In this communication, an experimental study of inverted absorber solar still (IASS) and single slope solar still (SS) at different water depth and total dissolved solid (TDS) is presented. ...Experiments are conducted for the climatic condition of Muscat, Oman. A thermal model is also developed for the IASS and validated with experimental results. A fair agreement is found for the daytime operation of the IASS. It is observed that higher water temperature can be achieved by using the IASS in comparison to the SS. The daily yield obtained from the IASS are 6.302, 5.576 and 4.299
kg/m
2-day at water depths (
d
w
) 0.01, 0.02 and 0.03
m respectively. At same respective water depths, the daily yield obtained from the SS are 2.152, 1.931, 0.826
kg/m
2-day respectively lower than that of the IASS. It is observed that for climatic condition of Muscat, Oman, the optimum water depth for the IASS is 0.03
m above which the addition of reflector under the basin does not affect its performance much more in comparison to that of the SS for sea water. The feed saline water and yielded distilled water are also compared for different TDS values, pH, and electrical conductance. On the basis of economic analysis of IASS, it is found that the annualized cost of distilled water in Indian rupees for Muscat climatic condition is Rs. 0.74, 0.66 and 0.62 (conversion factors: $ 1
=
Rs. 50 and 1 OMR
=
Rs. 120) for the life time of 15, 20 and 25
years respectively.
In scenarios like privacy protection or large-scale data transmission, data-free knowledge distillation (DFKD) methods are proposed to learn Knowledge Distillation (KD) when data is not accessible. ...They generate pseudo samples by extracting the knowledge from teacher model, and utilize above pseudo samples for KD. The challenge in previous DFKD methods lies in the static nature of their target distributions and they focus on learning the instance-level distributions, causing its reliance on the pretrained teacher model. To address above concerns, our study introduces a novel DFKD approach known as AdaDFKD, designed to establish and utilize relationships among pseudo samples, which is adaptive to the student model, and finally effectively mitigates the aforementioned risk. We achieve this by generating from “easy-to-discriminate” samples to “hard-to-discriminate” samples as human does. We design a relationship refinement module (R2M) to optimize the generation process, wherein we learn a progressive conditional distribution of negative samples and maximize the log-likelihood of inter-sample similarity of pseudosamples. Theoretically, we discover that such design of AdaDFKD both minimize the divergence and maximize the mutual information between the distribution of teacher and student models. Above results demonstrate the superiority of our approach over state-of-the-art (SOTA) DFKD methods across various benchmarks, teacher–student pairs, and evaluation metrics, as well as robustness and fast convergence.
A new graphical method for the fundamental design of pressure‐swing distillation (PSD) and valid heat integration of the designed process is demonstrated based on a pinch analysis. As the minimum ...pressure constraint, the pinch pressure is determined by this graphical method to secure the necessary temperature driving force and circumvent the distillation boundaries. For a feasibility evaluation of heat‐integrated PSD (HIPSD), the suggested approach requires visualization of the pinch temperature contours and the feasible composition region of an initial feed according to the pressure change and dynamic properties of ternary azeotropic systems. We thus carried out an in‐depth examination by conducting case studies to show the feasible integration into HIPSD. The new insight is that the HIPSD system can have different pinch pressures depending on the column sequence or the separation order of pure components. In addition, the analysis verifies that energy‐efficient HIPSD can be generated by considering pinch constraints.
Display omitted
Membrane distillation (MD) is an emerging technology for the desalination of brines. In some cases, liquid penetrates into the pores of the membrane, causing pore wetting. MD’s ...commercialization is hindered largely due to the occurrence of pore wetting phenomena since it results in the reduction of flux and/or permeate quality. Hence, it is of crucial importance for MD to prevent pore wetting from occurring. In this paper, the methods of detecting pore wetting and the membrane parameters related to this phenomenon are reviewed and possible sources of MD pore wetting occurrence are identified. Moreover, the methods to prepare membranes specifically designed for the mitigation of membrane wetting, such as the design of membrane materials, membrane surface modification, preparation of nanocomposite membranes by the addition of nanoparticles, dual-layered membranes, and membranes with a re-entrant structure are discussed. Finally, process-based approach for wetting mitigation, mainly by the pretreatment of the feed solution, is elucidated, and models for wetting phenomena are also outlined. Thus, attempts are made in this review to discuss all aspects of the pore wetting of MD membranes.
Display omitted
•Effects of pressure is investigated for an extractive distillation process separating DMC-MeOH mixture.•Increase in pressure of extractive column reduces the entrainer requirement ...significantly.•This results in a significant decrease of TAC compared with the conventional design.•CO2 emissions are also reduced by increasing the operating pressure.
Although dimethyl carbonate is defined as a green chemical, separation of dimethyl carbonate-methanol azeotropic mixture is an important issue for many dimethyl carbonate production processes. Extractive distillation process is considered as a favorable method for separating this mixture, but the recovery of entrainer still results in a significant loss of capital and operating costs. On the other hand, operating pressure is an important design variable for distillation columns since it has an important impact on column temperature and phase equilibrium. In this work, the effect of operating pressure is investigated for an extractive distillation process separating the dimethyl carbonate-methanol mixture using methyl isobutyl ketone as entrainer. It is observed that the increase in the operating pressure of extractive distillation column significantly decreases the amount of required entrainer flowrate. As the result, a process with an extractive distillation column operating at 10 bar reduces total annual cost and carbon dioxide emissions by 34.1% and 29.8%, respectively compared to the conventional process with an extractive distillation column operating at atmospheric pressure.
Federated learning is a distributed machine learning paradigm where the goal is to collaboratively train a high quality global model while private training data remains local over distributed ...clients. However, heterogenous data distribution over clients is severely challenging for federated learning system, which severely damage the quality of model. In order to address this challenge, we propose global prototype distillation (FedGPD) for heterogenous federated learning to improve performance of global model. The intuition is to use global class prototypes as knowledge to instruct local training on client side. Eventually, local objectives will be consistent with the global optima so that FedGPD learns an improved global model. Experiments show that FedGPD outperforms previous state-of-art methods by 0.22% ~1.28% in terms of average accuracy on representative benchmark datasets.
Display omitted
•The new processes for separating benzene/isopropanol/water is proposed.•Multi-objective genetic algorithm is applied to optimize the proposed processes.•The proposed hybrid ...reactive-extractive distillation processes have better economic and environmental performances.
The recycling of wastewater in chemical industry is of great significance because it can recover valuable organic solvents and prevent environmental pollution. In this work, benzene/isopropanol (IPA)/water ternary mixtures were separated by single and double reactive-extractive distillation column processes respectively. We conducted thermodynamic analysis on separation processes of ternary mixture to determine the feasible separation sequence. Furthermore, the processes are optimized by multi-objective genetic algorithm (MOGA) to obtain suitable operating conditions. And the processes were evaluated by the total annual cost (TAC), CO2 emissions (ECO2), extraction efficiency (Eext) and thermodynamic efficiency. The results showed that compared with the traditional triple column extractive distillation (TCED) process, the TAC and ECO2 of the double column reactive-extractive distillation (DCRED) process were dramatically reduced by 52.9% and 49.5%, respectively. Similarly, TAC and ECO2 decreased by 49.5% and 53.5% in reactive-extractive dividing wall column (REDWC) process compared with the traditional process. With application of intensive processes, the final proposed process is more economically and environmentally sustainable than the separation system proposed in currently available literatures. This study can provide reference for the efficient separation of industrial wastewater containing benzene and IPA.
•We present data-free knowledge distillation method for regression.•It adopts generator that creates synthetic data to transfer knowledge to student.•Given teacher, generator and student are trained ...in adversarial manner.•Generator is trained to synthesize data on which student is unable to mimic teacher.•Student is trained on synthetic data to mimic teacher’s predictions.
Knowledge distillation has been used successfully to compress a large neural network (teacher) into a smaller neural network (student) by transferring the knowledge of the teacher network with its original training dataset. However, the original training dataset is not reusable in many real-world applications. To address this issue, data-free knowledge distillation, which is knowledge distillation in the absence of the original training datasets, has been studied. However, existing methods are limited to classification problems and cannot be directly applied to regression problems. In this study, we propose a novel data-free knowledge distillation method that is applicable to regression problems. Given a teacher network, we adopt a generator network to transfer the knowledge in the teacher network to a student network. We simultaneously train the generator and student networks in an adversarial manner. The generator network is trained to create synthetic data on which the teacher and student networks make different predictions, with the student network being trained to mimic the teacher network’s predictions. We demonstrate the effectiveness of the proposed method on benchmark datasets. Our results show that the student network emulates the prediction ability of the teacher network with little performance loss.
Membrane distillation (MD) has shown potential as a means of desalination and water purification. As a thermally driven membrane technology which runs at relatively low pressure, which can withstand ...high salinity feed streams, and which is potentially more resistant to fouling, MD could be used for desalination where reverse osmosis is not a good option. The use of thermal energy, rather than electrical energy, and the fact that MD membranes can withstand dryout make this technology attractive for renewable power applications as well. However, most research on MD has focused on maximizing membrane flux as opposed to minimizing energy consumption and cost, and current MD systems suffer from poor energy efficiency compared to other desalination systems. In solar driven systems, the reported thermal performance has not been much better than a simple solar still. This paper examines the energy efficiency of single-stage MD-based desalination cycles in each of the MD configurations commonly used for desalination (direct contact, air gap, and vacuum) and compares the gained output ratio, or GOR, of each configuration across the range of membrane module geometries, and operating conditions. Limitations of each configuration are identified. Direct contact MD and air gap MD, in particular, have potential for high GOR.
PDF
