Predicting the time series energy consumption data of manufacturing processes can optimize energy management efficiency and reduce maintenance costs for enterprises. Using deep learning algorithms to establish prediction models for sensor data is an effective approach; however, the performance of these models is significantly influenced by the quantity and quality of the training data. In real production environments, the amount of time series data that can be collected during the manufacturing process is limited, which can lead to a decline in model performance. In this paper, we use an improved TimeGAN model for the augmentation of energy consumption data, which incorporates a multi-head self-attention mechanism layer into the recovery model to enhance prediction accuracy. A hybrid CNN-GRU model is used to predict the energy consumption data from the operational processes of manufacturing equipment. After data augmentation, the prediction model exhibits significant reductions in RMSE and MAE along with an increase in the R2 value. The prediction accuracy of the model is maximized when the amount of generated synthetic data is approximately twice that of the original data.

As global climate change intensifies, assessing product carbon footprints serves as a foundational step for quantifying greenhouse gas emissions throughout a product’s lifecycle, forming the basis for achieving sustainability and emission reduction goals. Traditional lifecycle assessment methods face challenges such as subjective boundary definitions and time-consuming inventory construction. This study introduces PCF-RWKV, a novel model based on the RWKV architecture with task-specialized low-rank adaptations (LoRAs). Trained on carbon footprint datasets, the model minimizes memory use and data interference, enabling efficient deployment on consumer-grade GPUs without relying on cloud computing. By integrating multi-agent technology, PCF-RWKV automates the creation of lifecycle inventories and aligns production processes with emission factors to calculate carbon footprints. This approach significantly improves the efficiency and security of corporate carbon footprint assessments, providing a potential alternative to traditional methods.

As global climate change intensifies, assessing product carbon footprints has become essential for measuring greenhouse gas emissions throughout a product’s lifecycle, a key factor in achieving sustainable development and emission reduction goals. Traditional lifecycle assessment methods encounter challenges such as subjective product boundary determination and lengthy lifecycle inventory construction. Recent advancements in large language models offer new opportunities for rapid, knowledge-based content generation, yet these models often lack domain-specific training, face data security concerns, and have complex deployment requirements. This study introduces the PCF-RWKV model, which utilizes the RWKV architecture with multiple stacked residual blocks and three task-specialized Low-Rank Adaptations (LoRAs). Trained on a carbon footprint assessment-specific dataset using low-rank adaptive techniques, the model minimizes data interference and memory waste, enabling efficient deployment on consumer-grade single GPUs without relying on cloud storage.By integrating Multi-Agents technology, PCF-RWKV automates the construction of LCI for production processes, aligns production processes with emission factors to calculate carbon footprints, thereby enhancing the efficiency and security of enterprise carbon footprint evaluations and addressing the limitations of traditional methods.

对新工科建设背景下清华大学3D创新打印实验室进行多层面、多维度、多系统的建设与管理,以基础建设为前提,以3D打印制度建设为依托,以团队建立为核心,构建了3D打印创新实验室管理模式,该模式具有一定示范作用,为3D打印实验室建设与管理提供了借鉴与参考.

Large-area silicon-nitride thin films deposited from silane and ammonia by plasma-enhanced chemical vapor deposition are investigated experimentally in a 300 mm apparatus with a vertical showerhead. The responses of deposition rate and refractive index to the process parameters are found and discussed. The effects of showerhead configuration on the full-wafer deposition rate and refractive index are further examined, and the inherent non-uniformity is improved by using a proper convex showerhead. The residual gases are analyzed online, and a good correlation between the partial pressure of hydrogen and the deposition rate is found.

随着高等教育全球化步伐的加快,全英文教学(EMI)在母语为非英语国家的高校得到了迅速发展,但EMI教师资源的不足是各国高校在推行EMI课程时所面临的主要困境之一.文章对当前国内外EMI教学的发展现状进行了概述,并以清华大学教师在牛津大学的EMI培训为例,介绍了牛津大学EMI中心的教学培训内容和思路,分析了其EMI教学培训的特点.最后基于我国高校EMI教学的发展需求,提出了推进高校EMI课程建设和教师培训的建议.

The robustness of a system refers to the ability of tolerating unknown perturbations which might affect the system, and there are many errors in the manufacturing or assembly process of gear systems.Therefore,a robust optimization design method of gearbox is proposed considering the uncertain factors. The optimization method contains three steps:(1)system simulation, or in other words, build the dynamic model of gearbox to acquire its dynamic property and calculate the robustness of gearbox using Monte-Carlo method;(2)constructing target function,goal programming method is used to collect mass, dynamic and robust property together to be a comprehensive target function;(3)target optimization, compound form method is used to optimize the comprehensive target function to get the optimum solution.Then,the robust optimization of a parallel spur gear train is given as an example to show how the proposed method works, and from the performance of the optimum solution, it can be concluded that the robustness of the optimum solution of robust optimization is obviously better than that of normal optimization.

Showerhead is a typical and important part in a vertical plasma-enhanced chemical vapor deposition (PECVD) reactor. A contrast experiment with the same recipe but different showerhead configurations was carried out, where the effects of the blocked area, position and drilling depth of the arrayed holes on the full-wafer deposition rate and the ratio of nitrogen to silicon of the silicon-nitride thin film were examined. The results of the blocking-hole scheme demonstrate that: the deposition rate is higher under the blocked region; the gradients of the deposition rate in the edge-opened configurations are larger than those in the center-opened configurations; the tendency of the ratio of nitrogen to silicon goes up towards to the centroid in the edge-opened configurations. The results of the drilling-hole scheme show a similar characteristic, however, the gradients are much smoother than those in the blocking-hole scheme. These results indicate that it is a feasible way to globally or locally refine the qualities of the large-area thin film through tailoring or varying the drilling depth of the arrayed holes on the showerhead face plate. This paper contributes to the structural design for the large-area silicon-nitride thin-film PECVD rector with vertical showerhead, and can also be referred to the other large-area thin-film reactor with multi-hole arrayed showerhead.

Process chamber is the core unit of chemical vapor deposition and etching, and the physical fields in it have fatal effect on process quality. It is significant and difficult for improving the process performance to regulate the fields' profiles finely. Two design solutions for the profile regulation are proposed: controllable type and resistance type. A novel profile error feedback method is presented, and a simulation-based auto-design framework is established. The profile error feedback method is in a quasi-closed-loop-control mode. It starts from an initial guessed sequence of the design/control variables and predicts a better sequence via feedback of the profile error between the output-targeted profile and the expected one. It never stops until the profile error is narrowed in the preset tolerance. Three kinds of numerical experiments about the regulation of the thermal, fluid, and plasma profile are set to test the effectiveness and feasibility of the profile error feedback method and the two kinds of design schemes.

近年来随着教育全球化进程的加快,在提升国际竞争力和加快高等教育国际化的驱动下,EMI教学在母语为非英语的国家和地区迅速发展.通过分析各国EMI教学的现状可以看出,师资资源和相应教学培训的不足是当前国内外EMI教学发展的最主要瓶颈之一,同时EMI教学还面临学生学习成效、教材选择和EMI课程建设等多方面的问题与挑战.对高校而言,当前应重点从EMI师资资源建设、创新EMI教学方法和完善EMI课程体系等3个方面进行突破,并在开展EMI教学实践和理论研究的同时,基于人才培养规律建立起完善的国际化人才培养体系,这还需要教育主管部门、高校、教师和学生等多方的共同努力.

读图能力是逻辑推理、空间推理和空间想象能力的综合,培养学生良好的读图能力是工程制图课程的重要教学目标之一.文章在比较本格推理与读图过程类似性的基础上,将本格推理思维引入工程制图课堂并通过2个教学案例讨论了其应用,引导学生将读图转化为逻辑严密的推理解谜过程,这不仅激发了学生学习的学术志趣,也提高了学生的读图能力和严谨性.

Components on waste printed circuit boards (PCB) are worthy reusing. Moreover components reusing is depended on appropriate PCB disassembly. At present researchers focus on the analysis and calculation of removal force. However, as there are melted solder between components and PCB when disassembling, even enough removal force cannot guarantee that components can be disassembled. Based on the removal force model put forward by our research group, the disassembly energy which takes both removal force and components separating displacement into account is treated as the disassembly criterion. According to it, the disassembly process under vibration is researched, the PCB modal on the condition of opposite side simply supported and the other opposite free is analyzed, based on the vibration plate theory PCB movement differential equation is built and PCB displacement and acceleration response is got. Separating process of through-hole devices and surface-mount devices under vertical disassembly is studied, and component disassembly is analyzed in situations of different excitation frequency, different location in PCB, components having different minimum disassembly acceleration. Disassembly figure method is proposed to compare the energy components get in disassembly process and the minimum energy components need, which will give a guild to disassembly parameters. At last, experiments are performed to prove the disassembly model.

The temperature distributionin chemical vapor deposition (CVD) reactor, especially on the surfaces of the Si-wafer, was mathematically modeled, theoretically analyzed, numerically simulated, and experimentally evaluated. The influence of the realistic situation, including but not limited to the pressure, gas flow-rate, and low-pressure induced temperature-jump at the wafer/heater gap, on the temperature profile on Si-wafer surfaces was investigated. The simulated results show that the average temperature on Si-wafer surface increased with an increase of the pressure and/or with a decrease of the gas flow-rate, and that the temperature decreases in radial direction. Moreover, the geometry of the impedance ceramics was designed via simulation to improve the uniformity of temperature distribution on Si-wafer surface because its insertion into the Si-wafer/heater gap is capable of adjusting the temperature profile. We suggest that the simulated results be of some technological interest in design of CVD chamber and control of the film growth. © 2016, Science Press. All right reserved.

Under conditions of melt solder,whether components might be disassembled was de-cided by the separating distance between components and the substrate (named as separating displace-ment)caused by removal force.Firstly,the models of separating displacement for surface mount de-vices (SMD)and through-hole devices (THD)were established under the conditions of vertical disas-sembly and horizontal disassembly.Secondly,based on the Laplace equation and the Newton’s equa-tion for hydrostatic pressure,tensile breaking height of solder was analyzed and modeled in vertical disassembly for 3 typical kinds of components and the programmed algorithm was elaborated in detail. Finally,to verify the models for tensile breaking height of the melted solder,based on SMD with mounting base experiments were performed and the results show that the models for tensile breaking height are effective for melting solder.Models of separating displacements for components will be helpful to quantify the distance between components and the substrate,which is helpful for preparing the dismantling process of waste PCBs.

电子产品的大量报废使得废弃线路板大量增加,从而使其作为整体丧失原有功能,但其上的元器件特别是集成芯片重用价值高.在拆解重用过程中,加热解焊是重要工艺,它将对已在一定湿度中吸潮饱和的芯片产生热冲击,进而使芯片中产生热应力及湿应力,塑封芯片可能产生分层.为研究芯片界面强度对分层的影响,首先在内聚力理论指导下,对新旧芯片的铜衬底-模塑料间界面强度通过试验-仿真方法进行测量,荻取断裂能、最大张力位移及剪切强度等参数;然后基于ANSYS软件的内聚力模型,建立基于湿-热综合膨胀系数的仿真模型;在此模型中,研究实际中热、湿的极限工况分别对新、旧芯片的界面处切向变形的影响,进而对拆解策略提供指导.

反应腔室晶片表面温度分布是影响薄膜沉积均匀性的重要因素.本文以载有晶片的典型化学气相沉积反应腔室为研究对象,考虑了晶片与加热盘之间的微小间隙导致的低压下流固接触面的温度跳跃现象,建立了腔室温度仿真模型.通过模型和实验研究了腔室气压和气体流量对晶片表面温度分布的影响,并研究了温度分布轮廓的调节方法及传热结构设计.研究结果表明,晶圆表面平均温度随气压的增加而增加,随气体流量的增加而降低,晶片从中心到边缘温度分布轮廓存在下降的趋势.通过在加热盘和晶片之间加入阻抗介质,可以有效调节晶片表面温度分布轮廓.本文最后通过仿真手段,得到了能使晶片表面温度均匀分布的阻抗介质表面形状.本文的研究结果对指导腔室结构设计和工艺控制具有重要意义.

In this paper, planet pin parallelism error, which is one of manufacturing error of planet carrier, is employed as a main variable to influence planet load sharing factor. This error is categorize two group: (i) pin parallelism error with rotation on the axis perpendicular to the tangent of base circle of gear (x axis rotation in this paper) (ii) pin parallelism error with rotation on the tangent axis of base circle of gear (y axis rotation in this paper). For this study, the planetary gear system in 1.5MW wind turbine is applied and pure torsional rigid body model of this planetary gear is built using Solidworks and MSC.ADAMS. Based on quantified parallelism error and simulation model, dynamics simulation of planetary gear is carried out to obtain dynamic mesh load results with each type of error and load sharing factor is calculated with mesh load results. Load sharing factor formula and the suggestion for planetary reliability design is proposed with the conclusion of this study.

Plasma enhanced vapor deposition is a complex multi-physics process which couples plasma dynamics, chemical reactions, fluid dynamics, heat, and mass transfer. The structure of a reactor has significant effects on the process performance. In order to simulate the behaviors of silicon-nitride thin-film processing in a 300mm reactor with a vertical showerhead, a comprehensive multi-model framework is developed to deal with the multi-physics phenomena in the entire reactor. Full-wafer deposition rate and ratio of nitrogen to silicon are compared between model and experiment to verify the multi-model simulation. The experiments are carried out with different showerhead configurations but the same process recipe, and the mechanisms of the thin-film profiles are analyzed based on the models. The results indicate that the spatial gradients of deposition rate and ratio of nitrogen to silicon are determined by the hole configurations on the showerhead.

Large-scale,high-reliability and long-life have become the development trends of wind power equipment.Therefore,the research on dynamic characteristics of wind turbine has great significance in improving its design and manufacturing technology.An integral dynamic modeling process was proposed based on the energy flow, material flow,information flow and topological relationship of wind turbine.Furthermore,with the integral dynamic model,the behaviors of wind tubrine were simulated by using SIMULINK,considering synthetically the wind energy model,wind load model,transmission system model and grid load model.As a result,the rotating speed characteristic and power characteristic simulated are very close to the operating parameters under rated condition,which verifies the reliability of the simulation.The torque fluctuating amplitude is decreasing from impeller to generator and the level of torque magnitude of impeller is E6 Nm,while that of generator is E4 Nm.In addition,the envelope lines of the torque curves at input and output ends of gearbox were fitted to provide macro boundary condition for gearbox power loss analysis.

The RF-discharge of the capacitively-coupled plasma (CCP), widely used in film-growth and/or ion-etching for integrated circuits (IC) fabrication, was approximated, modeled, analyzed, and simulated with software COMSOL. The influence of the discharge conditions, including the voltage, pressure and electrode gap, on the electron density and average electron temperature of the plasma in a 12-inch chamber was investigated. The simulated results show that the allowable major discharge conditions strongly affect the characteristics of CCP plasma. For example, (i) As the voltage increases, both the electron density and average electron temperature increase, accompanied by deteriorating uniformity of their radial distributions; (ii) As the pressure increases, the electron density increases with improving uniform; but the average electron temperature decreases with its uniformity changing in a decrease-increase mode; and (iii) As the electrode-gap increases, the electron density increases and the average electron temperature decreases, with increasingly improved uniformity of their distributions. We suggest that the simulated results be of much technological interest. ©, 2015, Zhenkong Kexue yu Jishu Xuebao/Journal of Vacuum Science and Technology. All right reserved.