Fushun, located in Northeast China, is prone to debris flow disasters due to its complex topographical and geological conditions. In 2013, large-scale debris flow disasters triggered by a rainstorm broke out, causing hundreds of casualties and serious economic losses. Therefore, it is significant to evaluate the susceptibility and hazard of debris flows in Fushun. Instead of adopting grid units, this research adopts the hydrological response units as the evaluation units and conducts debris flow susceptibility assessment for Fushun area with analytical hierarchical process. By combining the susceptibility with two different precipitation data, hazard assessment is further conducted. Comparison between the two hazard maps is conducted to explore the influence of precipitation on debris flow hazard assessment. Under different precipitation conditions, the debris flow hazard of the same area changes. Statistics on the accuracy of the susceptibility and hazard assessment results is conducted and further compared with the local existing debris flow disaster records, demonstrating that the evaluation results are generally in good consistency with the actual situation in Fushun.
Existing research combines acupuncture theory with network science and proposes a new paradigm for the study of acupoint selection patterns-a key acupoint mining algorithm based on acupoint networks. However, the basic idea of this study for finding key acupoints is based on binary acupoint synergy relationships, which ignores the higher-order synergy among multiple acupoints and does not truly reflect the implicit patterns of each acupoint among meridian systems. Moreover, the mining results assessment method, which this new paradigm involves, does not have wide applicability and universality. In this paper, with the introduction of higher-order interactions between multiple acupoints, a high-specificity key acupoint mining algorithm based on 3-node motif is proposed in the acupoint-disease network (ADN). In response to the narrow applicability of the new research paradigm involving the evaluation of algorithms' measures, new and widely applicable and universal evaluation criteria are introduced in terms of resolution, network loss, and accuracy, respectively. Based on the principles of acupoint selection involved in acupuncture clinics in Chinese medicine, the acupoints involved in the data were divided into a total of 19 regions according to their distribution characteristics. From these 19 regions, we selected the key acupoints that have a large impact on the global network. Finally, we compared this algorithm with five other acupoint importance assessment algorithms in terms of resolution, network loss, and accuracy, respectively. The comprehensive results show that the algorithm identifies key acupoints with an accuracy of 63%, which is 14% to 21% higher than other existing methods. The key acupoints identified by the algorithm have a significant disruptive effect on the connectivity of the network, indicating that the key acupoints are at the core of the acupoint-disease network topology. They have a significant propagation influence on other acupoints, which means that the key acupoints have high-synergistic cooperation with other acupoints. Meanwhile, the stability and specificity of the algorithm ensure the reliability of the key acupoints. We believe that the key acupoints identified by the algorithm can be used as core acupoints from the perspective of network topology and high synergy of other acupoints, respectively, and help researchers explore targeted and high-impact combinations of acupoints to optimize existing acupuncture prescriptions under condition constraints.
Feeding a cold steel strip into continuous casting (CC) mold is an effective method to improve the quality of large round bloom. However, it is difficult to assess the correlation between the bloom quality and the initial temperature of the strip. Therefore, a three‐phase mixed columnar–equiaxed solidification model is developed to evaluate the effect of feeding strip temperature on macrosegregation and solidification structure in a large vertical CC round bloom. The results show that feeding the steel strip induces “crystal rain” in the liquid core of the bloom and increases the volume fraction of the equiaxed phase. After feeding a strip, the flow direction of molten steel is changed, and the flow velocity is increased in the liquid core of the bloom. Furthermore, with initial temperature of the strip decreases, the production efficiency is improved. Compared to the four conditions investigated, the carbon distribution in the bloom is the most uniform, the negative segregation at the centerline of the bloom is the weakest with the value of −35.50%, and the extreme value difference of the macrosegregation index at 20 m below the meniscus is the smallest with the value of 38.35% under the condition of T initial = 500 K.
Aiming at the financing mode selection problem of a closed-loop supply chain consisting of a retailer,a recycler,and a capital constrainted manufacturer,the revenue functions of the manufacturer,retailer,recycler,and closed-loop supply chain are constructed when the capital constrained manufacturer can get financing from the retailer(internal financing)or the bank(external financing),respectively,and based on the Stackelberg game,the optimal pricing and return rate strategies under two financing modes are given.On this basis,by comparing the optimal decisions and revenue of the recycler,retailer,manufacturer,and closed-loop supply chain under different financing modes,the financing mode selection strategies are given from different perspectives.The research results show that under certain conditions,the retailer providing financing services are beneficial for the manufacturer,retailer,recycler and closed-loop supply chain.
相变材料(phase change materials,PCMs)是潜热储热技术的核心储热媒介,因此选择合适的PCMs需综合考虑到多种因素,而相变温度、储能密度和导热系数是衡量PCMs适用性的最关键因素.对典型的固-液相变的不同类型PCMs特点进行了归纳和比较,结果证明研究多种高性能兼容的复合材料是PCMs应用的发展方向.通过理论分析得出:在设定工作温差50℃范围内常用PCMs相变潜热≥100 kJ/kg时,单位质量储热密度与PCMs的相变温度不再相关;PCMs的相变温度越高,系统的有效输出温度范围越宽,系统输出越灵活;PCMs的相变潜热越大,系统在有效温度范围变化时释热效率越稳定.因此在50℃工作温差范围内选择相变潜热高、兼顾较高的固液平均比热容和相变温度高的PCMs时系统的输出性能更优越.
提出一种基于骨架邻近像素匹配的线结构光条中心提取方法,分别在图像预处理阶段和光条中心提取阶段对传统方法进行改进.在图像预处理阶段,将马尔可夫随机场理论应用于二值图像去噪中,同时提出了一种基于连通域面积特性的ROI(region of interest)提取方法.在光条中心提取阶段,首先提出了一种光条骨架剪枝算法,对细化ROI得到的光条骨架进行剪枝、平滑,之后综合考虑光条图像的几何特性和灰度分布特性,基于邻近分析对ROI内各像素进行划分,继而求取出灰度重心,最后经Savitzky-Golay滤波后实现光条中心提取.实验结果表明,所提方法对不同类型光条的提取适用性强,相较于Steger法精度更高,且速度在其基础上提高了约6.98倍.
基于粒子云网格法与蒙特卡罗法,使用VSim软件对微型溅射离子泵内部潘宁放电的工作过程进行了分析.建立了二维仿真模型并得到了氮气离子入射阴极板时的入射能量、入射角度和入射位置等参数.将非垂直溅射产额理论与仿真得到的结果相结合,分析了阴极板上溅射产额的分布规律.根据离子入射参数与溅射产额,计算得到了微型溅射离子泵对氮气的抽速.计算值与实验结果一致性好并且该方法可以给出溅射离子泵抽速阈值对应的压力.仿真结果中增加阳极筒长度可以降低微型溅射离子泵抽速阈值对应的压力,并通过实验证明.
针对永磁同步直线电机系统中存在的模型不确定性、状态约束以及输入非线性(如非线性电磁驱动力/输入受限)问题,提出了一种基于神经网络的自适应控制器.具体来说,为了降低噪声敏感性,进一步提高跟踪精度,采用仅依赖于参考轨迹的期望信号来替代测量信号.然后,设计神经网络在线逼近未知模型和非线性函数,并通过构造连续控制的方法来处理逼近误差.此外,构造障碍李雅普诺夫函数确保系统在运行过程中状态始终满足约束条件;并且通过严格的理论分析证明了跟踪性能满足要求.最后,通过仿真实验验证了所提控制器的有效性和鲁棒性.
在产前超声筛查过程中,为了能够帮助医生在丘脑标准平面上快速、精确地测量胎儿头围,提出一种新颖的双分支卷积神经网络直接分割胎儿颅骨边界,2个分支通过共享层相互促进,有效地提高了颅骨边界的分割精度,特别对局部不清晰或者不连续的边界仍然有着较好的分割效果,具有较高的鲁棒性.本方法的测量过程不需要过多的后处理操作,并且模型属于轻量级网络,便于部署.该方法应用在Grand-Challenge中的HC18数据集及从医院采集的300例数据上,均取得了较好的结果,对比其他主流分割网络如U-Net,Res-U-Net,U-Net++,CE-Net等,所提方法具有更高的分割精度及更小的测量误差.
为研究堆石料的湿化变形,首先通过室内颗粒破碎试验分析了粒径对岩石颗粒强度和软化系数的影响;然后依据散粒集合体应力与应变张量表达式和岩石颗粒的软化系数,建立了饱和与干燥试样的应力-应变转换关系,结合双线法预测了堆石料的湿化变形;最后通过算例分析了该方法的有效性.结果表明:浸水湿化会降低岩石颗粒的强度,饱和与干燥玄武岩颗粒的强度均服从Weibull分布;颗粒特征强度随粒径的增加而降低,但是Weibull模量和颗粒的软化系数没有明显的粒径相关性.堆石料饱和与干燥试样的应力-应变关系可通过岩石颗粒软化系数进行变换;由饱和试样的应力-应变关系和岩石颗粒软化系数可估算堆石料浸水后因颗粒破碎引起的湿化变形.算例中预测结果与试验结果整体比较接近,表明岩石颗粒浸水后强度降低导致的颗粒破碎是堆石料产生湿化变形的主要原因.
探明泥水盾构施工过程中关键掘进参数(如刀盘转速、主推进力、刀盘扭矩等)对开挖面泥浆支护效果及盾构能耗的影响规律,是保障盾构快速掘进和降低盾构机械损耗的重要前提.本研究依托济南穿黄隧道东线盾构项目掘进参数集计算各掘进环的场切深指数(FPI)和扭矩切深指数(TPI);采用掘进比能将掘进参数集划分为优配数据集以及待优化数据集,并分别基于支持向量回归和人工神经网络方法建立了关键掘进参数的预测模型.结果表明:盾构场切深指数和扭矩切深指数可有效描述掘进地层的同一性.盾构掘进比能呈对数正态分布特征,可有效表征盾构掘进工作状态并评估盾构各项掘进参数的配置水平.同一地层中当盾构掘进能耗水平波动较大时宜采用人工神经网络预测模型对刀盘转速以及盾构主推进力进行优化.
针对SNN-HRL等传统Skill discovery类算法存在的探索困难问题,本文基于SNN-HRL算法提出了融合多种探索策略的分层强化学习算法MES-HRL,改进传统分层结构,算法包括探索轨迹、学习轨迹、路径规划三层.在探索轨迹层,训练智能体尽可能多地探索未知环境,为后续的训练过程提供足够的环境状态信息.在学习轨迹层,将探索轨迹层的训练结果作为"先验知识"用于该层训练,提高训练效率.在路径规划层,利用智能体之前获得的skill来完成路径规划任务.通过仿真对比MES-HRL与SNN-HRL算法在不同环境下的性能表现,仿真结果显示,MES-HRL算法解决了传统算法的探索问题,具有更出色的路径规划能力.
混凝土宏观率相关效应研究已经取得了十分丰富的成果,其细观结构和材料参数显著影响混凝土宏观力学性能,因此,从细观尺度研究混凝土动态性能是揭示混凝土宏观率相关效应的重要途径.本文引入应变率影响,修正界面内聚力本构模型,从细观尺度研究了不同骨料含量和骨料形状的混凝土在不同加载速率下的单轴抗拉破坏模式和力学特性.分析结果表明:随着加载速率增加,混凝土破坏裂缝分布更加均匀;混凝土抗拉强度的增量与应变率的对数呈线性关系;混凝土抗拉强度随骨料含量增加近似线性减小,但骨料形状对其影响很小.
盾构荷载作为盾构的主要性能指标,准确的荷载预测对于保证盾构安全高效工作和周边环境稳定具有重要意义.鉴于传统预测方法精度差的局限性,本研究以数据的高维度特征和时序特征为切入点,提出一种结合卷积神经网络、双向长短期记忆神经网络和注意力机制的混合模型(CNN-BiLSTM-Multiattention,CBM),对盾构荷载进行精准预测.该模型不仅可以提取数据的高维度特征和时序特征,还能突出高维度特征的重要性和关键时间节点信息.通过实验证明了相较于4种现有的模型,本文所提出的模型在3种评价指标上均优于其他模型,对推力和扭矩预测的准确率达到94.2％ 和96.2％.
通过中间坯超快冷工艺,在0.2％C-2％Mn普碳钢中获得表层铁素体和心部马氏体的梯度层状组织,实现了钢板压下量约50％ 的大变形温轧.大变形马氏体经450℃和530℃退火后,制备出平均晶粒尺寸为0.52μm和0.66μm的超细晶组织.利用扫描电镜(SEM)、电子背散射衍射(EBSD)和准静态拉伸试验等手段,研究了超细晶钢板的微观组织与力学性能.结果表明,相较于610℃退火粗晶钢板,450℃和530℃退火超细晶钢的屈服强度可提升2～3倍,平均屈服强度分别达到了1475 MPa和1196 MPa,延伸率也显著下降.晶界强化和位错强化是超细晶钢强度提升的主要强化机制,而加工硬化率降低导致了超细晶钢的塑性下降.
为了明晰熔渣体系中超声的空化行为,应用Rayleigh-Plesset方程模拟1020℃条件下25％K2 O-30％Na2 O-45％SiO2熔渣中空化泡的运动行为.结果表明,在空化核尺寸15μm,声压振幅3.039 MPa的条件下,空化泡在频率25 kHz由一次振荡溃灭变为多次振荡才溃灭的瞬态空化,36 kHz变为无周期稳态空化,63 kHz变成周期稳态空化,呈现振幅降低和溃灭时间变长的特点.在空化核尺寸15μm,频率20 kHz的条件下,空化泡在声压2.4 MPa处由稳态空化变为一次振荡溃灭的瞬态空化,17 MPa处变为2次振荡溃灭的瞬态空化,170 MPa处变为多次振荡才溃灭的瞬态空化,呈现振幅增大和溃灭时间变长的特点.在声压振幅3.039 MPa,频率20 kHz的条件下,空化泡在空化核2μm由稳态空化变为一次振荡溃灭空化,21μm变为多次振荡才溃灭的空化,33μm变为非周期稳态空化,呈现振幅降低和溃灭时间变长的特点.
将纳维-斯托克斯(NS)方程方法与直接模拟蒙特卡罗(DSMC)方法耦合以实现过渡流态计算.相对过去欠缺反馈机制的单向耦合方法,提出以NS,DSMC速度场相互修正边界值的方法实现双向耦合方法的反馈机制,这对于非稳态研究和数值修正十分重要.结果表明,双向耦合结果与全局DSMC结果有高相似性,相对单向耦合具有更高的效率、稳定性及精度,提高耦合强度可以提升稳定性与精度.双向耦合通过采用全局NS结果边界值获得更高的子域收敛性、稳定性及精度.提出超前演算方法可以增强DSMC域的时间耦合性,为非稳态计算奠定基础.
以探索智能互联环境下场景化服务的创新动因及其对价值共创维度与结果的影响为研究核心,采用探索性案例研究方法,以海尔食联网为案例对象,基于场景理论与价值共创理论,对智能互联环境场景化服务影响价值共创的维度进行解析.研究得出智能互联环境下场景化服务的创新动因有外部的用户需求拉力、数智技术推力及内部的核心资源能力、战略发展能力;场景化服务在价值主张、价值共创主体、价值共创载体、价值共创过程4个维度产生影响,产出了用户价值、企业价值与生态价值.本文研究了智能互联环境下场景化服务价值共创实现机理,拓展了智能互联时代价值共创理论.数智时代需求升级趋势下,企业发展场景化服务有助于实现精准满足用户需求.
针对语音情感识别过程中特征不充分的问题,提出了约束式双通道模型,从全局和局部两方面充分挖掘特征所包含的情感信息,从而提高情感识别率.通道1是针对语音特征的全局信息,通过改进门控循环单元,构建了BAGRU(bidirectional attention gate recurrent unit)模型,提高了语音特征之间的相关性;通道2是针对语音特征的局部信息,卷积神经网络与对抗训练结合,避免了局部信息相互干扰.通过双通道融合模型,根据通道特征重要程度生成不同权重,同时引入正交约束,解决了融合时产生特征冗余的问题.研究结果表明,在IEMOCAP和EMO-DB情感语料库上分别达到了62.83％和82.19％的识别精度,表现出了良好性能.
Freezing of Gait (FOG) is the most common and disabling gait disorder in patients with Parkinson’s Disease (PD), which seriously affects the life quality and social function of patients. This paper proposes a FOG recognition method based on the Variational Mode Decomposition (VMD). Firstly, VMD instead of the traditional time-frequency analysis method to complete adaptive decomposition to the FOG signal. Secondly, to improve the accuracy and speed of the recognition algorithm, use the CART model as the base classifier and perform the feature dimension reduction. Then use the RUSBoost ensemble algorithm to solve the problem of unbalanced sample size and considerable limitations of a single classifier. Finally, the hyperparameters of the ensemble classifier are optimized by Bayesian optimization, and the experiment proves that the RUSBoost algorithm can complete the gait recognition task well. Compared with the Adaboost, Tomeklinks-Adaboost and ROS-Adaboost ensemble algorithms, the RUSBoost ensemble algorithm can complete the FOG recognition task more efficiently. When the maximum number of splits is 1023, and the number of base classifiers is 100, the performance of the RUSBoost ensemble algorithm can reach the best. The accuracy of the time recognition algorithm was 87.8%, the sensitivity was 89.7%, and the specificity was 87.5%.
