基于PCAU-Net的快速多通道磁共振成像方法

多通道磁共振成像方法采用多个接收线圈同时欠采样k空间以加快成像速度,并基于后处理算法重建图像,但在较高加速因子时,其图像重建质量仍然较差．本文提出了一种基于PCAU-Net的快速多通道磁共振成像方法,将单通道实数U型卷积神经网络拓展到多通道复数卷积神经网络,设计了一种结构不对称的U型网络结构,通过在解码部分减小网络规模以降低模型的复杂度．PCAU-Net网络在跳跃连接前增加了1×1卷积,以实现跨通道信息交互．输入和输出之间利用残差连接为误差的反向传播提供捷径．实验结果表明,使用规则和随机采样模板,在不同加速因子时,相比常规的GRAPPA重建算法和SPIRiT重建方法,本文提出的PCAU-Net方法可高质量重建出磁共振复数图像,并且相比于PCU-Net方法,PCAU-Net减少了模型参数、缩短了训练时间．     

结合PBL的思维导图法在大学化学“四大平衡”教学中的应用

以化学反应"四大平衡"的主体内容:酸碱电离平衡、配位平衡、沉淀平衡和氧化还原平衡为例,介绍结合问题导向学习的思维导图法在大学化学教学实践中的应用。     

演示实验制作促进动手动脑的大学物理探索式学习的改革尝试

介绍了我们通过制作演示实验装置促进动手动脑的大学物理探索式学习的教学改革实践.详细说明了动手动脑学习的试验过程、试验结果.自制物理演示实验装置给学生提供了一个可以自主选题的探索式学习平台.     

人工智能在锂离子电池研发中的应用

锂离子电池已成为解决现代社会储能问题的最佳解决方案之一。然而,电池材料和器件开发都是复杂的多变量问题,传统的依赖研究人员进行实验的试错法在电池性能提升方面遇到了瓶颈。人工智能（AI）具有强大的高速、海量数据处理能力,是上述突破研究瓶颈的最具潜力的技术。其中,机器学习 (ML) 算法在评估多维数据变量和集合之间的组合关联方面的独特优势有望帮助研究人员发现不同因素之间的相互作用规律并阐明材料合成和设备制造的机制。本综述总结了锂离子电池传统研究方法遇到的各种挑战,并详细介绍了人工智能在电池材料研究、电池器件设计与制造、材料与器件表征、电池循环寿命与安全性评估等方面的应用。最重要的是,我们介绍了AI和ML在电池研究中面临的挑战,并讨论了它们应用的缺点和前景。我们相信,未来实验科学家、数学建模专家和AI专家之间更紧密的合作将极大地促进AI和ML方法用以解决传统方法难以克服的电池和材料问题。     

Machine learning as an online diagnostic tool for proton exchange membrane fuel cells

Proton exchange membrane fuel cells are considered a promising power supply system with high efficiency and zero emissions. They typically work within a relatively narrow range of temperature and humidity to achieve optimal performance; however, this makes the system difficult to control, leading to faults and accelerated degradation. Two main approaches can be used for diagnosis, limited data input which provides an unintrusive, rapid but limited analysis, or advanced characterisation that provides a more accurate diagnosis but often requires invasive or slow measurements. To provide an accurate diagnosis with rapid data acquisition, machine learning methods have shown great potential. However, there is a broad approach to the diagnostic algorithms and signals used in the field. This article provides a critical view of the current approaches and suggests recommendations for future methodologies of machine learning in fuel cell diagnostic applications.     

Single-Molecule Sensing of Acidic Catecholamine Metabolites Using a Programmable Nanopore

Acidic catecholamine metabolites, which could serve as diagnostic markers for many diseases, demonstrate an importance of accurate sensing. However, they share a highly similar chemical structure, which is a challenge in the design of sensing strategies. A nanopore may be engineered to sense these metabolites in a single molecule manner. To achieve this, a recently developed programmable nano-reactor for stochastic sensing (PNRSS) technique adapted with a phenylboronic acid (PBA) adaptor was applied. Three acidic catecholamine metabolites, including 3,4-dihydroxyphenylacetic acid (DOPAC), 3,4-dihydroxymandelic acid (DHMA) and 3-methoxy-4-hydroxymandetic acid (VMA) were investigated by PNRSS. Specifically, DHMA, which contains an α-hydroxycarboxylate moiety and an adjacent cis-hydroxyl groups on its benzene ring, reports two binding modes simultaneously resolvable by PNRSS. Assisted with the high resolution of PNRSS, direct regulation of these two binding modes by pH can also be observed. A custom machine learning algorithm was also developed to achieve automatic event classification.     

Optimization of heterogeneous Catalyst-assisted fatty acid methyl esters biodiesel production from Soybean oil with different Machine learning methods

There is a growing attention to the bio and renewable energies due to fast depletion of fossil fuels as well as the global warming problem. Here, we developed a modeling and simulation method by means of artificial intelligence (AI) for prediction of the bioenergy production from vegetable bean oil. AI methods are well known for prediction of complex and nonlinear process. Three distinct Adaptive Boosted models including Huber regression, LASSO, and Support Vector Regression (SVR) as well as artificial neural network (ANN) were applied in this study to predict actual yield of Fatty acid methyl esters (FAME) production. All boosted utilizing the Adaptive boosting algorithm. The important influencing parameters on the biodiesel production such as the catalyst loading (CAO/Ag, wt%) and methanol to oil (Soybean oil) molar ratio were selected as the input variables of models while the yield of FAME production was selected as output. Model hyper-parameters were tuned to maintain generality while improving prediction accuracy. The models were evaluated using three distinct metrics Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R². Error rates of 8.16780E-01, 4.43895E-01, 2.06692E + 00, and 3.92713 E-01 were obtained with the MAE metric for boosted Huber, SVR, LASSO and ANN models. On the other hand, the RMSE error of these models were about 1.092E-02, 1.015E-02, 2.669E-02, and 1.01174E-02, respectively. Finally, the R-square score were calculated for boosted Huber, boosted SVR, and boosted LASSO as 0.976, 0.990, 0.872, and 0.99702, respectively. Therefore, it can be concluded that although the boosted SVR and ANN models were better models for prediction of process efficiency in terms of error, but all algorithms had high accuracy. The optimum yield of 83.77% and 81.60% for biodiesel production were observed at optimum operating values from boosted SVR and ANN models, respectively.     

Predictive modeling and computational machine learning simulation of adsorption separation using advanced nanocomposite materials

Adsorption process was simulated in this study for removal of Hg and Ni from water using nanocomposite materials. The used nanostructured material for the adsorption study was a combined MOF and layered double hydroxide, which is considered as MOF-LDH in this work. The data were obtained from resources and different machine learning models were trained. We selected three different regression models, including elastic net, decision tree, and Gradient boosting, to make regression on the small data set with two inputs and two outputs. Inputs are Ion type (Hg or Ni) and initial ion concentration in the feed solution (C₀), and outputs are equilibrium concentration (Ce) and equilibrium capacity of the adsorbent (Qe) in this dataset. After tuning their hyper-parameters, final models were implemented and assessed using different metrics. In terms of the R2-score metric, all models have more than 0.97 for Ce and more than 0.88 for Qe. The Gradient Boosting has an R2-score of 0.994 for Qe. Also, considering RMSE and MAE, Gradient Boosting shows acceptable errors and best models. Finally, the optimal values with the GB model are identical to dataset optimal: (Ion = Ni, C₀ = 250, Ce = 206.0). However, for Qe, it is different and is equal to (Ion = Hg, C0 = 121.12, Ce = 606.15). The results revealed that the developed methods of simulation are of high capacity in prediction of adsorption for removal of heavy metals using nanostructure materials.