求解复杂优化问题的双层多种群粒子群优化算法

为解决粒子群优化算法存在的易早熟和精度低问题,提出了一种双层多种群粒子群优化算法.此算法采用上下两层,即下层N个基础种群和上层一个精英种群.各个基础种群相互独立进化,并从精英种群中得到优良信息指导自己的进化.上层精英种群首先通过接受各基础种群的当前最优粒子来更新自己的粒子集合,然后执行自适应变异操作,最后随机地向每一个基础种群输送出本次进化后的一个最优粒子来改进其下一轮搜索.该算法的并行双进化机制增加了群体的随机性和多样性,提高了全局搜索能力和收敛精度.实例仿真表明该算法具有较好的性能,尤其对于复杂多峰函数优化,成功率显著提高.     

基于MPSO的有限缓冲区多产品厂间歇调度问题的研究

研究了以最小化最大完工时间为目标的有限缓冲区多产品厂间歇调度问题,提出了一种基于多种群粒子群优化(MPSO)的间歇调度算法.该算法采用多种群,增加了种群初始粒子的多样性,在每一代子种群并行进化的过程中引入移民粒子,使子种群之间相互影响和促进,避免算法过早地陷入局部最优,提高了算法的全局搜索能力;每代进化后选出子种群中的优秀粒子作为精华种群,并对其进行变邻域搜索(VNS),进一步提高了算法的收敛精度.通过对不同规模调度问题的仿真,以及与其它算法的对比,证明了该算法解决有限缓冲区多产品厂间歇调度问题的有效性和优越性.     

基于遗传算法与粒子群优化法的机载合成孔径雷达天线赋形波束综合

采用了遗传算法和粒子群优化法两种现代进化算法对赋形波束天线进行了综合,得到了很好的结果,并比较了两种算法的优缺点.首先通过对宽测绘带的机载合成孔径雷达距离向回波特性的分析,给出了天线距离向赋形辐射方向图表达式.接着利用两种现代优化算法结合谢昆若夫法对天线阵分别进行综合,得到了近似相等的天线各单元激励幅度和相位.然后选择5个单元激励并利用微带贴片天线阵列实现了赋形波束方向图,实验结果和理论优化结果基本吻合,从而证明了这两种现代进化算法对赋形天线方向图综合方法的有效性.     

基于速度交流的多种群多目标粒子群算法研究

为提高多目标优化算法的收敛精度和搜索性能,提出一种基于速度交流的多种群多目标粒子群算法。算法引入速度交流机制,将种群划分为多个子种群以实现速度信息共享,改善粒子单一搜索模式,提高算法的全局搜索能力。采用混沌映射优化惯性权重,提高粒子搜索遍历性和全局性,为降低算法在运行后期陷入局部最优Pareto前沿的可能性,对各个子种群执行不同的变异操作。将算法与NSGA-Ⅱ、SPEA2、AbYSS、MOPSO、SMPSO和GWASF-GA先进多目标优化算法进行对比,实验结果表明:该算法得到的解集具有更好的收敛性和分布性。     

基于绿色生产的混合流水车间调度问题研究

针对绿色可持续发展问题,通过量化绿色指标评价方法,构建最小化最大完工时间、碳排放和噪声的多目标混合流水车间调度模型,并提出一种混合离散多目标帝国竞争算法(hybrid discrete multi-objective imperial competition algorithm,HDMICA)对模型进行求解。采用基于混沌反向学习策略的种群初始化方式提高初始化种群的多样性;基于本文模型设计3种有效的局部搜索策略以提升算法局部搜索能力;通过实验验证所提算法的有效性及优越性。     

基于速度交流的共生多种群粒子群算法

为了提高粒子群算法的求解精度,改善算法的搜索性能,提出一种基于速度交流的共生多种群粒子群算法(SMPSO)。该算法采用速度交流机制划分整个从种群为多个子种群,负责解空间的全局搜索,将获得的最优信息分享给主种群;主种群综合从种群与自身最优经验,负责局部深度优化,获得最优信息反馈给从种群,从而建立主从群间的共生关系,实现解空间的充分搜索。迭代后期,在主种群中引入自适应变异策略,提高算法跳出局部最优的能力。将提出的SMPSO算法应用于基准测试函数中,与其它改进的PSO算法进行比较。实验结果表明,SMPSO算法在求解精度、搜索能力、稳定性等方面均有较大的提高。     

“一刀切”排样问题的遗传算法与蚁群算法整合研究

对于"一刀切"矩形件优化排样问题,采用遗传算法与蚁群算法的混合算法进行研究.针对两种算法的传统混合策略和现有混合策略的不足,对两种算法的混合策略进行改进,并利用种群本身的染色体适值来判断种群进化是否停滞,确定了算法的最佳融合时机.对具体算例的分析验证表明,改进后的混合策略可有效减少算法的冗余迭代次数,提高搜索速度,是一种行之有效的排样算法.     

基于改进反向差分进化算法的多股簧响应模型参数辨识

针对传统算法在辨识多股簧响应模型参数方面存在的不足,提出一种改进反向差分进化算法。改进的算法采用新的反向学习机制引导种群逼近全局最优解,并使用正弦混沌序列计算缩放因子以提高种群的多样性。这两种机制协同操作可以较好地平衡算法的全局勘探和局部开采能力。通过建立参数辨识的目标函数和进行多股簧的动态试验,然后使用改进算法优化目标函数得到辨识结果。计算结果表明,改进算法能够有效地辨识多股簧模型参数,参数的收敛速度和计算的成功率优于标准反向差分进化算法和其他算法;即使在噪声级别较高的情况下,改进反向差分进化算法也可以准确地求出多股簧的模型参数。     

基于改进遗传算法的海底地层参数反演计算

为了提高对海底地层参数变量的反演计算能力,设计了一种基于双种群协同进化策略的改进遗传算法。针对标准遗传算法局部搜索能力差且易于出现早熟现象的缺点,在标准遗传算法基础上引入双种群同时进行优化搜索,两个种群分别给予不同的控制参数,实现协同进化,最终给出一个综合的最优解。通过两个算例对遗传算法的寻优能力进行测试,实验结果表明,提出的改进算法不仅提高了搜索性能,并且对遗传控制参数的依靠度大大降低,特别是对大型复合参数反演问题的求解计算更为有效。     

基于多变异策略与拥挤积距的多目标优化算法

为了进一步提升进化迭代中的群体多样性和分布性,提出一种基于多变异策略与拥挤积距的多目标优化算法(mcMOA)。该算法设计模糊记忆变异算子采集和利用进化中成功变异的尺度信息,以引导后续变异,增强了局部搜索效率。算法采用多变异策略,将模糊记忆变异、多项式变异、非一致性变异3种变异方式有机融入整个进化周期,提升了进化种群的多样性和全局搜索效率。针对拥挤距离不能有效表达个体局部分布性的问题,算法采用个体与相邻个体之间的距离乘积来替代拥挤距离,拥挤积距能有效表示个体的局部密度和局部分布性。通过标准测试函数的仿真实验并与多个采用单变异策略的经典算法比较,新算法在收敛性和分布性方面表现更优。     

质子交换膜燃料电池膜电极的结构优化

膜电极(membrane electrode assembly,MEA)是质子交换膜燃料电池(proton exchange membrane fuel cell,PEMFC)的核心部件,为PEMFC提供了多相物质传递的微通道和电化学反应场所。为了实现燃料电池商业化目标,需要制备高功率密度、低Pt载量、耐久性好的MEA。在MEA中除了催化剂以外,各功能层结构、层与层之间的界面都对MEA的性能具有重要影响。传统方法(CCS法和CCM法)制备的MEA在结构上有很多缺陷,明显制约了Pt的利用率和系统传质能力。通过优化各功能层结构消除缺陷,将有利于进一步提升PEMFC综合性能。本文从传统MEA结构存在的问题出发,梳理了近年来关于催化层、质子交换膜和气体扩散层结构优化方面的文献,归纳总结了各先进结构的制备方法、构效关系以及优缺点,对未来高性能、低成本和长寿命的MEA的开发具有指导意义。     

Facile patterning of reduced graphene oxide film into microelectrode array for highly sensitive sensing

In this study, we develop a new technique to fabricate a reduced graphene oxide (rGO)-based microelectrode array (MEA) with low-cost soft lithography. To prepare patterned rGO, a polydimethylsiloxane (PDMS) mold with an array of microwells on its surface is fabricated using soft lithography, and GO is assembled on an indium tin oxide (ITO) electrode with a layer-by-layer method. The rGO pattern is formed by closely contacting the assembled GO film onto the ITO electrode with the PDMS mold filled with hydrazine solution in the microwells to selectively reduce the localized GO into the rGO. The MEA with patterned rGO as the microelectrode is characterized with Kelvin probe force microscopy (KFM), atomic force microscopy (AFM), and cyclic voltammetry (CV) with ferricyanide in aqueous solution as the redox probe. The KFM and AFM results demonstrate that each rGO pattern prepared under the present conditions is 3 μm in diameter, which is close to that of the PDMS mold we use. The CV results show that the rGO patterned onto the ITO exhibits a sigmoid-shaped voltammogram up to 200 mVs(-1) with a microampere level current response, suggesting that the rGO-based electrode fabricated with soft lithography behalves like a MEA. To demonstrate the potential electroanalytical application of the rGO-based MEA, prussian blue (PB) is electrodeposited onto the rGO-based MEA to form the PB/rGO-based MEA. Electrochemical studies on the formed PB/rGO-based MEA reveal that MEA shows a lower detection limit and a larger current density for the detection of H(2)O(2), as compared with the macroscopic rGO electrode. The method demonstrated here provides a simple and low-cost strategy for the fabrication of graphene-based MEA that are useful for electroanalytical applications.     

Pulsed Activation of a Fuel Cell on the Basis of a Proton-Exchange Polymer Membrane

The influence of pulsed activation on the electrochemical characteristics of fuel cells based on the Nafion proton-conducting membrane is studied. The activation was a repeated pulsed operation of the membrane–electrode assembly (MEA), which was characterized by alternating a constant load of the MEA near a short-circuit current at voltage U_L = 0.1 V and its “idle time” corresponding to the open circuit voltage of the MEA, U_OCV ≈ 0.95 V. It is shown that there is an optimal ratio of the time parameters of the pulsed load at which the activation is much more efficient than in the case of constant potentiostatic or galvanostatic regimes often used in practice.     

Achieving excellent strength-ductility synergy in twinned NiCoCr medium-entropy alloy via Al/Ta co-doping

Alloying is an effective strategy to tailor microstructure and mechanical properties of metallic materials to overcome the strength-ductility trade-offdilemma.In this work,we combined a novel alloy design prin-ciple,i.e.harvesting pronounced solid solution hardening (SSH) based on the misfit volumes engineering,and simultaneously,architecting the ductile matrix based on the valence electron concentrations (VEC)criterion,to fulfill an excellent strength-ductility synergy for the newly emerging high/medium-entropy alloys (HEAs/MEAs).Based on this strategy,Al/Ta co-doping within NiCoCr MEA leads to an efficient synthetic approach,that is minor Al/Ta co-doping not only renders significantly enhanced strength with notable SSH effect and ultrahigh strain-hardening capability,but also sharply refines grains and induces abnormal twinning behaviors of (NiCoCr)92Al6Ta2 MEA.Compared with the partially twinned NiCoCr MEA,the yield strength (σy) and ultimate tensile strength (σUTS) of fully twinned Al/Ta-containing MEA were increased by ～102 ％ to ～600 MPa and ～35 ％ to ～ 1000 MPa,respectively,along with good ductility beyond 50 ％.Different from the NiCoCr MEA with deformation twins (DTs)/stacking faults (SFs) domi-nated plasticity,the extraordinary strain-hardening capability of the solute-hardened (NiCoCr)92Al6Ta2 MEA,deactivated deformation twinning,originates from the high density of dislocation walls,micro-bands and abundance of SFs.The abnormal twinning behaviors,i.e.,prevalence of annealing twins (ATs)but absence of DTs in (NiCoCr)92Al6Ta2 MEA,are explained in terms of the relaxation of grain boundaries(for ATs) and the twinning mechanism transition (for DTs),respectively.     

Production new chelating fibers of polyacrilonitrile for removing metal ions

In this work, raw acrylic fibers (RAF) have been modified with ethanolamine to provide new chelating fibers and prepared polyacrylonitrile‐monoethanolamine (PAN‐MEA). PAN‐MEA fiber was prepared by conversion nitrile groups into hydroxyle groups using ethanolamine solution with different concentration under refluxing at different temperatures and different times. The fiber structure has been investigated by different experimental techniques of characterization such as Fourier transform infrared spectroscopy (FT‐IR), thermogravimetric analysis (TGA), scanning electron microscopy (SEM) and the physical and mechanical properties has also been investigated in this study. The PAN‐MEA was examined as chelating material in a series of batch adsorption experiments for removal Cu (II), Pb (II) and Ni (II) ions. The ion adsorption of PAN‐MEA was determined with the atomic adsorption method. PAN‐MEA show adsorption metal ions more than usual fiber because of the modification.     

水泥中MgO膨胀剂水化产生的膨胀应力表征

本文研制了一套膨胀应力测试装置,并以此为基础测试了约束条件下MgO膨胀剂(MEA)压实体和掺MgO膨胀剂水泥浆体中水化产生的膨胀应力;采用X射线衍射法定量分析了MgO的含量,并计算MgO的水化程度。结果表明:活性指数为23s和46s的MgO膨胀剂压实体水化22d产生的膨胀应力分别为38.9MPa和62.7MPa;掺水泥浆体的膨胀应力随MEA掺量增大、MEA活性指数提高和MgO水化程度增加而增大,不掺MgO膨胀剂的水泥浆体63d时的收缩应力为1.3MPa,掺12%活性指数为23s和46s的MgO膨胀剂水泥浆体在63d时的膨胀应力分别为1.8MPa和3.0MPa。     

The Ionic Resistance of a Nanostructured Membrane Electrode Assembly under Nonequilibrium Conditions

Technical Physics Letters - The ionic resistance of the membrane electrode assembly (MEA) of O2/H2 fuel cells is studied in situ in a broad range of MEA compositions by using electrochemical...     

高浓度CO2在MEA溶液中平衡溶解度的实验研究

采用气液搅拌平衡装置,通过酸碱滴定和皂泡流量计测定了高浓度CO2在不同MEA溶液中的平衡溶解度。利用正交实验设计方法,考察了温度、CO2分压力、MEA质量浓度等因素对CO2平衡溶解度的影响。结果表明,在实验参数选定范围内,影响MEA中CO2平衡溶解度的因素主次顺序为：CO2分压力〉MEA质量浓度〉温度;MEA溶液中CO2的平衡溶解度随压力的增大而增大,随MEA质量浓度的增大而减小,随温度的增大而减小。     

PEMFC膜电极电催化层的制备和性能

膜电极催化层的组成和电催化剂的活性对质子交换膜燃料电池的性能有很大影响．采用浸渍还原法制备出了Pt平均粒径为3．1nm的Pt／C催化剂．催化剂中Pt的粒径和在碳黑载体（VulcanXC-72）表面的分散程度采用透射电镜（TEM）进行测试．用Pt／C催化剂、适量的Nation溶液和PrFE乳液制备出质子交换膜燃料电池（PEMFc）膜电极的催化剂层,并研究了该催化剂层中PTFE含量对其性能的影响．实验表明,PTFE强烈的疏水性可以迫使部分水分向阳极反扩散,催化层中加入适量的PTFE可以使膜电极具有一定的水管理能力,在去掉辅助增湿系统的条件下具有良好的极化性能．     

Microelectrode Array-evaluation of Neurotoxic Effects of Magnesium as an Implantable Biomaterial

Magnesium(Mg)-based biomaterials have shown great potential in clinical applications. However,the cytotoxic effects of excessive Mg~(2+) and the corrosion products from Mg-based biomaterials,particularly their effects on neurons,have been little studied. Although viability tests are most commonly used,a functional evaluation is critically needed. Here,both methyl thiazolyl tetrazolium(MTT) and lactate dehydrogenase(LDH) assays were used to test the effect of Mg~(2+) and Mg-extract solution on neuronal viability.Microelectrode arrays(MEAs),which provide long-term,real-time recording of extracellular electrophysiological signals of in vitro neuronal networks,were used to test for toxic effects. The minimum effective concentrations(EC_(min)) of Mg~(2+) from the MTT and LDH assays were 3 mmol/L and 100 mmol/L,respectively,while the EC_(min) obtained from the MEA assay was 0.1 mmol/L. MEA data revealed significant loss of neuronal network activity when the culture was exposed to 25% Mg-extract solution,a concentration that did not affect neuronal viability. For evaluating the biocompatibility of Mg-based biomaterials with neurons,MEA electrophysiological testing is a more precise method than basic cell-viability testing.