Handwritten Mathematical Expression Recognition (HMER) remains challenging because mathematical notation is both spatially structured and highly variable across writers. Existing methods either operate on offline raster images, thereby discarding pen-stroke dynamics, or rely on modular online pipelines in which segmentation and relation errors propagate downstream. We present STRAMER (Structural Tree-Aware Mathematical Expression Recognizer), an end-to-end multi-modal architecture that integrates a DenseNet image encoder and a BiLSTM stroke encoder through bidirectional cross-modal fusion, combines a coverage-augmented Transformer decoder with a tree-aware biaffine dependency module for explicit syntactic supervision, and appends a spatial relation head that yields a complete Stroke Label Graph (SLG). To ensure globally valid structure, tree decoding is performed with maximum spanning arborescence inference and SLG relation labels are obtained with constrained CRF decoding. The model is trained with a triple joint objective comprising sequence, tree, and relation losses. Over five independent runs, STRAMER achieves ExpRate on CROHME 2019 while additionally producing interpretable SLG outputs. We further report an inference latency of per expression on a single GPU and provide a full complexity analysis of the relation component. These results indicate that jointly modelling visual appearance, stroke dynamics, and explicit structure is a promising direction for robust HMER.
A high-precision spectral implementation framework is presented for computing eigenvalues of second- and fourth-order Sturm–Liouville problems on finite intervals. The framework employs the trigonometric substitution to assemble all derivative operations on Chebyshev polynomials entry-by-entry from closed-form trigonometric identities, bypassing the repeated matrix multiplication that causes the standard Chebyshev differentiation matrix (CDM) approach to accumulate roundoff at large , high derivative order, and in clustered spectral regimes. By reformulating the differential problem as a generalised eigenvalue system solved via the QZ algorithm, the framework combines numerical reliability with high-order geometric convergence. Multiple boundary conditions at the same endpoint are incorporated through a node-reduction procedure; an empirical condition-number study is included to characterise its stability. All numerical experiments are conducted in MATLAB at 34-digit precision, enabling eigenvalue errors at the level of and providing self-convergence verification independent of double-precision saturation. The framework is benchmarked across six problems including the Coffey–Evans equation and free vibration of exponentially functionally graded beams, demonstrating good accuracy beyond double-precision references.
Introduction: Dynamic capabilities implies the professional functions by which an organization initiates carrying out self-adjustment and self-improvement thereby improving the efficiency and growth of the organization itself. Higher education institutions can be described as complex adaptive systems that change their behaviour as needed in order to survive, thrive & avoid deterioration. HEI in developing countries are required to increase performance to enlarge their contribution to socio-economic development. Organizational capacity is considered a pre-requisite for this performance. Organizational dynamics is enhanced by capacity development processes and activities. Performance measures must be based on a set of objectives that are linked to the mission of the organisation and its vision for the future. Objectives: The research aims to find out the relationship between perception of the respondents on their organisational dynamics and performances and course stream. This study investigates the role of organizational dynamics in the performance of the respondents. Methods: Organisational Leadership & Organisational Culture was considered to be the precedents of Organisational Dynamics and Organisational Capacity, Organisational Competency, Organisational Environment, Organisational Development and Organisational Motivation were taken as the measures of Organisational Dynamics. The primary data were collected through questionnaire and analysed using statistical tools. Results: Based on SEM analysis using data from 203 respondents studying in arts and science colleges in Tirunelveli, it is shown that that measures of Organisational Dynamics has a mediating role on the performance of students. Conclusions: The study aimed at analysing the influence of measures of organisational dynamics on the performance of students specifically academic and research performance. It opens ground for further research in the dimension of the relationship between organisational dynamics and faculty performance which is relatively focussed less.
To investigate the effects of side ratio and turbulent characteristics on the spatial correlation of along-wind fluctuating wind loads on rectangular tall buildings,synchronization pressure wind tunnel tests for rectangular tall buildings with side ratios ranging from 1/9～9 were carried out in four wind fields.Based on the experimental findings,the coherence function and vertical correlation coefficient of along-wind fluctuating wind loads were examined in relation to the effects of side ratio,turbulence intensity,and turbulence integral scale.The mathematical model of vertical correlation of rectangular high-rise buildings with a side ratio of 1/9～9 was obtained by the nonlinear least square method.The results show that the correlation coefficient of along-wind fluctuating wind load is affected by both the side ratio and separation distance,and decreases exponentially with the increase of separation distance.The along-wind fluctuating coherence function decays exponentially with the increase of frequency,and the decay rate of the coherence function is roughly positively related to the ratio of separation distance to average wind speed.The effects of turbulence integral scale and turbulence intensity on the correlation of along-wind fluctuating wind loads are different for buildings with different side ratios.It can serve as a guide for structural design and load code revision since the coherence functions and correlation coefficients of along-wind fluctuating wind loads on rectangular high-rise buildings suggested in this work correspond well with the experimental data.
Around the world, concrete is widely used as an essential building material, however it can fracture and let water and salts in, which can cause corrosion and shorten the structure's lifespan. In particular, the robustness of bacterial concrete in marine environments where it is exposed to more extreme conditions is examined in this extensive study. Making use of the special properties of bacteria that may produce calcium carbonate, bacterial concrete, also known as Bio-Concrete, takes advantage of this. Sealing the cracks helps the concrete self-heal, hence increasing the structure's lifespan. The goal of the study is to shed light on how bacterial concrete might lessen the negative impacts of environmental stressors on concrete structures in maritime environments. The study highlights how crucial it is to take into account the sustainability and economic viability of bacterial concrete in coastal environments before implementing it widely. The study immersed bacterial concrete beams in seawater for 365 days, showing no rebar corrosion, a common issue with normal concrete. It also compared the strength of normal and bacterial concrete. The research explored using RHA to strengthen M40-grade concrete and adding 5% to 10% corn starch to improve flowability and setting time without compromising strength and durability. Additionally, 0.5% silica fume was introduced to enhance concrete strength and durability. Concrete's capacity to survive weathering, chemical abrasion, and other difficulties throughout time is largely dependent on its durability. Crack prevention is essential for maintaining structural integrity, water tightness, and aesthetic appeal. Moreover, concrete structures' longevity and functioning could be greatly improved by the development of a trustworthy automated crack repair system. For this reason, maintaining the longevity of concrete structures depends critically on controlling the fracture widths. The study also examines the sustainability over the long run and the difficulties involved in using bacterial concrete in marine construction, providing important information for researchers and engineers and advancing our knowledge of the material's potential for durable and robust marine applications.
To improve the low accuracy of the zero-dimensional combustion model established by BP-NN, a particle swarm-neural network (PSO-NN) algorithm was proposed. The PSO optimize weights and thresholds of NN, and the operating and combustion parameters are constructed, and then compared with NN algorithm. The results show that comparing with NN algorithm, the zero-dimensional combustion model constructed by PSO-NN algorithm has higher prediction accuracy, and the mean square error of the main combustion period m is 0.0034, which is 78.21% lower than that before optimization. The particle swarm algorithm has quicker convergence and stronger versatility, which is suitable for the study of diesel engine 0-D model.
Ship detection using synthetic aperture radar (SAR) images has important applications in military and civilian fields, but the different sizes of the ship downgrade the detection accuracy of multiscale ships. Aiming at the problem of the poor accuracy and low efficiency of multiscale ship detection in complex scenes, this paper proposes a lightweight and anchor-free frame detection strategy for multiscale ships in SAR images. First, to deal with the problems of limited training samples, different sizes, attitudes, and angles of the ships in SAR images, a data augmentation strategy suitable for SAR images is adopted to expand the training space, followed by multiscale training to enhance the model generalization ability for multiscale ship detection. Second, a lightweight and anchor-free ship detection model based on the improved CenterNet is proposed, which abandons the dense anchor frame generation and extracts the key point of the ships for detection and positioning. Compared with the anchor frame-based detection method, this proposed detection model does not need to use the post-processing method to remove redundant anchor frames, and can accurately locate the center point of the ships with a better detection performance. Third, to reduce the model size and simplify the model parameters, a more lightweight network design is adopted in combination with the characteristics of SAR images. Hence, a residual network (ResNet) with fewer convolutional layers is constructed as the backbone network, and the cross-stage partial network (CSPNet) and spatial pyramid pooling (SPP) network are designed as the bottleneck network. The shallow ResNet can fully extract the SAR image features and reduce the training overfitting, and CSPNet and SPP can effectively combine the low-level image features to obtain the high-level features, reducing the model computation while at the same time enhancing the feature extraction ability. Finally, the evaluation index of the common objects in the context dataset is introduced, which can provide higher-quality evaluation results for ship detection accuracy and provide comprehensive evaluation indicators for multiscale ship detection. Experimental results show that the proposed strategy has the advantages of high detection efficiency, strong detection ability, and good generalization performance, which can achieve real-time and high-precision detection of the multiscale ship in complex SAR images.
This paper proposes to model and simulate the radio frequency interference (RFI), especially the narrowband RFI and the wideband RFI, for the high-frequency over-the-horizon (OTH) radar. Based on the theories of random process and linear filtering, the proposed models uses the white Gaussian noise as the input of a linear filter, so that the output is a random process with small bandwidth. In the filter design, seven kinds of response functions are proposed for the convolution method, and four sets of coefficients are proposed for the auto-regressive and moving-average method. Besides, extended approaches are also provided to increase the diversity of the RFI simulation. Numerical experiments demonstrate that the proposed methods can simulate most range-Doppler maps of real RFI.
To further elucidate the distribution characteristics of interfacial structure parameters such as void frac-tion and interfacial area concentration in rod bundle channels,this study investigates the air-water two-phase flow in the 5×5 rod bundle channels using a prototype spacer grid.The local distribution of phase interface flow parameters such as void fraction,interfacial area concentration,bubble velocity,and Sauter mean diameter were measured u-sing a four-sensor conductivity probe at different height sections in the flow direction.The results reveal that the ra-dial distribution of the phase interface parameters is mainly affected by the combined effect of lateral forces such as lift force,wall lubrication force,and turbulent dispersion force.The spacer grid mainly affects the distribution characteristics of phase interface parameters through turbulent vortex aggregation and grid shear fracture,turbulent vortex enhances bubble coalescence and grid shear enhances bubble breakup.There is a pressure drop zone down-stream of the spacer grid,and the distribution of interfacial parameters inside it undergoes drastic changes.This study provides a reference for optimizing two-phase flow phase interfacial transport models for rod bundle channels.
This review aims to provide a comprehensive analysis of active solar stills, a promising technology for sustainable water purification. Active solar stills utilize external energy sources and the sun's energy to evaporate and condense water, thereby removing impurities and producing clean drinking water. However, existing reviews on active solar stills have certain limitations. This review addresses these limitations through consistent experimental setups and meta-analyses. Additionally, the review takes a holistic approach, considering not only productivity and efficiency but also practicality, maintenance requirements, and economic feasibility. The use of phase change material (PCM) in solar stills is a potential approach to entrap the heat and reduce the losses. The review provides practical guidelines and recommendations for optimizing the performance and feasibility of active solar stills. By integrating these novel aspects, this review offers valuable insights to advance the understanding, implementation, and adoption of active solar stills as a sustainable water purification solution.
针对局部异常因子(local outlier factor,LOF)异常检测算法时间空间复杂度高、对交叉异常及低密度簇周围异常点不敏感等局限,提出了基于近邻搜索空间提取的LOF异常检测算法(isolation-based data extracting LOF,iDELOF),将基于隔离思想的近邻搜索空间提取(isolation-based KNN search space extraction,iKSSE)前置于LOF算法,以高效剪切掉大量无用以及干扰数据,获得更加精准的搜索空间.基于此完成了理论以及4 组实验分析,每组实验分别进行iDELOF算法与LOF、iForest、iNNE等多种典型算法的对比分析.结果表明:iDELOF算法通过拉大正异常点局部离群因子的差距,增强了对交叉异常以及低密度簇周围异常点的识别能力,提升了LOF的检测效果;iDELOF算法在识别轴平行异常方面与LOF同样具有明显优越性;iDELOF算法通过iKSSE所获数据子集显著小于原数据集,多数子集数据量小于原数据集的 1%,因此iDELOF的时间空间复杂度显著降低,且原数据集数据量越大,优越性越明显,当数据量足够大时,iDELOF算法的运行时间将低于IF算法.
Now a day the construction industries are facing many problems due to improper planning and time and cost overruns. This has made to discuss how it could be improved. Many techniques have been raised in recent times one of them is Earned worth administration (EWA), which gives better administration of time and cost requirements. BIM is an approach where it improves the arrangement and improvement stages of a venture plan. Here the scheduling, quantity takeoff, representation module, and earned worth administration platform play a vital role in tracking the project as per the actual planning of the project and avoiding time delay and cost overrun, then with the final updated architectural model the 5D building architectural model with reference to the time and cost is setup and keeps on updating as the calendar of the project gets updated, by incorporating this tool of the integrated system properly into the construction practice the above said problems of cost and time overruns can be overcome.
深海矿产资源储量巨大、品位高、种类丰富,世界各国都在加紧深海矿产资源开发关键核心技术攻关.面向深海矿产资源开发高挑战技术和高端装备需求,本文从深海探矿、深海采矿和海底环保等方面系统梳理了国内外深海矿产资源开发技术研究现状,总结归纳了我国深海矿产资源开发中存在的关键问题,并据此凝练出我国亟待着重解决的关键核心技术和未来深海采矿研究的机遇、挑战和新方向,最后提出发展建议和对策.研究表明:深海矿产资源开发应秉持绿色、智能、精准、高效协同、技术创新的发展理念,加强基础性科学研究的同时加快深海矿产资源开发关键核心技术联合攻关;面向深海资源开发的需求牵引和技术导向,大力推动深海资源开发相关配套技术产业的发展和科技创新成果的转化和应用,不断激发科技创新成果的内生动力,以期加快我国深海矿产资源开发的商业化进程.
为实现精细化中子物理计算程序HNET高效、精确的共振计算,本文基于自主开发的HDF5 格式多群数据库研究并优化了子群共振计算方法.针对子群参数的计算,研究并实施了经调整拟合点方法优化后的帕德近似法,以保证子群参数计算的准确性;为提高计算效率,需解决传统子群方法频繁调用中子输运求解器的问题,采用等效单共振群方法,并对共振核素进行分组,只对每组的代表性核素进行固定源方程求解.基于上述方法,开发了HNET共振计算模块,并针对典型基准例题进行分析验证.数值结果表明:优化后的子群方法能在保证计算效率的前提下,具有较高的计算精度.
In the Wireless sensor networks, the main crucial part as we observe based on literature analysis managing Real-time data, load balancing, security of data, managing data, and also transmitting large (multimedia data) from source to destination using WMSN. Using the proposed system work on energy efficiency and robustness of the system and based on result analysis achieve the better result as compared to the existing system. Introduction: The growth of wireless sensor networks during the past ten years has altered how data is gathered and retrieved from different areas. With the expansion of communication infrastructure, the requirement and use of Multimedia Wireless Sensor Networks is becoming more widespread on these days. These networks confront a number of difficulties in ensuring user data is secure, trustworthy, and private, just like any other WMSNs. Objectives: This paper works on multimedia data transmission in WMSN and for that works on 3-bit LSB Embedding for transmitting large-size data with min energy consumption rate as compared to other methods like AES, ECC, etc. Methods: 3-bit LSB Embedding will be done which provides secrecy of the data and then we will implement the Energy Efficient routing to the embedded multimedia data for WMSN . Results: The shortest path optimization, using energy efficient routing protocol, by sensor nodes to transfer multimedia-data from source-node to destination-node achieved in WMSN. The average energy consumption and throughput of 400 multimedia sensor nodes transferring large amount multimedia-data transfer from source-to-destination WMSN. Existing work shows that it has high energy consumption and low throughput as compare to the proposed work. Conclusions: In this paper, work on energy-efficient real time multimedia data trans receive using Wireless sensor network use 3-bit LSB data embedding in multimedia data transmission in a wireless sensor network for compress data. Based 3-bit LSB data embedding for digital audio, image, video, and 3D media. Given the tremendous developments in digital media communications ranging from conventional digital audio to immersive media, LSB data hiding plays an important role in providing high capacity and maintaining imperceptibility by considering mechanisms of the HAS and HVS. This paper uses a hybrid approach for secure data as well as compressing it using a 3-bit LSB embedding approach as well as for making energy optimization using a trust mechanism for set initial threshold values of parameters like min energy required, transmission power, location, transmission power, etc. in the transmission side source node and another surrounding node will be selected based on a parameter that we reserve than apply 3-bit LSB embedding on data and on the receiver and apply same de-embedding after selecting receiver node. And for establishing communication select the most appropriate and optimized path based on energy and location-based estimation. With the proposed structure we achieve significantly improve in results as achieve more throughput and less energy consumption as compared to the existing system.
为明确随机有限断层法在俯冲带板内地震动模拟中的适用性,以2021 年日本千叶Mj6.1 俯冲带板内地震为例,使用该方法模拟了震中范围100 km内25 组KiK-net台站井上和井下记录,并分析了模拟与观测记录的频谱、持时、峰值和空间分布等地震动特征.结果表明:模拟与观测记录的5%阻尼比拟加速度反应谱(Aps)在 0.1～10 Hz频带范围内吻合较好;基于70%能量持时模型的模拟记录在强震动段和持时上与观测记录匹配良好;井上台模拟与观测记录的地表峰值加速度(Apg)接近,且两者Apg衰减特征基本一致;模拟和观测记录所得Apg等值线相似.此外,将模拟和观测记录与现有日本俯冲带板内地震动衰减关系式(Zhao16)进行了比较,发现Zhao16 的Apg预测值普遍出现高估,其Aps预测值在低频和高频分别出现一定程度的低估和高估,这可能与研究区域的盆地效应和软土层有关.研究结果为随机有限断层法模拟俯冲带板内地震动的适用性提供了依据,进而为探索将该方法应用于中国具有相似俯冲带构造的地区提供参考.
In the field of UAV aerial video smoke detection,existing smoke detection algorithms are often suffering from low detection accuracy and slow speed due to the great diversity between different detection scenarios.To address this issue,we constructed a UAV smoke dataset（USD） in multiple scenes,and proposed an improved UAV smoke detection algorithm based on YOLOx in multiple scenes.First,an improved attention mechanism was introduced into the YOLOx network to improve the extraction process of channel features and spatial features,so as to extract more representational smoke features.Then,a bidirectional feature fusion module was presented to enhance the fusion ability of multi-scale feature fusion module for small target features.Finally,a Focal-EIOU loss function was used to solve the problems such as the imbalance of positive and negative samples in the training process,and the distance and coincidence degree of two frames cannot be reflected when the prediction frame and real frame are not intersected.Experimental results show that the proposed algorithm had good robustness when applied to UAV video smoke detection tasks in multiple scenarios.Compared with several traditional smoke detection algorithms,the accuracy of the proposed smoke detection method on different datasets was improved.In comparison with the original YOLOx-s model,the accuracy was improved by 2.7%,the recall rate was improved by 3%,and the speed reached 73.6 frames per second.
针对变速控制力矩陀螺(variable speed control moment gyro,VSCMG)作为执行机构应用在敏捷遥感卫星上进行姿态机动时末端模式切换的平稳性和快速性冲突问题,在考虑框架转速误差的基础上,设计姿态误差参数作为切换指标,制定误差参数切换区域内的过渡规则,将指令力矩实时分配给控制力矩陀螺(control moment gyro,CMG)和飞轮并分别求解,提出了一种控制力矩陀螺/反作用飞轮工作模式模糊平滑切换操纵律.为了使得姿态机动末端卫星姿态达到姿态稳定度和指向精度要求的时间更短,以该时间为优化指标提出聚类变异改进粒子群算法对该操纵律参数寻优,确定最佳的切换区域和切换参数,并进行了仿真验证.结果表明:改进后粒子群算法在相同的迭代次数中总是表现出比传统粒子群算法更优的适应度,具有更快的收敛速度和更高的收敛精度,参数优化后的模糊平滑切换操纵律相比于现有操纵律能够在较短时间内完成双模式的平滑切换,并在姿态机动末端更迅速地达到姿态稳定度和指向精度要求,提高了遥感卫星敏捷机动与高稳指向的控制性能,有利于高质量完成成像任务.
为实现古建筑木构件的无损检测,提出基于无损检测技术的木材密度和抗压强度的线性预测公式.运用超声波-针阻仪联合检测技术测量了12 种木材不同纹理方向的超声波波速和抗压强度以及径向和斜向的阻抗比,拟合得到了预测公式中的回归系数,再将预测公式应用于对2 种新木材和1 种古木的密度和抗压强度的预测中.结果表明:木材超声波波速和阻抗比与其密度和抗压强度均呈正相关线性关系,而采用超声波-针阻仪联合检测技术时,通过多元回归方法得到线性关系的拟合优度有显著提升.对新、古木材密度进行预测时,仅采用针阻仪技术即可实现精确预测,其中对新木材的预测误差小于5%,对古木的预测误差小于1%;对新、古木材抗压强度进行预测时,采用单一的检测技术预测误差较大,而采用联合检测方法的预测效果较好,对新木材的预测误差小于5%,对古木的预测误差小于9%.试验验证了采用超声波-针阻仪联合检测技术对木材密度和抗压强度进行精确预测的可行性,成果可为古建筑木构件的健康监测提供重要技术支持.
海水海砂混凝土(seawater sea sand,SSC)在岛礁和临海工程建设中有广阔的应用空间.海洋环境下,混凝土容易开裂,严重影响结构耐久性.为确保该新型混凝土在海洋环境下的安全服役,对SSC的断裂力学性能研究以及断裂参数的合理确定至关重要.然而采用传统方法,基于线弹性断裂理论确定的断裂参数,由于未考虑材料非均匀性的影响,不可避免存在尺寸效应.鉴于此,本文利用基于边界效应的非线性断裂理论,考虑材料的非连续性与非均匀性,确定SSC的真实断裂参数.配制最大骨料粒径dmax为10 和20 mm的海水海砂混凝土,分别进行高度为100 和200 mm的三点弯曲梁试验,初始缝高比为0.1～0.7;并用淡水、河砂等质量替代海水、海砂配制普通混凝土(ordinary Portland cement,OPC)作为对照组进行试验.基于边界效应模型,通过引入混凝土平均骨料粒径,得到了SSC的真实无尺寸效应拉伸强度ft及断裂韧度KIC,进而结合正态分布分析确定两断裂参数的均值以及具有95%保证率的上下限值,并利用得到的拉伸强度成功预测任意尺寸条件下SSC试件的极限荷载.结果表明:相同骨料级配下,与普通混凝土相比,海水海砂混凝土断裂截面上骨料断裂的比例更高,SSC的拉伸强度及断裂韧度高于OPC;随着最大骨料粒径的增大,SSC和OPC骨料断裂的比例均减小,其断裂韧度KIC均增加,ft降低.所提模型及相关结果可为海水海砂混凝土实际工程设计提供参考.
