Electrocatalytic oxidation of methanol on platinum nanoparticles electrodeposited onto porous carbon substrates

To achieve methanol fuel cell electrodes allowing a high catalyst use, thin layers of various carbon powders and recast Nafion® were electrochemically plated with platinum. The resulting Pt deposits were characterized by hydrogen and carbon monoxide electrosorption, as well as by transmission electron microscopy. Methanol oxidation was then carried out using cyclic voltammetry. The influence of the amount of carbon surface oxides and the effect of Pt specific surface area on the Pt catalytic activity were thus investigated.     

Influence of doctor blade gap on the properties of tape cast NiO/YSZ anode supports for solid oxide fuel cells

In this study, various tape cast NiO/YSZ anode support layers with similar geometric properties are fabricated by varying the doctor blade from 100?µm to 200?µm with an increment of 25?µm. The mechanical properties of the anode support layers are investigated by three point bending tests of 30 samples for each doctor blade gap. The reliability curves of the flexural strength data are also obtained via two-parameter Weibull distribution method. The effects of the doctor blade gap on the microstructure and the electrochemical performance of the anode support layers are determined via SEM investigations and single cell performance-impedance tests, respectively. The apparent porosities of the samples are also measured by Archimedes’ principle. The results indicate that the doctor blade gap or the resultant tape thickness influences the microstructure of tape cast NiO/YSZ anode supports significantly, yielding different mechanical and electrochemical characteristics. At a reliability level of 70%, the highest flexural strength of 110.20?MPa is obtained from the anode support layer with a doctor blade gap of 175?µm and the 16?cm² active area cell with this anode support layer also exhibits the highest peak performance of 0.483?W/cm² at an operating temperature of 800?°C. Thus, a doctor blade gap of 175?µm is found to have such a microstructure that provides not only better mechanical strength but also higher electrochemical performance.     

基于模糊核聚类的乙烯裂解深度DE-LSSVM多模型建模

乙烯裂解深度的建模与控制对于裂解炉的实时优化具有重要意义。针对石脑油原料组分复杂、油品特性波动大等状况,采用模糊核聚类对石脑油数据库进行最优划分,建立最小二乘支持向量机的多模型,对于最小二乘支持向量机中模型的参数选取,利用差分进化算法进行参数寻优,提高了模型的精度和泛化能力。通过对现场数据的建模实验,结果表明：基于模糊核聚类的乙烯裂解深度最小二乘支持向量机多模型跟踪性能良好,预测精度较高。     

京沪高速铁路聚脲防护工程要点剖析

“喷涂聚脲弹性体”（SprayPolyureaElastomer）技术以其优异的综合性能和先进的施工工艺应用于京沪高速铁路桥梁混凝土防护。本文对已投入运营的京沪高速铁路聚脲防护工程聚脲防护技术的施工工艺、聚脲材料性能指标等进行了较为全面的分析;对京沪高速铁路聚脲防护技术遇到的问题进行了深刻剖析,并提出富有成效的改进措施;详细叙述了青岛理工大学功能材料研究所在制定技术文件、基材处理工艺、冬季野外施工以及涂层低温脱落等问题上,针对京沪高铁聚脲防护工程做出的大量富有指导意义的研究工作,确保了京沪高铁聚脲防护工程的顺利进行。     

工业管道中非金属管道的设计

简述了非金属管道的分类以及与金属管道相比较的优缺点,根据自身工程经验,介绍了非金属管道的连接方法、管道布置和支吊架选择的一般要求。     

载体对Co-Cu催化剂催化CH4-合成气梯阶转化反应的影响

以钴、铜为活性金属,采用等体积浸渍法制备Co-Cu双金属催化剂,考察了不同载体的Co-Cu双金属催化剂对CH4-合成气梯阶转化直接合成C2+含氧化合物的影响,并进行了XRD、NH3-TPD和H2-TPR表征及相应的分析。实验结果表明,载体对催化剂催化性能有较大的影响,不同载体的催化剂对CH4-合成气两步梯阶转化反应的活性高低顺序为Co-Cu/TiO2＞Co-Cu/Al2O3＞Co-Cu/SiO2,Co-Cu/TiO2催化剂对C2+含氧化合物的选择性为79.9%。这归因于TiO2载体与活性金属Co、Cu之间适宜的相互作用及载体对Co-Cu良好的分散性能。此外,Co-Cu/TiO2催化剂还具有反应所需的表面酸量及适度的中强酸度。     

基于改进聚类和加权bagging的多模型软测量建模

针对化工生产过程中软测量模型估计精度的问题,提出一种基于改进聚类和加权bagging的多模型建模方法。该方法在传统FCM聚类的基础上,利用K-近邻处理进一步降低错分率,改善聚类效果;通过相关性分析对训练样本集进行特征分组,将原始集划分为多个特征集;最后根据加权bagging的集成学习算法,融合支持向量机自适应地实现多模型建模。仿真结果表明,该建模方法可以合理地加权分配特征子模型,使得模型估计精度得到提高,具有更强的泛化能力。     

生物反应过程的在线支持向量机建模优化

针对生物反应过程具有较强的非线性、时变性,建立准确的机理模型较为困难,并且复杂的机理模型也无法用于在线控制和优化。将在线支持向量机和机理模型结合,提出串并联在线自校正混合建模方法。通过对典型生化过程谷氨酸的生产过程分析,找到影响谷氨酸浓度的关键参数;从现场历史数据中选取样本,建立基于在线向量机的软测量模型。实验结果表明该模型对谷氨酸浓度预测效果较好。     

基于加权LS-SVM的青霉素发酵过程建模

青霉素发酵过程中,一些重要参数的检测存在一定的误差,给生产过程的监测及重要参数的实时监控等带来一定困难。样本数据中自变量、因变量均有可能包含误差数据,影响模型建立的准确性,本文采用加权最小二乘算法,给各个样本的误差平方赋予不同权重用于克服异常训练样本的影响,利用Pensim仿真平台数据,采用粒子群算法(PSO)对加权最小二乘向量机算法(WLS-SVM)的参数寻优,建立青霉素发酵过程模型,通过仿真实验表明了该算法用于青霉素发酵过程建模的有效性。