Data-driven modeling on anisotropic mechanical behavior of brain tissue with internal pressure

Brain tissue is one of the softest parts of the human body,composed of white matter and grey matter.The mechanical behavior of the brain tissue plays an essential role in regulating brain morphology and brain function.Besides,traumatic brain injury(TBI)and various brain diseases are also greatly influenced by the brain's mechanical properties.Whether white matter or grey matter,brain tissue contains multiscale structures composed of neurons,glial cells,fibers,blood vessels,etc.,each with different mechanical properties.As such,brain tissue exhibits complex mechanical behavior,usually with strong nonlinearity,heterogeneity,and directional dependence.Building a constitutive law for multiscale brain tissue using traditional function-based approaches can be very challenging.Instead,this paper proposes a data-driven approach to establish the desired mechanical model of brain tissue.We focus on blood vessels with internal pressure embedded in a white or grey matter matrix material to demonstrate our approach.The matrix is described by an isotropic or anisotropic nonlinear elastic model.A representative unit cell(RUC)with blood vessels is built,which is used to generate the stress-strain data under different internal blood pressure and various proportional displacement loading paths.The generated stress-strain data is then used to train a mechanical law using artificial neural networks to predict the macroscopic mechanical response of brain tissue under different internal pressures.Finally,the trained material model is implemented into finite element software to predict the mechanical behavior of a whole brain under intracranial pressure and distributed body forces.Compared with a direct numerical simulation that employs a reference material model,our proposed approach greatly reduces the computational cost and improves modeling efficiency.The predictions made by our trained model demonstrate sufficient accuracy.Specifically,we find that the level of internal blood pressure can greatly influence stress distribution and determine the possible related damage behaviors.

Data Driven Vibration Control:A Review

With the ongoing advancements in sensor networks and data acquisition technologies across various systems like manufacturing,aviation,and healthcare,the data driven vibration control(DDVC)has attracted broad interests from both the industrial and academic communities.Input shaping(IS),as a simple and effective feedforward method,is greatly demanded in DDVC methods.It convolves the desired input command with impulse sequence without requiring parametric dynamics and the closed-loop system structure,thereby suppressing the residual vibration separately.Based on a thorough investigation into the state-of-the-art DDVC methods,this survey has made the following efforts:1)Introducing the IS theory and typical input shapers;2)Categorizing recent progress of DDVC methods;3)Summarizing commonly adopted metrics for DDVC;and 4)Discussing the engineering applications and future trends of DDVC.By doing so,this study provides a systematic and comprehensive overview of existing DDVC methods from designing to optimizing perspectives,aiming at promoting future research regarding this emerging and vital issue.

A Data-Oriented Method to Optimize Hydraulic Fracturing Parameters of Tight Sandstone Reservoirs

Based on the actual data collected from the tight sandstone development zone, correlation analysis using theSpearman method was conducted to determine the main factors influencing the gas production rate of tightsandstone fracturing. An integrated model combining geological engineering and numerical simulation of fracturepropagation and production was completed. Based on data analysis, the hydraulic fracture parameters wereoptimized to develop a differentiated fracturing treatment adjustment plan. The results indicate that the influenceof geological and engineering factors in the X1 and X2 development zones in the study area differs significantly.Therefore, it is challenging to adopt a uniform development strategy to achieve rapid production increase. Thedata analysis reveals that the variation in gas production rate is primarily affected by the reservoir thickness andpermeability parameters as geological factors. On the other hand, the amount of treatment fluid and proppantaddition significantly impact the gas production rate as engineering factors. Among these factors, the influence ofgeological factors is more pronounced in block X1. Therefore, the main focus should be on further optimizing thefracturing interval and adjusting the geological development well location. Given the existing well location, thereis limited potential for further optimizing fracture parameters to increase production. For block X2, the fracturingparameters should be optimized. Data screening was conducted to identify outliers in the entire dataset, and adata-driven fracturing parameter optimization method was employed to determine the basic adjustment directionfor reservoir stimulation in the target block. This approach provides insights into the influence of geological,stimulation, and completion parameters on gas production rate. Consequently, the subsequent fracturing parameteroptimization design can significantly reduce the modeling and simulation workload and guide field operations toimprove and optimize hydraulic fracturing efficiency.

Making Data-Driven Transportation Decisions for Freight Operations

Using Louisiana’s Interstate system, this paper aims to demonstrate how data can be used to evaluate freight movement reliability, economy, and safety of truck freight operations to improve decision-making. Data mainly from the National Performance Management Research Data Set (NPMRDS) and the Louisiana Crash Database were used to analyze Truck Travel Time Reliability Index, commercial vehicle User Delay Costs, and commercial vehicle safety. The results indicate that while Louisiana’s Interstate system remained reliable over the years, some segments were found to be unreliable, which were annually less than 12% of the state’s Interstate system mileage. The User Delay Costs by commercial vehicles on these unreliable segments were, on average, 65.45% of the User Delay Cost by all vehicles on the Interstate highway system between 2016 and 2019, 53.10% between 2020 and 2021, and 70.36% in 2022, which are considerably high. These disproportionate ratios indicate the economic impact of the unreliability of the Interstate system on commercial vehicle operations. Additionally, though the annual crash frequencies remained relatively constant, an increasing proportion of commercial vehicles are involved in crashes, with segments (mileposts) that have high crash frequencies seeming to correspond with locations with recurring congestion on the Interstate highway system. The study highlights the potential of using data to identify areas that need improvement in transportation systems to support better decision-making.

Data-Driven Model Identification and Control of the Inertial Systems

In the synthesis of the control algorithm for complex systems, we are often faced with imprecise or unknown mathematical models of the dynamical systems, or even with problems in finding a mathematical model of the system in the open loop. To tackle these difficulties, an approach of data-driven model identification and control algorithm design based on the maximum stability degree criterion is proposed in this paper. The data-driven model identification procedure supposes the finding of the mathematical model of the system based on the undamped transient response of the closed-loop system. The system is approximated with the inertial model, where the coefficients are calculated based on the values of the critical transfer coefficient, oscillation amplitude and period of the underdamped response of the closed-loop system. The data driven control design supposes that the tuning parameters of the controller are calculated based on the parameters obtained from the previous step of system identification and there are presented the expressions for the calculation of the tuning parameters. The obtained results of data-driven model identification and algorithm for synthesis the controller were verified by computer simulation.

一种Wi-Fi RTT/数据驱动惯性导航行人室内定位方法

为了研究基于智能手机的行人室内定位方法,并提高其精度,本文提出了一种基于Wi-Fi往返时间(RTT)、惯性测量单元(IMU)的定位系统。该方法主要包括3部分:(1)使用扩展卡尔曼滤波融合测距信息的Wi-Fi RTT室内定位方法;(2)适用于多手机使用模式的航位推算方法,该方法基于长短时记忆模型(LSTM)建立神经网络模型,预测行人运动速度及航向;(3)基于误差状态卡尔曼滤波的Wi-Fi RTT/数据驱动惯性导航融合定位方法,进一步提高定位精度。试验结果表明,与单一的基于Wi-Fi RTT方法和数据驱动惯性导航方法相比,本文方法的平均定位精度提升了10%~20%。

基于LSTM-MPC的PEMFC运行状态建模与容错控制

质子交换膜燃料电池(proton exchange membrane fuel cell,PEMFC)具有多物理场耦合特性易产生不同故障且难以控制。为了能在故障状态下快速有效控制,提出基于模型预测控制(modelpredictivecontrol,MPC)的容错控制方案。首先,以长短时记忆神经网络(long short-term memory,LSTM)的预测误差为遗传算法的适应度函数,寻优获取LSTM的最优超参数组合,基于数据驱动构建PEMFC系统在4种不同运行状态下的LSTM预测模型作为预测模型模块。然后,建立基于神经网络的控制器作为优化控制器模块,根据上述模块制定以PEMFC系统阴阳极输入气体压强为控制量、电堆电压为输出量的容错控制方案。最后,仿真验证LSTM预测模型与容错控制方案得到,LSTM预测模型在训练集和测试集的评估指标均方根误差(root mean square error,RMSE)指标值分别为0.0489和0.0558,具有较好的拟合效果。在不同故障状态下,MPC相较于传统PID容错控制方案电压恢复时间缩短50%及以上,并在氢气泄露故障状态下,最大压降降低22.2%,证明了所提控制策略的有效性和正确性。

智慧引领、数据赋能--全国城市交通跟踪监测与评估

通过智慧化和数据驱动的方式,提升我国城市交通基础设施监测和评估水平。首先,分析我国城市交通基础设施规模庞大但监测和管理水平有待提高的现状,并强调利用大数据技术进行城市交通监测的重要性。其次,回顾国内外城市交通监测理论与实践的发展,包括城市交通年报的编制和大范围城市群的交通状况监测报告。介绍全国主要城市交通监测技术方面的探索,包括技术路线、数据来源、指标筛选、算法优化和效果回顾等。结果显示,报告发布引起了社会各界对城市交通问题的广泛关注,大量城市也出台了相应改善措施,全国层面的交通监测让交通设施水平和服务水平透明化,进而推动城市交通治理。最后,提出未来的研究方向,从监测走向决策支持,开展精细化交通治理研究等。通过智慧化和数据化手段提升城市交通监测评估水平,为我国城市交通治理提供了科学支撑。

知识-数据驱动的沉管隧道接头安全状态分析方法——以港珠澳大桥海底沉管隧道为例

为解决沉管隧道接头安全状态不断变化且难以直接感知的技术难点,以港珠澳大桥海底沉管隧道为工程背景,通过分析既有监测数据,总结管节接头的变形模式,揭示管节接头张合量与结构温度的强相关性,明确潮位变化对接头剪切变形的显著影响。在此基础上提出一种基于知识-数据驱动的沉管隧道接头变形快速推演方法,通过建立沉管隧道精细化有限元模型,开展海量典型变形模式下的沉管隧道结构力学行为分析,构建沉管隧道变形服役行为数据集;利用BP神经网络,建立基于仿真接头服役行为特征的沉管隧道接头全断面变形推演模型,实现基于有限实测数据的接头全断面变形快速重构。该方法在港珠澳大桥海底沉管隧道的现场管养中得到成功应用。以2023年台风“苏拉”为例,基于台风登陆过程中接头的局部位移实测数据,推演接头剪力键及止水带关键点位处管节接头的变形情况。结果表明,该沉管隧道接头系统整体受台风影响较小。

电磁目标表征:知识-数据联合驱动新范式

电磁目标表征是电磁空间态势感知中的一项共性基础性问题。早期目标表征基于专家经验知识,需要设计者具有较强的专业背景与先验知识,其在复杂信号环境下的性能不佳。近年来发展起来的深度学习为复杂电磁环境下的目标信号表征提供了新的途径,它通过模拟人脑的深层结构建立机器学习模型,以端到端的方式自动表征和处理目标数据,在电磁目标检测、分类、识别、参数估计、行为认知等感知任务中显示出良好的性能。然而,深度学习严重依赖海量高质量标注数据,在现实电磁环境中存在一定局限。将知识融入智能系统一直是人工智能的研究方向,结合知识与数据进行电磁目标表征,将有望提升目标感知精度与泛化能力,正在成为电磁目标表征中新的方向。本文回顾了电磁目标表征技术的发展过程,对新的知识-数据联合驱动的电磁目标感知范式进行了展望。

基于LDA-BP神经网络的高校思政课教师数据驱动决策力评价研究

数据驱动决策力为高校思政课教师提供科学合理的教学判断,对数据驱动决策力进行评价研究,有助于提高高校思政课教师的数据决策水平,进而提升思想政治教育教学质量。鉴于传统评测方法缺乏客观性与可重复性,运用LDA-BP神经网络技术构建高校思政课数据驱动决策力的指标体系与评价模型。首先,运用LDA方法对高校思想政治教育相关的政策文本与研究文献进行主题提取,并将主题信息作为指标构建基础;其次,通过研读文献与政策文本,并结合主题分析结果构建高校思政课教师数据驱动决策力评价指标体系;最后,通过对BP神经网络的训练及测试来生成高校思政课教师数据驱动决策力的评价模型。研究表明,高校思政课教师的专业知识、教学水平以及数据分析与解读能力是影响数据驱动决策能力的关键因素,据此,理应从素养提升、文化培育、管理革新、政府支持等方面入手增强数据驱动决策力。

电网优化调度的模型-数据-知识融合方法研究评述及展望

随着新型电力系统时变不确定因素日益增多和非线性更为增强,电网调度决策的复杂度剧增,充分利用物理模型、数据驱动和调度员知识经验的互补特性,有望实现电网优化调度决策效率和场景适应性的提升。首先,从不确定性调度、数据驱动和知识驱动3方面分别梳理了电网优化调度相关新发展;其次,分析了模型-数据-知识融合的内涵,将3者融合架构分为主从驱动模式和对等驱动模式,分别评述了国内外学者的相关研究工作;最后,针对模型-数据-知识融合在电网优化调度中的应用现状分析了存在问题,并从模型-数据-知识融合效果的量化评价、调度可信知识的主动筛选和演绎、多模式融合的电网调度可靠智能决策和自主趋优演化4个方面对未来研究方向进行了展望。

基于数据-模型混合驱动的电力系统机电暂态快速仿真方法

数据驱动建模方法改变了发电机传统的建模范式,导致传统的机电暂态时域仿真方法无法直接应用于新范式下的电力系统。为此,该文提出一种基于数据-模型混合驱动的机电暂态时域仿真(data and physics driven time domain simulation,DPD-TDS)算法。算法中发电机状态变量与节点注入电流通过数据驱动模型推理计算,并通过网络方程完成节点电压计算,两者交替求解完成仿真。算法提出一种混合驱动范式下的网络代数方程组预处理方法,用以改善仿真的收敛性;算法设计一种中央处理器单元-神经网络处理器单元(central processing unit-neural network processing unit,CPU-NPU)异构计算框架以加速仿真,CPU进行机理模型的微分代数方程求解;NPU作协处理器完成数据驱动模型的前向推理。最后在IEEE-39和Polish-2383系统中将部分或全部发电机替换为数据驱动模型进行验证,仿真结果表明,所提出的仿真算法收敛性好,计算速度快,结果准确。

基于ReliefF-RBF的路面不平度识别算法研究

路面不平度对道路车辆行驶安全性及车辆动力学响应具有重要影响。通过将路面不平度识别与先进悬架控制结合,有望能进一步提升乘员舒适性和车辆的操纵稳定性。现有基于数据驱动的路面分类方法难以高效处理时变参数与车速,现有基于模型的路面识别算法需要已知精确车辆模型,在实际应用中面临车辆物理参数难以获得的问题。提出一种融合模型和数据驱动的路面分类算法,采用基于模型的方法反算等效路面轮廓,结合数据预处理方法,对车辆响应和反算等效路面轮廓数据进行滤波;对等效路面轮廓和响应信息进行时域频域特征计算,采用ReliefF算法进行关键特征提取,构建基于径向基函数神经网络的路面分类器,进行路面分级识别;通过仿真试验和实车试验验证了不同车辆参数和车速下所提出的算法鲁棒性。

基于数据驱动的电-热-气综合能源系统概率多能流计算分析

针对可再生能源与系统负荷波动对综合能源系统多能流分布的不确定性量化问题,提出一种基于数据驱动的综合能源系统概率多能流计算方法。首先,提出了考虑压缩机不同工作模式的综合能源系统多能流计算统一模型,并探讨了压缩机不同工作模式对能流分布的影响;其次,提出基于支持向量回归的概率能流计算方法,先通过多次重复的确定性多能流计算,构建数据样本集,再用支持向量回归挖掘出综合能源系统中已知负荷、网络节点信息与未知节点参数的非线性映射关系;最后,通过算例分析对提出的多能流计算统一模型在不同压缩机工作模式下的有效性进行了验证;通过与传统概率多能流计算方法对比研究,证明提出的数据驱动概率能流计算方法具有更高的计算精度与效率。

基于数据-模型融合驱动的新能源场站宽频阻抗在线辨识及稳定性评估

风电、光伏等新能源在电网中的渗透率不断提高,导致一系列宽频振荡问题。阻抗法是研究新能源并网系统宽频振荡稳定性的有效方法之一,然而由于商业化新能源机组的“黑/灰箱”问题、新能源场站运行状态的随机不确定性等因素,导致新能源场站的精确阻抗/导纳建模困难。为此,提出基于数据-模型融合驱动的新能源场站宽频阻抗/导纳在线辨识方法,针对输入随机不确定的新能源机组,采用数据驱动方法建立其覆盖整个稳态运行工况的宽频阻抗/导纳辨识模型,然后结合新能源场站的物理结构模型,构建数据-模型融合驱动的新能源场站宽频阻抗/导纳在线辨识模型。最后,以风电并网系统为例,利用建立的导纳辨识模型对风电并网系统的宽频振荡稳定性进行在线评估,并通过时域仿真验证稳定性在线分析的正确性。

融合K-means聚类和序列分解的实车锂电池剩余使用寿命预测

电动汽车锂离子动力电池健康状态(SOH)衰退过程受使用工况影响存在较多波动,导致模型预测精度下降,在锂电池剩余使用寿命(RUL)短期预测时,SOH波动情况不可忽略,为了准确预测SOH短期内波动情况,须从实车上传的锂电池运行数据中提取有效的健康因子。本文建立一种联合分布特征输入和序列分解融合的锂电池RUL预测方法,使用K-means聚类方法构建车辆锂电池运行过程的联合分布特征,并通过S-G滤波器对SOH衰退曲线进行序列分解,分别使用长短时记忆神经网络(LSTM)和多层感知机(MLP)对趋势部分和波动部分进行预测,融合得到最终预测结果。理论分析和实车采集数据验证表明,融合模型可以在预测车辆锂电池RUL短期衰退趋势的同时预测SOH的波动情况,有较高的短期预测精度。

面向新型电力系统运行的数据-物理融合建模综述

构建新型电力系统是我国实现碳达峰、碳中和目标的关键,将给电力工业带来深刻变革与挑战。数据-物理融合建模(简称融合建模)是一类新兴的建模技术,能够同时发挥物理机理与数据的价值,有望成为新型电力系统重要的分析工具。为此,该文首先梳理融合建模的相关概念与应用场景,讨论近年来国内外的研究趋势与热点。进而从技术特征和融合模式两方面,提出针对融合模型的分析框架。同时,聚焦于新型电力系统运行领域,全方位总结整理融合建模在应对现有技术挑战方面的潜在优势及不足,并展望未来研究与工程实践的重点发展领域。

从“技术驱动”转向“数据驱动”的智慧农业教育改革--以植物与环境课程中的病虫害防治为例

随着智慧农业信息化、数字化和智能化技术的迅猛发展,如何从“技术驱动”转向“数据驱动”的智慧农业教育改革成为亟需解决的问题。针对涉农专业学生多数呈现低迷学习状态,对传统农业基础知识与生产技术不感兴趣等现实问题。该文利用深度学习技术,对智慧农业教学模式、实例、教学方式及学生学习成绩评定等方面进行教改创新探索。例如,围绕智能温室病虫害防治等关键技术问题,引导学生查阅数据,提出病虫害防治的立项与设计方案;观看课外教学视频的“微课件”,学生与教师实时交互探讨问题;将农业相关数据、资料汇集、整理,建立数据模型、构建病虫害防治预警方案。这种“数据驱动”的教学模式,为学生创造出多样化的学习环境,提升学习体验,使学生认识到掌握智慧农业先进科学技术、知识的实用价值,从而提高教学效果与涉农教育的“适切性”。

基于改进MMD-GAN的可再生能源随机场景生成

针对可再生能源出力不确定性的准确表征问题,提出了一种基于改进的最大均值差异生成对抗网络(maximum mean discrepancy generative adversarial networks,MMD-GAN)的可再生能源随机场景生成方法。首先,阐述了GAN及MMD-GAN的基本原理,提出了MMD-GAN的改进方案,即在MMD-GAN的基础上改进鉴别器损失函数,并采用谱归一化和有界高斯核提升生成器和鉴别器的训练稳定性。然后,设计了基于改进MMD-GAN的可再生能源随机场景生成流程。最后,分析了所提方法在可再生能源随机场景生成中的效果,比较了改进MMD-GAN方法与MMD-GAN方法及典型GAN方法的性能差异。结果表明,改进MMD-GAN方法在生成分布和真实分布的Wasserstein距离上较对比方法降低超过50%,生成的场景精度得到有效提升。