Agent-based architecture for context-aware and personalized event recommendation

In order to offer context-aware and personalized information, intelligent processing techniques are necessary. Different initiatives considering many contexts have been proposed, but users preferences need to be learned to offer contextualized and personalized services, products or information. Therefore, this paper proposes an agent-based architecture for context-aware and personalized event recommendation based on ontology and the spreading algorithm. The use of ontology allows to define the domain knowledge model, while the spreading activation algorithm learns user patterns by discovering user interests. The proposed agent-based architecture was validated with the modeling and implementation of eAgora? application, which was illustrated at the pervasive university context.     

Swelling activation energy of κ‐carrageenan in its gel state: A fluorescence study

A steady‐state fluorescence technique was employed to study the swelling of κ‐carrageenan gels at various temperatures. Pyranine was used as a fluorescence probe. The fluorescence intensity of pyranine was measured during the insitu swelling process of κ‐carrageenan gels. The fluorescence intensity increased exponentially as the swelling time increased. The increase in the fluorescence intensity was modeled with the Li–Tanaka equation, from which the swelling time constants and cooperative diffusion coefficients were determined. The swelling time constants decreased and the cooperative diffusion coefficients increased as the swelling temperature was increased. The swelling activation energies were measured to be 47.05 kJ/mol. © 2007 Wiley Periodicals, Inc. J Appl Polym Sci 2007     

Electrochemical insertion of lithium ions into disordered carbons derived from reduced graphite fluoride

Both covalent (obtained by direct fluorination at high temperature) and semi-ionic carbon fluorides (synthesized at room temperature) were reduced in order to obtain disordered carbons containing very small content of fluorine and different physical properties according to the reduction treatment (chemical, thermal or electrochemical). After a physical characterization (X-ray diffraction, electron spin resonance and FT-IR spectroscopies), the electrochemical behaviours of the pristine carbon fluorides and of the treated samples were investigated during the insertion of lithium using liquid carbonate-based electrolytes (LiClO₄-EC/PC, 50:50%, v/v). Both galvanostatic and voltammetric modes were performed and revealed that the voltage profiles and the capacities differed according to the starting material and the reduction treatment. Semi-ionic carbon fluoride treated in F₂ atmosphere for 2 h at 150 °C and then chemically reduced in KOH exhibits high reversible capacities (the reversible capacity is 530 mAh g⁻¹ in the second cycle); in this case, the voltage profiles show a large flat portion at potentials lower than 0.3 V which is attributed to the insertion/deinsertion of lithium ions between the small graphene sheets and/or the absorption of pseudo metallic lithium into the microporosity of the sample. Nevertheless, a part of the lithium ions are removed at potentials higher than 0.5 V versus Li⁺/Li limiting the useful capacity.     

Electrochemical insertion of sodium into hard carbons

The electrochemical insertion of sodium ions into different types of hard carbons was achieved in electrolytes composed of ethylene carbonate as the solvent and NaClO₄ as the salt. For all the materials studied the sodium uptake increases when the carbon highest heat treatment temperature (HTT) decreases. PAN-based carbon fibres appear to be suitable structures to allow significant sodium insertion. Thus, T650 ex-PAN fibres lead to a reversible capacity close to 209 mAh g⁻¹. In that case, sodium insertion occurs in two main ways: one is the adsorption on the single graphene layers and the other is the concomitant insertion into the porosity that occurs below 0.1 V versus Na⁺/Na. This second mechanism, which is indicated by a low-voltage plateau on the electrochemical curves, allows significant insertion. The compared electrochemical study of two saccharose-coke samples corresponding to different regions of Dahn's classification underlines the importance of the carbon precursor and of the manufacture process. The reversible capacity is equal to 184 mAh g⁻¹ for the sample heat treated at 800 °C which presents a high hydrogen content whereas it is close to 145 mAh g⁻¹ for the one characterized by a HTT of about 1500 °C and a low hydrogen content. The best electrochemical performances are obtained for pyrolyzed cellulose carbons. Indeed, the reversible capacity is about 279 mAh g⁻¹. Outgassing these carbons at 950 °C results in such a decrease of the reversible capacity down to 145 mAh g⁻¹. That can be related either to the thermal elimination of heteroelements or to modifications of the pore size distribution. Consequently, the most suitable hard carbon material for anodic applications in rechargeable sodium-ion batteries should both present a high residual hydrogen content and a significant microporosity.     

铁电极氰化物电镀过程中的活化现象

采用循环伏安法和恒电流法研究了铁电极在氰化物电解液中的行为,证实了在阴极过程中电极首先极化至铁表面钝态层的还原,实现了表面的活化;随后极化至金属在活化的铁基体上的电沉积过程。     

Methane activation over La2xY2-2xO3 (0 ⩽x ⩽ 1) catalysts

The influence of the structure on the catalytic methane activation coupling activity has been shown for La_2x Y_2–2x O₃ (0 x 1) compounds. A perovskite structure leads to total oxidation (CO, CO₂) whereas a bixbyite structure results in C₂ hydrocarbon formation. This change in selectivity is interpreted in terms of the difference in oxygen environment of lanthanum in the two stuctures.     

固体超强酸S_2O_8~(2-)/TiO_2-Al_2O_3催化合成乙酸戊酯的动力学研究

研究了以固体超强酸S2O82-/TiO2-A l2O3为催化剂合成乙酸戊酯的反应,在反应温度分别为100,108,116℃下,测出合成乙酸戊酯的动力学方程参数,建立了动力学方程式,并与无催化剂酯化反应的活化能及动力学方程式进行了比较。结果表明,最佳合成条件为:醇酸摩尔比1.3∶1,催化剂用量1 g,带水剂用量10 mL。固体超强酸S2O82-/TiO2-A l2O3使反应活化能明显降低,是合成乙酸戊酯的有效催化剂。     

A study on structure and electrochemical properties of (La, Ce, Pr, Nd)2MgNi9 hydrogen storage electrode alloys

(La, Ce, Pr, Nd)₂MgNi₉ hydrogen storage alloys were prepared through induction melting followed by a long annealing treatment. The structure and electrochemical properties of annealed alloys have been investigated by orthogonal design experiments. Both the individual effects of each substituting element and their interaction in alloys were studied systemically. It has been shown that the structure of main phase in alloys belongs to PuNi₃-type with a space group R-3m. Substituting rare-earth elements have a significant effect on both the phase structure of alloys and microstructure. The anisotropic change in the crystal structure of alloys can cause the acceleration of pulverization of alloy particles and result in the deterioration of the cyclic stability of alloy electrodes. Misch metals can raise the plateau pressure of hydrogen absorption/desorption. The discharge capacity of alloy ranges from 342.97 to 380.68 mAh g⁻¹ depending on the sort and content of substituting elements. Both cerium and neodymium can obviously reduce the discharge capacity of alloy electrodes. When compared to the La₂MgNi₉ alloy electrode, mish metals can significantly improve the high rate dischargeability of alloy electrodes. The improvement of the kinetic characteristic of alloy electrodes mainly results from the increase of the hydrogen diffusion rate in alloy bulk.     

双电层电容器用沥青焦基活性炭的制备工艺条件对其孔结构及电化学性能的影响

以沥青焦为原料，KOH为活化剂在不同的工艺条件下制备了双层电容器用活性炭电极材料。分别考察了活化剂用量、活化时间、以及加入Cu、Ni催化活化等工艺条件对活性炭孔结构及作为双电层电容器电极的电化学性能的影响。结果表明：在实验范围内增加KOH用量及活化时间，活性炭的比表面积和比电容增加，比电容最高达到247F／g。添加Cu、Ni催化活化后活性炭的比表面积及比电容增加，高功率放电性能明显改善。     

基于马尔可夫链的自适应性神经网络训练算法

提出一种基于马尔可夫链的自适应性神经网络训练方法,对传统的S型激励函数进行了改进,建立了自适应性的神经网络分类器。在假设样本中噪声服从于正态分布的情况下建立最大似然估计,通过后验概率建立马尔科夫链对样本进行训练,提高了神经网络训练速度。在轴承故障诊断中的测试结果表明,该算法可以迅速稳定地训练出神经网络,有效提高诊断的分类结果。