活性炭吸附甲基橙模拟废水预测模型研究

分别采用BP人工神经网络算法及多元线性回归法,以实验所得的36组数据为样本,建立了以吸附时间、活性炭投加量及甲基橙废水浓度为输入变量,以活性炭吸附处理后甲基橙溶液的吸光度为输出变量的吸附预测模型,并进行了两模型预测效果的对比。结果表明,BP神经网络模型获得了比多元线性回归更好的拟合预测效果。使用BP神经网络模型可以实现同时考虑三个操作因素条件下活性炭吸附特性的预测,而且预测结果与实验数据吻合度较高,其预测样本最大和最小相对偏差分别为2.92%和0.029%,残差绝对值小于0.050 5。     

神经网络预测磁絮凝处理矿井水效果的研究

构建了磁絮凝处理矿井水实验装置,并使用RS485接口经串口协议与计算机MFC控制程序交换数据。根据实验装置所得数据,采用C++调用Matlab Engine实现GA-BP和GRNN神经网络的训练、预测与准确度的分析。由分析可知,GA-BP神经网络初始化参数较多,预测结果平均绝对误差为21.849,个体适应度在进化次数为35时趋于平稳,适应度达到82%。GRNN神经网络初始化仅需要调整光滑因子,当光滑因子取0.75时,预测平均绝对误差为10.726。通过比较GA-BP与GRNN神经网络的R~2和RMSE的数值,可知GRNN准确性更高,在磁絮凝处理含悬浮物矿井水中适用性更强。     

BP神经网络在膜分离处理废水中的应用

为能更好地预测工艺条件对膜分离过程的影响,运用BP神经网络技术建立输入变量为压差、流速、浓度和温度,输出变量为膜通量的预测模型。通过大量实验数据训练预测模型,得到的网络模型整体误差平方和仅为0.014 5;计算值与模拟值相比,10组不同条件的膜通量平均预测误差仅为1.1,证实了所建立的BP神经网络膜通量预测模型与实验值吻合程度较好,有较好地预测能力。在此基础上进一步考察了工艺参数对膜分离过程的影响。     

室内环境中细木工板甲醛释放预测模型

采用环境测试舱模拟室内环境,测定其中的细木工板的甲醛释放浓度,考察环境温度和相对湿度对其释放的影响;分析细木工板中甲醛气体扩散机理,并进行实际室内环境中细木工板释放甲醛实验,与模拟室内环境对比;最后运用灰色预测模型和神经网络模型建立灰色神经网络模型,对实际室内环境中细木工板甲醛释放规律进行预测. 结果表明,随环境温度和相对湿度升高,板材释放的甲醛浓度增加,且温度对甲醛释放活跃期影响更显著,低温和低湿度时板材中甲醛释放更易达到稳定;细木工板释放甲醛浓度与室内外温差呈正相关性,热压渗风作用对室内细木工板释放甲醛浓度的变化有重要影响;灰色神经网络模型的预测与实验数据吻合较好,平均绝对误差为-0.0007 mg/m3,相对误差为0.208%~5.981%.     

基于遗传BP神经网络预测硫在高含硫气体中溶解度

为更精确地关联预测硫在高含硫气体中的溶解度,提出将遗传算法(GA)和LM-反向传播神经网络(LM-BP ANN)相结合的预测模型。设计了该模型的计算过程,讨论了模型参数的设置。以温度、压力和气体组分作为BP神经网络预测模型的输入变量,利用GA优化了BP神经网络的初始权值和阈值,采用遗传算法优化后的BP神经网络计算了元素硫在高含硫气体中的溶解度。结果表明,该模型训练结果与实测值之间的平均相对误差为5.90%,测试结果与实测值的平均相对误差为5.54%;该方法较BP神经网络模型具有预测精度高、收敛速度快的优点;该模型具有较好的模拟及内推、外推功能。     

基于不同神经网络模型的橡胶减振器性能预测的应用研究

利用BP和GRNN神经网络建立胶料性能和减振器性能之间的预测模型,并对预测结果的误差进行对比分析。结果表明,利用BP神经网络建立的预测模型预测平均相对误差在16%以内,利用GRNN神经网络建立的预测模型预测平均相对误差在5%以内,说明使用GRNN神经网络进行减振器性能预测精度更高、效果更好。     

基于聚类分析与粒子群算法优化神经网络的灰熔点预测

针对解决燃煤锅炉或气化炉的结渣现象，影响锅炉安全性问题，以灰成分金属氧化物为自变量，灰熔点流动温度为因变量，建立了K-Means-PSO-BPNN的灰熔点预测模型，误差分析结果表明，经过粒子群算法优化，BP神经网络模型在聚类分析后的预测效果得到了显著提高，表现出更好的相关性，相关系数为0.967,高于未优化的0.917,平均绝对误差为5.81,小于未聚类的26.98,并且模型的准确性提高到98.89%。因此，聚类分析以及粒子群算法优化后的神经网络模型能够更准确预测煤灰的流动温度(FT)。     

基于遗传算法优化的BP神经网络气化用煤灰流动温度预测模型

采用遗传算法优化的BP神经网络建立煤灰流动温度预测模型，模型以灰成分及酸碱质量比、硅铝质量比等组合参数作为输入变量，以煤灰流动温度作为输出量，对126组来自中国北部地区的煤灰样数据进行训练与测试，并建立常规BP神经网络模型，研究了各输入变量对网络模型预测精度的影响并对比与常规BP神经网络模型的预测能力。结果表明：不同输入层变量的GA-BP神经网络模型对训练集和测试集样本数据都具有较好的学习和泛化能力，所有预测结果相对平均预测误差均不超过4%。酸碱质量比和硅铝质量比参数作为神经网络输入层的添加，虽略微提高模型对训练样本的拟合程度，但也导致验证时过拟合现象的发生，模型对新样本的拟合优度下降。采用SiO₂,Al₂O₃,Fe₂O₃,CaO,MgO和Na₂O+K₂O的质量分数6个参数作为输入变量的GA-BP模型最为适合，其对测试集数据的预测相对平均误差为3.45%,低于常规BP神经网络模型3.79%的误差。     

双向门控循环神经网络的SO_2排放浓度预测模型

火力发电站脱硫系统数据具有大惯性和延时性等特点,且影响SO_2排放浓度的因素众多。为此,建立了基于双向门控循环神经网络(biGRU)的SO_2排放浓度预测模型。以分析得到的主成分为输入变量,SO_2排放浓度为输出变量,通过训练对脱硫系统SO_2排放浓度数据进行预测,并进行比较。结果表明,与传统的RNN以及LSTM模型相比,biGRU模型能够获得较高的预测精度,其对称平均绝对百分比误差相较于RNN和LSTM分别下降了4.235%,0.718%,其均方根误差分别下降了1.942,0.443 mg/Nm³。该模型预测误差较低,泛化能力较好,具有较高的实际应用价值,有利于实现排放控制和节能减排。     

基于卷积神经网络的轮胎花纹噪声值预测

利用图像处理和卷积神经网络（CNN）搭建轮胎花纹结构与轮胎花纹噪声值之间的数学模型，分别采用CNN模型和BP神经网络对轮胎花纹噪声值进行预测，并对比预测精度。结果表明：采用CNN模型，轮胎花纹噪声值的预测值与实测值的平均绝对误差为0.591 dB，平均相对误差为0.81%；采用BP神经网络，轮胎花纹噪声值的预测值与实测值的平均绝对误差为0.713 dB，平均相对误差为0.95%；相较于BP神经网络，CNN模型对轮胎花纹噪声值的预测精度更高。     

Insect antifeedant properties of anthranoids from the genusVismia

Vismiones and ferruginins, representatives of a new class of lypophilic anthranoids from the genusVismia were found to inhibit feeding in larvae of species ofSpodoptera, Heliothis, and inLocusta migratoria.     

Direct observation of freeze-thaw instability of latex coatings via high pressure freezing and cryogenic SEM

Despite its industrial importance, the subject of freeze-thaw (F/T) stability of latex coatings has not been studied extensively. There is also a lack of fundamental understanding about the process and the mechanisms through which a coating becomes destabilized. High pressure (2100 bar) freezing fixes the state of water-suspended particles of polymer binder and inorganic pigments without the growth of ice crystals during freezing that produce artifacts in direct imaging scanning electron microscopy (SEM) of fracture surfaces of frozen coatings. We show that by incorporating copolymerizable functional monomers, it is possible to achieve F/T stability in polymer latexes and in low-VOC paints, as judged by the microstructures revealed by the cryogenic SEM technique. Particle coalescence as well as pigment segregation in F/T unstable systems are visualized. In order to achieve F/T stability in paints, latex particles must not flocculate and should provide protection to inorganic pigment and extender particles. Because of the unique capabilities of the cryogenic SEM, we are able to separate the effects of freezing and thawing, and study the influence of the rate of freezing and thawing on F/T stability. Destabilization can be caused by either freezing or thawing. A slow freezing process is more detrimental to F/T stability than a fast freezing process; the latter actually preserves suspension stability during freezing. Presented at the 82nd Annual Meeting of the Federation of Societies for Coatings Technology, October 27–29, 2004 in Chicago, IL. Tied for first place in The John A. Gordon Best Paper Competition.     

Ethanol and (−)-α-Pinene: Attractant Kairomones for Bark and Ambrosia Beetles in the Southeastern US

In 2002–2004, we examined the flight responses of 49 species of native and exotic bark and ambrosia beetles (Coleoptera: Scolytidae and Platypodidae) to traps baited with ethanol and/or (−)-α-pinene in the southeastern US. Eight field trials were conducted in mature pine stands in Alabama, Florida, Georgia, North Carolina, and South Carolina. Funnel traps baited with ethanol lures (release rate, about 0.6 g/day at 25–28°C) were attractive to ten species of ambrosia beetles (Ambrosiodmus tachygraphus, Anisandrus sayi, Dryoxylon onoharaensum, Monarthrum mali, Xyleborinus saxesenii, Xyleborus affinis, Xyleborus ferrugineus, Xylosandrus compactus, Xylosandrus crassiusculus, and Xylosandrus germanus) and two species of bark beetles (Cryptocarenus heveae and Hypothenemus sp.). Traps baited with (−)-α-pinene lures (release rate, 2–6 g/day at 25–28°C) were attractive to five bark beetle species (Dendroctonus terebrans, Hylastes porculus, Hylastes salebrosus, Hylastes tenuis, and Ips grandicollis) and one platypodid ambrosia beetle species (Myoplatypus flavicornis). Ethanol enhanced responses of some species (Xyleborus pubescens, H. porculus, H. salebrosus, H. tenuis, and Pityophthorus cariniceps) to traps baited with (−)-α-pinene in some locations. (−)-α-Pinene interrupted the response of some ambrosia beetle species to traps baited with ethanol, but only the response of D. onoharaensum was interrupted consistently at most locations. Of 23 species of ambrosia beetles captured in our field trials, nine were exotic and accounted for 70–97% of total catches of ambrosia beetles. Our results provide support for the continued use of separate traps baited with ethanol alone and ethanol with (−)-α-pinene to detect and monitor common bark and ambrosia beetles from the southeastern region of the US.     

Methyl 2,4,6-decatrienoates Attract Stink Bugs and Tachinid Parasitoids

Halyomorpha halys (Stål) (Pentatomidae), called the brown marmorated stink bug (BMSB), is a newly invasive species in the eastern USA that is rapidly spreading from the original point of establishment in Allentown, PA. In its native range, the BMSB is reportedly attracted to methyl (E,E,Z)-2,4,6-decatrienoate, the male-produced pheromone of another pentatomid common in eastern Asia, Plautia stali Scott. In North America, Thyanta spp. are the only pentatomids known to produce methyl 2,4,6-decatrienoate [the (E,Z,Z)-isomer] as part of their pheromones. Methyl 2,4,6-decatrienoates were field-tested in Maryland to monitor the spread of the BMSB and to explore the possibility that Thyanta spp. are an alternate host for parasitic tachinid flies that use stink bug pheromones as host-finding kairomones. Here we report the first captures of adult and nymph BMSBs in traps baited with methyl (E,E,Z)-2,4,6-decatrienoate in central Maryland and present data verifying that the tachinid, Euclytia flava (Townsend), exploits methyl (E,Z,Z)-2,4,6-decatrienoate as a kairomone. We also report the unexpected finding that various isomers of methyl 2,4,6-decatrienoate attract Acrosternum hilare (Say), although this bug apparently does not produce methyl decatrienoates. Other stink bugs and tachinids native to North America were also attracted to methyl 2,4,6-decatrienoates. These data indicate there are Heteroptera in North America in addition to Thyanta spp. that probably use methyl 2,4,6-decatrienoates as pheromones. The evidence that some pentatomids exploit the pheromones of other true bugs as kairomones to find food or to congregate as a passive defense against tachinid parasitism is discussed.     

2007～2008年世界塑料工业进展

收集了2007年7月～2008年6月世界塑料工业的相关资料,介绍了2007～2008年国外塑料工业的发展情况,提供了世界塑料产量、消费量及全球各类树脂的需求量及产能情况.按通用热塑性树脂(聚乙烯、聚丙烯、聚苯乙烯、聚氯乙烯、ABS树脂)、工程塑料(尼龙、聚碳酸酯、聚甲醛、热塑性聚酯、聚苯醚)、特种工程塑料(聚苯·硫醚、液晶聚合物、聚醚醚酮)、通用热固性树脂(酚醛、聚氨酯、不饱和聚酯树脂、环氧树脂)不同品种的顺序,对树脂的产量、消费量、供需状况及合成工艺、产品应用开发、树脂品种的延伸及应用的进一步扩展等技术作了详细介绍.     

2005～2006年国外塑料工业进展

收集了2005年7月～2006年6月国外塑料工业的相关资料,介绍了2005—2006年国外塑料工业的发展情况。提供了世界塑料产量、消费量及全球各类树脂生产量以及各国塑料制品的进出口情况。作为对比,介绍了中国塑料的生产情况。按通用热塑性树脂（聚乙烯、聚丙烯、聚苯乙烯、聚氯乙烯、ABS树脂）、工程塑料（聚酰胺、聚碳酸酯、聚甲醛、热塑性聚酯、聚苯醚）、通用热固性树脂（酚醛、聚氨酯、不饱和树脂、环氧树脂）、特种工程塑料（聚苯硫醚、液晶聚合物、聚醚醚酮）的品种顺序,对树脂的产量、消费量、供需状况及合成工艺、产品开发、树脂品种的延伸及应用的扩展作了详细的介绍。