In order to comprehensively sort out the differences in research on urban vitality between China and foreign countries,and identify the urgent problems and weak links in research on urban vitality in China,CiteSpace software is used to conduct visual analysis of literature from 1992 to 2023 based on annual publication volume,institutional co-occurrence,author co-occurrence,keyword co-occurrence,keyword clustering,and prominent words,targeting relevant research in CNKI and Web of Science databases.Results indicate that in recent years,the number of literature on urban vitality research at home and abroad has rapidly increased,entering a period of great development.Compared with domestic studies that focus more on quantitative measurement of urban space,economy,and socio-cultural vitality,foreign research tends to focus more on areas such as climate and environment,economy and resilience,life and quality,and cultural diversity.The hot research topics are more extensive and have more practical guidance value.As China's urbanization process enters a critical transition period,based on multi-source data,interdisciplinary integration,constantly enriching research objects,continuous innovation of research concepts,and increasingly diversified research methods will become the development trend of future research on urban vitality in China.
为确保水产品冷链物流供需双方信息对称,降低供应链中断风险及供需不匹配造成的浪费,水产品冷链物流需求预测显得尤为关键.选取影响水产品冷链物流需求的 18 个因素并用灰色关联法(GRA)筛选验证,运用主成分分析法(PCA)提取主要特征,通过粒子群算法(PSO)优化的径向基神经网络(RBF)构建PCA-PSO-RBF预测模型,对水产品需求预测,并与PCA-PSO-BP、PCA-RBF、PCA-BP、SVM、BP模型对比.结果表明,构建的PCA-PSO-RBF预测模型具有较强的非线性系统处理能力与全局寻优能力,对小样本多特征的数据具有较好包容性和预测精度,通过MAE/RMSE/MAPE预测误差评价验证了PCA-PSO-RBF预测模型的有效性及优越性.
为获得不同埋深下具有岩爆倾向性的灰岩相似模型,引入岩爆能量原理建立岩爆相似准则,参考原岩相关参数结合岩爆相似准则推算相似模型对应的原岩目标相关参数.引入岩爆倾向性指数、弹性能和耗散能,以石英粉含量、重晶石粉含量、石膏水泥质量比、水含量为影响因素,采用 4 因素 5水平的正交试验方案,进行单轴抗压试验得到各项参数.对试验结果进行敏感性分析及多元线性回归分析,将不同埋深下灰岩相似模型参数带入线性回归方程,并考虑敏感性分析结果对材料配比进行优化,得出不同埋深下灰岩相似模型优化配比及物理参数,为灰岩室内研究提供参考.
提出一种基于注意力机制融合轻量化网络的桥梁裂缝图像分类方法.以轻量化卷积神经网络为理论基础分类识别桥梁裂缝图像,并在轻量化网络中加入注意力机制以解决网络无法自主关注所感兴趣区域的问题.根据桥梁图像中裂缝所占比例较小且边缘突出的特点,选用适合于识别桥梁裂缝的注意力机制——CBAM(convolutional block attention module),并将其嵌入轻量化卷积神经网络Ef-ficientNetv2 中,建立CBAM-EfficientNetv2 模型.实验结果表明:CBAM-EfficientNetv2 模型与VGG16、ResNet34 等常用深度学习模型对比,可获得最优的桥梁裂缝图像分类效果,其分类识别准确率达到95.64%.
针对拆卸方案评价指标具有的模糊性和复杂耦合关系特点,提出基于二元语义云模型的拆卸方案评价模型.在查阅文献和咨询相关行业专家的基础上结合废旧产品拆卸特点,建立废旧产品拆卸方案评价指标体系,该指标体系包括技术性、经济性、环境性 3 个一级指标和 9 个二级指标;引入二元语义表征各指标相对重要度,利用层次分析法构建二元语义判断矩阵,确定评价指标权重.通过各指标分级标准计算获得云模型的数字特征,在此基础上结合指标权重,构建基于二元语义云模型的废旧产品拆卸方案的评价模型;以圆柱圆锥齿轮减速器拆卸方案为例进行评价方法应用,结果表明,所提方法不仅兼顾当前政策导向且评价结果真实、可靠.
