Novel transparent and flexible nanocomposite film prepared from chrysotile nanofibres

In the present study, chrysotile nanofibres, obtained from physicochemical dispersion of natural chrysotile, were used to prepare nanofibre sheets by vacuum filtration. As-prepared sheets were then impregnated by UV-curable resin and cured by ultraviolet light to fabricate the flexible and transparent nanocomposite films. Observed from SEM, the transparent films showed a smooth surface and a typical sandwich structure in cross section, viz. nanofibre sheet filled with resin was sandwiched by two layers of resin. XRD patterns indicated the amorphous nature of cured resin and characteristic crystallographic structure of chrysotile in nanocomposite films. Though the nanofibre sheets were white in colour, and nanofibre contents in nanocomposites were as much as 43.4 wt%, the nanocomposite films displayed an excellent optical transparency with about 85% light transmittance in the visible light range. Tensile tests showed that the addition of nanofibres resulted in a great improvement in mechanical strength of the nanocomposite films; with the increase of nanofibre contents, the modulus and tensile strength of nanocomposite films increased gradually.     

Modulating lateral strain in GaN-based epitaxial layers by patterning sapphire substrates with aligned carbon nanotube films

A novel carbon nanotube-patterned sapphire substrate (CPSS) has been utilized for the growth of GaN material and fabrication of a InGaN/GaN light emitting diode (LED) by metal-organic vapor phase epitaxy. Different lateral strain distributions and stress reductions were observed in a GaN thin film on CPSS compared with those on a conventional sapphire substrate. Nanoheteroepitaxy induced by small size nucleation islands of about 50 nm is ascribed to this significant strain modulation. The crystalline quality of the GaN thin film was also improved, as illustrated by X-ray diffraction. Performances of 1 mm × 1 mm LEDs on CPSS were also enhanced, with an operational power increase of 37.5% and higher saturation current.      

Synthesis, characterization and osteoconductivity properties of bone fillers based on alendronate-loaded poly(ε-caprolactone)/hydroxyapatite microspheres

A superior drug controlled release system capable of achieving efficient osteogenesis is in imperative demand because of limited bone substitute tissue for the treatment of bone defect. In the present study, we investigated the potential of using poly(ε-caprolactone)–hydroxyapatite (PCL–HA) composite microspheres as an injectable bone repair vehicle by controlled release of alendronate (AL), a medicine that belongs to the bisphosphonates family. The PCL/HA–AL microspheres were prepared with solid/oil/water emulsion technique, which included two processes: (1) AL was loaded on the hydroxyapatite nanoparticles; (2) the HA–AL complex was built in the PCL matrix. The spherical PCL/HA–AL microspheres were characterized with its significantly improved encapsulation efficiency of hydrophilic AL and better sustained release. Human bone mesenchymal stem cells (hMSCs) were cultured on the surface of these microspheres and exhibited high proliferative profile. Specifically, in osteogenic medium, hMSCs on the surface of PCL/HA–AL microspheres displayed superior osteogenic differentiation which was verified by alkaline phosphatase activity assay. In conclusion, by presenting strong osteogenic commitment of hMSCs in vitro, the PCL/HA–AL microspheres have the potential to be used as an injectable vehicle for local therapy of bone defect.     

Study on the forecast method for underground coal mine

At present, coal mine fires were forecasted with some temperature, smog, CO, CO2, etc, however, this method can＇t meet the requirements for safe production of coal mines in monitoring accuracy and validity. Overcoming these problems of foregone monitoring methods, using multi-parameters which include fire image, smog, CO, CO2, O2, etc, the paper put forward a synthetical analysis monitor with advanced technology of neural network. The research and application of this method has significance in theory and practical value for coal mine fire forecast.     

显像管玻壳退火炉热平衡测试及评估

通过对4台玻壳退火炉的热平衡测试、计算及分析,得到了玻壳退火炉热平衡收支数据,计算出玻璃成品的单位质量产品耗热量、单位质量产品燃气耗量等主要技术经济指标。提出了进一步提高热效率的技术措施。     

基于神经网络的图像来源识别方法比较研究

随着手机等便携式智能电子设备的普及,图像已成为最重要的信息载体之一,在新闻、社交及司法等领域发挥着重要作用.在享用电子图像带来便捷性的同时,图像处理工具给不法分子通过篡改电子图像实施诈骗等犯罪活动提供了可能,识别图像来源、辨别图像真伪已成为遏制和惩罚此类犯罪活动的重要技术手段.该文讨论了神经网络在图像源识别中的应用方法...     

循环函数法在回热系统中的局部定量计算

根据等效焓降法在热力系统计算时的特点,推导了循环函数法在回热系统中的局部定量计算方法,导出了回热系统局部定量计算的几个重要结论.最后,通过实例计算对所得结论进行了验证.验证结果表明,该方法在回热系统局部定量计算过程中具有简洁、准确、概念清晰等特点,适合推广.     

H.264/AVC编解码器在TMS320C6400上的优化实现

分析了基于TMS320C6400平台的H.264/AVC编码器运算量在各个部分的分布,并且将其中运算量较大的部分进行线性汇编,分析了这些部分在进行线性汇编时需要注意的技术细节,能够提高线性汇编的并行程度.     

基于渐进式分解架构的风电时间序列预测

准确预测风电机组各项指标对准确管控机组和调控电网的供需有着重要意义. 预测指标任务可抽象为风电时间序列预测任务. 目前时间序列预测模型主要采用深度学习模型, 但是风电时间序列具有较强的波动性和随机性, 导致绝大部分模型不能较好挖掘风电时间序列的复杂演化特性. 为解决上述问题, 提出了一种基于渐进式分解架构的风电时间序列预测方法, 该方法首先应用神经网络池化分解方法将复杂的依赖关系简化并应用注意力机制学习长期趋势, 然后运用多变量融合捕捉模块增强了网络整体的多变量关联挖掘能力, 最后, 融合趋势项和周期项对风电时间序列做出准确的预测. 实验结果表明, 该方法在风电时间序列的多步预测中均方误差相比基线模型至高可提升24%, 在多尺度预测长度下表现出预测性能稳定提升的同时, 计算效率显著优于同类模型.