基于RF-LSTM的鸡舍恶臭气体预测研究

以鸡舍氨气为研究对象,对鸡舍氨气预测模型进行了研究.首先,利用随机森林算法（RF）对影响鸡舍氨气浓度的环境变量进行重要性排序,选取温度、湿度、光照、气象温度、降雨量作为模型的输入变量;在此基础上,构建了基于长短时记忆神经网络（LSTM）的鸡舍氨气浓度预测模型,并将提出的预测模型应用于江苏省宜兴市某养鸡场的氨气浓度预测中,并与LSTM模型、RF-Elman模型和RF-BP模型进行了对比实验,结果表明,基于RF-LSTM模型的预测效果最好,其平均绝对误差（MAE）、平均绝对百分误差（MAPE）和均方根误差（RMSE）分别为0.9183、4.9637%和1.4262;同时,为了验证该模型的性能,本文还实现了不同时间尺度的鸡舍氨气浓度预测,提前2h、3h、4h、5h氨气预测的平均绝对误差（MAE）分别为1.6218、2.1991、2.8553和3.0677.本文提出的预测模型提高了鸡舍氨气浓度的预测精度,可为减少鸡舍恶臭气体排放提供科学依据.     

乡村文化记忆资源的“文—旅”协同评价模型与应用——以苏州金庭镇为例

乡村文化记忆资源是乡村“文—旅”协同的链接点。基于文化记忆理论从文化信息层、物质载体层与社会行为层解析其内涵和协同框架,构建“文—旅”空间协调度、文化记忆连续性和旅游市场认同度的协同评价模型,以苏州金庭镇为例进行实证分析。结果表明:（1）时间格局上,旅游景点与生活生产类空间关联性不断提升,与功能象征类空间依赖减弱,与社会表征类和精神意识类空间关系有所降低但维持高度耦合水平。（2）空间格局上,与生活生产类空间呈极核带状格局,与其余三类空间呈现一核多极格局。（3）旅游景点与生活生产类空间协同关系从“文—旅”空间协调度、文化记忆连续性转向旅游市场认同度为主导,其余空间以“文—旅”空间协调度为主导。（4）影响协同关系的机制包括记忆延续与文化交融、空间优化与旅游融合和社会支持与政策保障等。研究结论可为乡村旅游地的文化振兴和文旅融合提供理论指导和应用模式。     

Nonstationary, long memory and anti-persistence in several climatological time series data

In this article we examine the stochastic behaviour of several daily datasets describing sun (total irradiance at the top of the atmosphere and sunspot numbers) and various climatological anomaly series by looking at their orders of integration. We use a testing procedure that permits us to consider fractional degrees of integration. The tests are valid under general forms of serial correlation and deterministic trends and do not require estimation of the fractional differencing parameter. Results show that the series are all nonstationary, with increments that might be stationary for those variables affecting sun, and anti-persistent for those affecting air temperatures.     

合成水滑石治理水体阴离子染料污染研究

利用不同层电荷的合成水滑石（层状双氢氧化物）和它们的焙烧态为吸附剂，系统考察了此类材料对于甲基橙染料的吸附保留行为，通过粉末X射线衍射，傅立叶红外和扫描电镜测试说明经水热处理得到的合成水滑石具有良好的结晶度和相纯度．水滑石及其焙烧态对于甲基橙阴离子染料的吸附研究主要采用间歇平衡法实验，吸附体系的电导率变化显示了水滑石的记忆效应，并有可能是焙烧态水滑石再生的动态过程反应低浓度甲基橙在焙烧态水滑石上的吸附实验说明，水滑石层电荷密度对于其焙烧态氧化物吸附低浓度阴离子具有显著的影响。     

鼓风机叶轮快速断裂的检测及失效分析

针对焦化厂鼓风机叶轮运行近15 d后产生穿透性裂纹、并延伸至叶轮半径2/3处的严重断裂事故,采用具有早期缺陷诊断功能的金属磁记忆新技术及其他多种手段,从宏观力学及微观机理角度对其进行综合性能评定及失效分析.研究结果表明,叶轮材料强度、韧性均过低,且韧性指标分散性大、同类型试样的韧性指标最大值与最小值之间相差3倍多;金相组织晶粒粗大,断口表现为脆性开裂.沿叶轮的断裂与非断裂面分别进行磁记忆扫描,磁场强度分析表明,断裂面内存在两处具有明显应力集中的危险源.分析表明,锻造及锻后热处理不良造成了快速疲劳断裂.最后,结合5年中3次叶轮断裂事故,对失效原因进行了相关性分析.     

Integrated biological responses of zebrafish (Danio rerio) to analyze water quality in regions under anthropogenic influence

This study analyzed water quality in regions around Patos lagoon (Southern Brazil) that are under anthropogenic pressure. Water samples were collected from five different sites, including one used as a source for human consumption (COR) and others known to be influenced by human activities (IP). Danio rerio (Teleostei, Cyprinidae) organisms were exposed for 24 h to these water samples, plus a control group. It was observed that: (1) reactive oxygen species levels were lower in COR and IP than in the control group; (2) glutamate-cysteine ligase (catalytic subunit) expression was higher in COR than in other sites; (3) exposure to all water samples affected long-term memory (LTM) when compared to control group. Thus, some water samples possess the ability to modulate the antioxidant system and to induce a decline in cognitive functions, as measured by LTM. The obtained results indicate that a combination of variables of different organization level (molecular, biochemical and behavioral) can be employed to analyze water quality in impacted regions.     

α7nAChR在氯化甲基汞神经毒性中的作用

为探讨环境染污染物氯化甲基汞(methylmercury chloride,Me Hg Cl)的神经毒性作用机制,采用MTT、免疫细胞荧光、dot Blot、qRT-PCR等实验技术检测Me Hg Cl对PC12细胞活性及其α7亚型烟碱型胆碱能受体(α7 nicotinic acetylcholine receptor,α7n AChR)蛋白和mRNA表达水平的影响。结果显示Me Hg Cl抑制PC12细胞活性,显著降低α7n AChR的蛋白和mRNA的表达水平,呈浓度依赖性。上述结果表明,Me Hg Cl对神经细胞毒性作用可能与抑制α7n AChR表达有关。     

基于BERT-BILSTM-CRF模型的电力行业事故文本智能分析*

为解决电力行业事故报告文本较长、语义复杂,难以进行有效文本识别问题,提出1种以BERT作为底层的预训练模型,并设计1种双重注意力机制编码器,结合BILSTM-CRF深度挖掘事故文本语义特征,从而实现文本智能分析。首先构建电力词典,通过对BERT预训练,进行BIO标注,然后引入BILSTM-CRF模型实现对文本标签智能分类,最后将该模型与现行其他4种深度学习模型进行对比。研究结果表明:该模型智能识别精确率、召回率及F1值(查准率)均达到约97%,较其他4种模型中效果最好的模型分别提高0.02,0.03,0.02。研究结果可为电力行业事故报告文本分析提供1种新思路。     

基于缓冲区重采样的LSTM林火预测模型*

为提高林火风险预测精度,挖掘地图上隐含的空间信息、时间序列上隐含的长期趋势和循环波动,提出1种基于缓冲区重采样的长短期记忆（LSTM）林火预测模型,选取15个与林火相关的影响因素,以方差膨胀因子为评价指标对其进行多重共线性检验,方差膨胀因子大于10的因素具有共线性,并采用信息增益率验证筛选结果的合理性。考虑到火灾的空间聚集特性,采用缓冲区分析与过采样相结合方法减少样本不均衡现象的影响,最终得到176 732条样本。对12个影响因素和研究时间段的火点建立LSTM预测模型,对森林火灾发生风险进行预测。研究结果表明:基于缓冲区重采样的LSTM林火预测模型有效考虑时空上隐含的信息,预测模型准确率为87.06%,特异性为97.99%,敏感度为76.12%,阳性预测率为97.43%,阴性预测率为80.41%,ROC曲线与AUC值均优于随机森林（RF）和支持向量机（SVM）这2种基准算法。维尔克松秩和检验发现,本文提出的模型与基准算法结果具有显著性差异。研究结果可为提高林火风险预测精度提供参考。     

Hunting in unfamiliar space: echolocation in the Indian false vampire bat, Megaderma lyra, when gleaning prey

The literature suggests that in familiar laboratory settings, Indian false vampire bats (Megaderma lyra, family Megadermatidae) locate terrestrial prey with and without emitting echolocation calls in the dark and cease echolocating when simulated moonlit conditions presumably allow the use of vision. More recent laboratory-based research suggests that M. lyra uses echolocation throughout attacks but at emission rates much lower than those of other gleaning bats. We present data from wild-caught bats hunting for and capturing prey in unfamiliar conditions mimicking natural situations. By varying light level and substrate complexity we demonstrated that hunting M. lyra always emit echolocation calls and that emission patterns are the same regardless of light/substrate condition and similar to those of other wild-caught gleaning bats. Therefore, echoic information appears necessary for this species when hunting in unfamiliar situations, while, in the context of past research, echolocation may be supplanted by vision, spatial memory or both in familiar spaces.Communicated by T. Czeschlik