Output-relevant Variational autoencoder for Just-in-time soft sensor modeling with missing data

Main challenges for developing data-based models lie in the existence of high-dimensional and possibly missing observations that exist in stored data from industry process. Variational autoencoder (VAE) as one of the deep learning methods has been applied for extracting useful information or features from high-dimensional dataset. Considering that existing VAE is unsupervised, an output-relevant VAE is proposed for extracting output-relevant features in this work. By using correlation between process variables, different weight is correspondingly assigned to each input variable. With symmetric Kullback–Leibler (SKL) divergence, the similarity is evaluated between the stored samples and a query sample. According to the values of the SKL divergence, data relevant for modeling are selected. Subsequently, Gaussian process regression (GPR) is utilized to establish a model between the input and the corresponding output at the query sample. In addition, owing to the common existence of missing data in output data set, the parameters and missing data in the GPR are estimated simultaneously. A practical debutanizer industrial process is utilized to illustrate the effectiveness of the proposed method.     

Preparation,physicochemical characterization and release behavior of the inclusion complex of trans-anethole and β-cyclodextrin

Trans-anethole (AT) has a variety of antimicrobial properties and is widely used as food functional ingredient. However, the applications of AT are limited due to its low water solubility, strong odor and low physicochemical stability. Therefore, the aim of this work was to encapsulate AT with β-cyclodextrin (β-CD) for obtaining inclusion complex by co-precipitation method. The measurements effectively confirmed the formation of inclusion complex between AT and β-CD. The results showed that the inclusion complex presented new solid crystalline phases and was more thermally stable than the physical mixture and β-CD. The phase solubility study showed that the aqueous solubility of AT was increased by being included in β-CD. The calculated stability constant of inclusion complex was 1195 M^− 1, indicating the strong interaction between AT and β-CD. Furthermore, the release study suggested that β-CD provided the protection for AT against evaporation. The release behavior of AT from the inclusion complex was controlled.     

Hydrogen reduced graphene oxide/metal grid hybrid film: towards high performance transparent conductive electrode for flexible electrochromic devices

The metal grid and reduced graphene oxide (RGO) are both promising transparent conductive materials for replacing the indium tin oxide (ITO) in flexible optoelectronics. However, the large empty area that exists in the grid together with the relatively high sheet resistance of RGO hinder both the materials for practical applications. In this work, we report for the first time a novel strategy for efficient combination of the metal grid and RGO by using a newly developed room-temperature reduction technique. The obtained RGO/metal grid hybrid films not only overcome the shortcomings of individual components but exhibit enhanced optical and electrical performances (R_s = 18 Ω sq⁻¹ and T = 80%) and excellent flexural endurance. With this hybrid film as the window electrode, a highly flexible electrochromic device with excellent stability and ultra-fast response shorter than 60 ms has been successfully fabricated. Considering its high efficiency, high quality, low cost and large area, the strategy would be particularly useful for economically fabricating various metal grid/RGO films which are quite promising high performance transparent and conductive materials for next generation optoelectronic devices.     

Monomer casting nylon-6-b-polyether amine copolymers: Synthesis and properties

MC nylon-6-b-polyether amine copolymers were prepared with macro-initiator based on amino-terminated polyether amine functionalized with isocyanate via in-situ polymerization. It was found that the introduction of polyether amine delayed the polymerization process of caprolactam by increasing apparent activation energy and pre-exponential factor, resulting in the decrease of molecular weight of nylon-6. The motion of molecular chain of the copolymers was easy because of the decreased hydrogen bonds and weakened inter-molecular forces. The physical entanglement of molecular chains of the copolymers was significant and strong which increased the entanglement density. Only the nylon-6 phase crystallized in the copolymers and the crystal grain size, spherulite size and crystallinity of the copolymers decreased. A small amount of γ crystal formed at high polyether amine content. The copolymers presented obvious strain hardening behavior in stress-strain curves and the loss factor dramatically increased while the glass transition temperature and storage module decreased. The fracture surface of the copolymers became rough and presented hairy structure, indicating that the toughening mechanism of the copolymers corresponded to the multi-layer crack extension mechanism.     

Integrated circuit and system design for renewable energy inverters

This study proposes that a novel integrated circuit (IC) and system design for renewable energy inverters can harvest renewable energy to power direct current (DC) and alternating current (AC) loads. In addition, an intelligent synthesis and management tool is developed to design the proposed system and to judge the system’s operational maintenance decisions. Finally, a renewable energy inverter’s information is posted to an online system. Users can obtain the proposed system’s information at any time and place. The accurate and superior performance of the proposed IC and system is confirmed by computer simulations and experimental results.     

Industrial sugar beets to biofuel: Field to fuel production system and cost estimates

Specialized varieties of sugar beets (Beta vulgaris L.) may be an eligible feedstock for advanced biofuel designation under the USA Energy Independence and Security Act of 2007. These non-food industrial beets could double ethanol production per hectare compared to alternative feedstocks. A mixed-integer mathematical programming model was constructed to determine the breakeven price of ethanol produced from industrial beets, and to determine the optimal size and biorefinery location. The model, based on limited field data, evaluates Southern Plains beet production in a 3-year crop rotation, and beet harvest, transportation, and processing. The optimal strategy depends critically on several assumptions including a just-in-time harvest and delivery system that remains to be tested in field trials. Based on a wet beet to ethanol conversion rate of 110 dm³ Mg⁻¹ and capital cost of 128 M$ for a 152 dam³ y⁻¹ biorefinery, the estimated breakeven ethanol price was 507 $ m⁻³. The average breakeven production cost of corn (Zea mays L.) grain ethanol ranged from 430 to 552 $ m⁻³ based on average net corn feedstock cost of 254 and 396 $ m⁻³ in 2014 and 2013, respectively. The estimated net beet ethanol delivered cost of 207 $ m⁻³ was lower than the average net corn feedstock cost of 254–396$ m⁻³ in 2013 and 2014. If for a mature industry, the cost to process beets was equal to the cost to process corn, the beet breakeven ethanol price would be $387 m^-3 (587 $ m⁻³ gasoline equivalent).     

Electrochemically deposited Fe2O3 nanorods on carbon nanofibers for free-standing anodes of lithium-ion batteries

Fe₂O₃ nanorod/carbon nanofiber (CNF) composites were prepared by the electrochemical deposition of Fe₂O₃ on a web of CNFs, which was then used as a free-standing anode. The conductive, three-dimensional structure of the CNF web allowed for the electrodeposition of the Fe₂O₃ nanorods, while its high conductivity made it possible to use the composite as a free-standing electrode in lithium-ion batteries. In addition, it was easy and cheap to fabricate by a simplification of a process of cell preparation. The nanorod-like Fe₂O₃ structures could only be electrodeposited on the CNFs; flake-like Fe₂O₃ was formed on flat conductive glass substrates. It can be attributed to the different growth mechanism of Fe₂O₃ on the CNFs because of the large number of reaction sites on the CNFs, differences in the precursor concentration and diffusivity within the CNF web. The formation of aggregates of the Fe₂O₃ particles on thicker CNFs also indicated that the CNFs had determined the Fe₂O₃ growth mechanism. The synthesised Fe₂O₃/CNF composite electrode exhibited stable rate capacities at different current densities. This suggested that CNF-based composite did not exhibit the intrinsic disadvantages of Fe₂O₃. Finally, carbon coatings were deposited on the Fe₂O₃/CNF composites to further improve their electronic conductivity and rate capability.     

Highly active sites of low spin FeⅡN4species:The identification and the ORR performance

Over recent years,catalytic materials of Fe-N-C species have been recognized being active for oxygen reduction reaction(ORR).However,the identification of active site remains challenging as it generally involves a pyrolysis process and mixed components being obtained.Herein Fe₃C/C and Fe₂N/C samples were synthesized by temperature programmed reduction of Fe precursors in 15%CH₄/H₂and pure NH₃,respectively.By acid leaching of Fe₂N/C sample,only single sites of FeN₄species were presented,providing an ideal model for identification of catalytic functions of the single sites of FeN₄in ORR.A correlation was conducted between the concentration of Fe^ⅡN₄in low spin state by Mossbauer spectra and the kinetic current density at 0.8 V in alkaline media,and such a structure-performance correlation assures the catalytic roles of low spin Fe^ⅡN₄ species as highly active sites for the ORR.     

水热法一步合成磷酸铁锂及其性能研究

为优化液相法一步制备磷酸铁锂（LiFePO₄）技术,以七水合硫酸亚铁、磷酸二氢铵、一水合氢氧化锂为原料,通过添加十二烷基苯磺酸钠（SDBS）作为表面活性剂,采用液相水热法合成技术,一步合成了LiFePO₄正极材料。研究了水热法一步合成技术对LiFePO₄材料的组成、结构、形貌、粒度等的影响,通过电感耦合等离子体发射光谱仪（ICP-OES）、X射线衍射仪（XRD）、扫描电镜（SEM）、粒度分析仪等对材料进行了表征分析,并测试了材料的电化学性能。研究结果表明,合成得到的LiFePO₄材料为微米级球形颗粒形貌的正交晶系非化学计量比的Li_1.02Fe_0.994PO₄材料。电化学性能测试结果表明,在0.1C倍率下首次充、放电比容量分别为162.0、159.9 mA·h/g,库伦效率达到98.7%、倍率性能（以1C/0.1C保持率计）为92.3%,0.1C倍率循环100次容量保持率为96.4%,展现出良好的电化学性能。     

Hydrothermal synthesis of 3D NixCo1−xS2 particles/graphene composite hydrogels for high performance supercapacitors

In this work, three dimensional (3D) Ni_xCo_1−xS₂/graphene composite hydrogels with different Ni contents (denoted as Ni_xCo_1−xS₂/GH (x = 0, 0.31, 0.56, 0.66, 1)) have been synthesized by a simple one-step hydrothermal method and utilized as the active materials of supercapacitors. The as-prepared samples present a 3D interconnected porous network with the pore sizes in the range of several to tens micrometers. Interestingly, the Ni_xCo_1−xS₂ particles are uniformly located on the graphene network and the particle size is evolved from ∼50 nm to ∼1.5 μm with the increase of Ni content. The electrochemical measurements revealed that the specific capacitance, rate capability and cyclability of different Ni_xCo_1−xS₂/GH electrodes are strongly affected by their different Ni content. Among these, the 3D Ni_0.31Co_0.69S₂/GH composite has the highest specific capacitance of 1166 F/g at a current density of 1 A/g. Furthermore, a specific capacitance of 559 F/g can be still maintained at high current density of 20 A/g. After 1000 charge–discharge cycles at 5 A/g, the specific capacitance remains a high value of 755 F/g.