Making costly manufacturing smart with transfer learning under limited data: A case study on composites autoclave processing

The integration of advanced manufacturing processes with ground-breaking Artificial Intelligence methods continue to provide unprecedented opportunities towards modern cyber-physical manufacturing processes, known as smart manufacturing or Industry 4.0. However, the “smartness” level of such approaches closely depends on the degree to which the implemented predictive models can handle uncertainties and production data shifts in the factory over time. In the case of change in a manufacturing process configuration with no sufficient new data, conventional Machine Learning (ML) models often tend to perform poorly. In this article, a transfer learning (TL) framework is proposed to tackle the aforementioned issue in modeling smart manufacturing. Namely, the proposed TL framework is able to adapt to probable shifts in the production process design and deliver accurate predictions without the need to re-train the model. Armed with sequential unfreezing and early stopping methods, the model demonstrated the ability to avoid catastrophic forgetting in the presence of severely limited data. Through the exemplified industry-focused case study on autoclave composite processing, the model yielded a drastic (88%) improvement in the generalization accuracy compared to the conventional learning, while reducing the computational and temporal cost by 56%.     

A novel data-driven approach to support decision-making during production scale-up of assembly systems

In today's manufacturing settings, a sudden increase in the customer demand may enforce manufacturers to alter their manufacturing systems either by adding new resources or changing the layout within a restricted time frame. Without an appropriate strategy to handle this transition to higher volume, manufacturers risk losing their market competitiveness. The subjective experience-based ad-hoc procedures existing in the industrial domain are insufficient to support the transition to a higher volume, thereby necessitating a new approach where the scale-up can be realised in a timely, systematic manner. This research study aims to fulfill this gap by proposing a novel Data-Driven Scale-up Model, known as DDSM, that builds upon kinematic and Discrete-Event Simulation (DES) models. These models are further enhanced by historical production data and knowledge representation techniques. The DDSM approach identifies the near-optimal production system configurations that meet the new customer demand using an iterative design process across two distinct levels, namely the workstation and system levels. At the workstation level, a set of potential workstation configurations are identified by utilising the knowledge mapping between product, process, resource and resource attribute domains. Workstation design data of selected configurations are streamlined into a common data model that is accessed at the system level where DES software and a multi-objective Genetic Algorithm (GA) are used to support decision-making activities by identifying potential system configurations that provide optimum scale-up Key Performance Indicators (KPIs). For the optimisation study, two conflicting objectives: scale-up cost and production throughput are considered. The approach is employed in a battery module assembly pilot line that requires structural modifications to meet the surge in the demand of electric vehicle powertrains. The pilot line is located at the Warwick Manufacturing Group, University of Warwick, where the production data is captured to initiate and validate the workstation models. Conclusively, it is ascertained by experts that the approach is found useful to support the selection of suitable system configuration and design with significant savings in time, cost and effort.     

Decision-based process planning for wire and arc additively manufactured and machined parts

The conventional manufacturing of aircraft components is based on the machining from bulk material and the buy-to-fly ratio is high. This, in combination with the often low machinability of the materials in use, leads to high manufacturing costs. To reduce the production costs for these components, a process chain was developed, which consists of an additive manufacturing process and a machining process. To fully utilize the process chain’s capabilities, an integrated process planning approach is necessary. As a result, the work sequence can be optimized to achieve the economically most suitable sequence. In this paper, a method for a joint manufacturing cost calculation and subsequent decision-based cost minimization is proposed for the wire and arc additive manufacturing (WAAM) & milling process chain. Furthermore, the parameters’ influence on the results and the magnitude of their influence are determined. These results make it possible to design an economically optimal work sequence and to automate the process planning for this process chain.     

基于星敏感器的姿态优化估计算法

对于采用星敏感器的航天器姿态确定问题,提出了一种快速的姿态优化估计方法;首先,根据Rodrigues参数和观测向量之间的线性关系,引入四元数,构造了不同的非奇异的优化准则;其次,借助四元数,求解了基于该准则的姿态优化算法;误差协方差分析和数值仿真结果表明,该算法与著名的QUaternion ESTmation(QUEST)算法相比,在相同的操作系统环境下,具有相同的姿态估计精度,但计算速度更快,对两种算法选取间隔为2000次的仿真步长,统计数据长度为100个采样点,最后对结果加权求平均值,统计结果显示仅为QUEST的三分之一;这种姿态优化估计方法具有一定的工程应用价值。     

空中交通预警与防撞系统(TCAS)风险及对策研究

空中交通预警与防撞系统TCAS对飞行安全非常重要,它能够通过对附近的飞机进行冲突检测,预测未来可能发生的危险,提示飞行员避免危险,帮助机组维持与其他飞机之间的空中交通安全间隔;但TCAS本身依然存在风险;在此对TCAS进行了半物理仿真测试,包括数字样机开发测试、物理样机开发测试、虚拟试飞以及数据统计分析等,对其各项功能进行测试与验证;在参阅众多相关文献的基础上,结合所做的研究工作,讨论了TCAS的主要风险,并针对这些风险,提出TCAS在设计、测试、操作方面需要改进的地方,从各方面完善TCAS,降低其风险,进一步保障飞行安全。     

基于Ziegler-Nichols频率响应方法的自适应PID控制

提出一种基于Ziegler-Nichols频率响应方法的自适应PID控制器,它通过控制回路正常运行中的过程对象输入输出数据在线辨识出过程对象重要的临界频率响应特性,然后基于Zieger-Nichols整定规则或改进的方法在线更新PID控制器参数.PID控制器的自适应过程不需要系统的任何先验知识,也不需要建立任何对象模型,可以保证控制回路始终运行在最佳状态.仿真实验表明了自适应PID控制的有效性和可行性.     

Development of human-machine interface for teleoperation of a mobile manipulator

Human-robot control interfaces have received increased attention during the past decades for conveniently introducing robot into human daily life. In this paper, a novel Human-machine Interface (HMI) is developed, which contains two components. One is based on the surface electromyography (sEMG) signal, which is from the human upper limb, and the other is based on the Microsoft Kinect sensor. The proposed interface allows the user to control in real time a mobile humanoid robot arm in 3-D space, through upper limb motion estimation by sEMG recordings and Microsoft Kinect sensor. The effectiveness of the method is verified by experiments, including random arm motions in the 3-D space with variable hand speed profiles.     

基于综合优化方法的风力发电机故障诊断

支持向量机(SVM)一种新型的统计学习方法。但是作为分类算法,它存在计算量大、运行时间长的缺点。针对LSSVM的参数选择问题,引入物理学中的黑洞概念,建立黑洞模型,结合模拟退火算法,提出了黑洞粒子群-模拟退火算法(BH-PSOSA)。该算法可以增加粒子的多样性,克服PSO算法优化过程中陷入局部极值的问题,提高了优化性能,改善了收敛特性。利用BHPSO-SA算法对LSSVM的参数进行优化选择,用UCI数据库的数据进行分类验证,相比CV参数优化的LSSVM,提高了分类速度和精度。最后把BHPSOSA-LSSVM算法应用到风机齿轮箱的故障诊断中,取得了良好的效果。     

一种新型分形微带天线的仿真优化

研究分形微带天线优化问题,针对多频通信需求,要求回波损耗特性实现多频性。传统分形天线结构单一、多频性能差。为了改善其多频特性,利用Cantor集和Koch曲线原理,通过建立不同宽度和长度的导带模型,并将导带宽度较窄的短截线替换为Koch曲线,提出了一种新型的适合TD-SCDMA、WiMax及WLAN的无线通信的三频段分形微带天线。利用CST软件进行仿真,得到了天线仿真回波损耗及方向图,并实测了回波损耗,实测结果与软件仿真结果吻合较好。经实验证明,所设计的天线能很好满足多频通信的需求。