed to the anti-fouling performance of CDIMD. Considering the side reactions at voltage >1.2 v, setting working voltage at around 1.2 v seemed to be optimal for CDIMD operation. With an overall consideration of pure water production and anti-fouling performance, a properly thicker CNTs layer (40–60 μm) efficiently enhanced the fouling mitigation of CDIMD. A frequent switching of electric field direction (9/1 min) facilitated the timely regeneration of electrode and composite membranes, accordingly promoting a reliable purification of organic wastewater by CDIMD technique.

Suitable Pore Confinement with Multiple Interactions in a Low-Cost Zeolitic Imidazolate Framework for Efficient Xe Capture and Separation

The increasing industrial demand for the noble gases xenon (Xe) and krypton (Kr), along with growing environmental concerns on their radionuclides, has made the efficient separation and purification of Xe and Kr a critical issue. Leveraging advanced porous materials to achieve highly efficient adsorptive Xe/Kr separation has been recognized as an energy-saving and economical alternative to the widely used cryogenic distillation technology. Herein, we present a stable, low-cost, and easily scalable zeolitic imidazolate framework (ZIF) material, ZIF-4, for effective xenon adsorption and separation. ZIF-4 possesses uniformly distributed pore cavities with ideally matched pore size (4.3 Å) with that of Xe atom (4.047 Å), providing appropriate pore confinement for Xe binding. Sorption isotherms of ZIF-4 demonstrate the remarkably high Xe uptake (1.64 mmol/g at 0.2 bar and 298 K) and excellent Xe/Kr separation selectivity (16.2 for IAST selectivity of 20/80 Xe/Kr gas mixtures). Breakthrough experiments further validate the efficient separation performance of ZIF-4 in both binary Xe/Kr (20/80) gas mixtures and under dilute conditions. Single-crystal X-ray diffraction studies of activated and Xe-loaded ZIF-4 crystals reveal that the flexible pore cavities of ZIF-4 can self-adaptively adsorb and accommodate Xe atoms through the rotation of imidazolate rings, forming multiple Xe···H and Xe···π interactions to create strong binding sites for Xe capture and discrimination. The stability, economic feasibility, and scalability of ZIF-4 underscore its significant potential for industrial Xe/Kr separation.

Dynamic Control Analysis of Various Side‐Stream Quaternary Extractive Distillation Configurations

AbstractA thermally coupled distillation technology can bring energy‐saving benefits, but it poses challenges to process control. This article explores dynamic control of different side‐stream quaternary extractive distillation configurations. One the one hand, the open‐loop controllability of these processes is analyzed in terms of various criteria by the control design interface technology of Aspen Plus Dynamics. On the other hand, their control structures are established and examined by introducing large feed flow and composition perturbations. The results show that the triple‐side‐stream distillation still performs the best state and input–output controllability in spite of the strongest nonlinearity, and is also well resistant to large feed perturbations using a simple control structure.

UAWC: An intelligent underwater acoustic target recognition system for working conditions mismatching

Underwater acoustic target recognition (UATR) systems are crucial to both military and civilian activities. However, the complex ship working conditions will largely affect the performance of recognition systems, especially in the case of working conditions mismatching (WCMM). For WCMM problems, an intelligent UATR system for working condition mismatching (UAWC) is proposed. UAWC uses auditory features as input to the system and uses knowledge distillation to learn the intrinsic connections of target features under different working conditions. In the proposed approach, the teacher network obtains initial knowledge by utilizing a large amount of existing working condition data. Next, the student network uses a small amount of target working data for training, and extracts incremental knowledge in teacher network through knowledge distillation technology to enhance the accuracy of its classification of target working data, so as to effectively deal with the WCMM problem.The tests make use of datasets for ship-radiated noise under various working conditions.The results showed that UAWC performs better than other methods on a wide range of WCMM problems.

Exergoeconomic assessment of a multi-section solar distiller coupled with solar air heater: Optimization and economic viability

This study focuses on optimizing the solar distiller and provides a foundation for future research and development in solar distillation technologies. This paper does an exergoeconomic analysis of a multi-section solar distiller (MS-SS) system that is combined with a solar air heater (SAH). Therefore, its objective is to determine the most effective operational and design parameters while taking into account the system's lifespan and interest rate. This will offer significant knowledge for improving the MS-SS. Key parameters examined in the study include exergoeconomic indicators, exergy production factor (EPF), exergy-based loss rate, exergy efficiency, and water cost. Variations in exergy loss rate across different seasons were observed, with implications for system efficiency. The exergoeconomic analysis demonstrates the impact of interest rates on the system's economic viability, with corresponding effects on water costs. The key findings suggest varying daily output and exergy efficiency, with the lowest values recorded in winter and the highest in summer. Summer demonstrates superior output exergy, averaging of 145 × 103 Wh per month. In the normal case of output salinity between 0 and 500 ppm, the cost was reduced by 96.7 %, and by 96.3 % in the case of four interior sections. Water cost reduction is also observable at the same output salinity, with 54.1 % at 0 ppm and 62.4 % at 200 ppm between number of interior sections = 0 and 4.