面向复杂制造过程质量的模糊Petri网预测及诊断方法研究

在复杂制造过程中，存在质量异常预测及诊断能力弱、智能化程度低、效率低等问题。如何针对制造过程质量问题特点采用合适的预测与诊断方法，满足日益提高的过程自动化水平的要求，是该领域研究人员面临重要的亟待解决的问题。 由于模糊Petri网是模糊集理论与Petri网理论有机结合的一种网络理论，其突出优势在于知识表示、推理和处理模糊信息的能力; 目前，尽管模糊Petri网已有成功的应用案例，但仍存在某些不足，需不断地改进与完善。因此，对模糊Petri网理论方法的研究，有利于提高其知识表示能力、模糊动态推理能力、推理结果可靠性与准确性等，对模糊Petri网理论的广泛应用具有推动作用。 本文以制造过程质量问题的预测与诊断为研究和应用背景，对模糊Petri网预测与诊断方法的研究为主线，以研发的系统为辅助分析工具，重点从方法的层面上对模糊Petri网理论进行了研究和探讨。旨在进一步完善模糊Petri网相关理论，并应用于制造过程质量问题的解决，提高过程的质量监控能力、事故预防能力、缩短故障原因查找周期、提高定位准确性及可靠性奠定方法基础。 针对制造过程质量预测与诊断问题特点，在广泛阅读相关文献并深入探索的基础上，对模糊Petri网理论方法进行了较深入的研究和探讨，重点解决了以下问题： 1）模糊Petri网自动建模方法：对模糊Petri网理论研究的基础和前提是建立模糊Petri网模型。为解决当前模糊Petri网建模效率低、工作量大、易出错等问题，本文提出了模糊Petri网的自动建模方法。该方法的提出，易于保证知识库与模型库更新的同步和一致，提高了建模效率，避免了建模的人为失误。 2) 模糊Petri网参数确定：模型建立后，为实现可靠有效地推理，需进行相关参数的确定。提出了确定模糊Petri网的初始库所token的方法。通过模糊统计的方法来获得模糊token，减少确定token时的主观臆断性和不一致性，为物理量与模糊token的实时转换提供了技术支持。由于构建符合客观实际的、连续的隶属函数是确定模糊token的前提条件，本文提出采用最小二乘拟合来构造模糊隶属函数方法。该方法简单，拟合能力强，人工干预少。由于变迁阈值影响推理的正确性及可靠性，这里对阈值设定进行了初步探讨。阈值设定越高，预测及诊断的漏报率越高；反之，误报率越高。给出了阈值设定的总代价计算式，阈值选择的目标是使总代价最小。 在建立了模糊Petri网模型、确定了相关参数后，便可对异常事件进行预测及诊断推理。 3）模糊Petri网预测方法：对预测模式进行了分类与定义，便于对不同模式下进行预测分析。提出了改进的FPN四种基本推理模型，通过禁止弧的引入，避免了激发过的变迁反复被激发，减少不必要的计算，实现了推理与模型结构的一致性。从而提高了推理效率和基于规则系统的响应能力。 4）模糊Petri网诊断方法：给出了一种模糊Petri网诊断推理方法。该方法充分利用模糊Petri网自身的结构与数学特性这一突出优势，实现了并行推理。以矢量计算方式获得中间库所能力，取代了常规的搜索方式，提高了推理效率。通过引入人机交互的处理策略，减少了模糊Petri网的复杂性及规模。指出在实践中，推理方法的效率、成本及实际的应用效果， 在重要性方面，要远大于方法自身的运算效率。 5）模糊着色Petri网推理方法：在建模复杂大型系统时，为解决模糊Petri网存在模型空间过大，模型数据结构松散等问题，提出了FCPN并行推理方法及FPR与FCPN模型转换算法。提出的FCPN与现有方法的主要区别在以下方面：首先，算法实现变迁的单次激发，避免推理激发变迁的重复计算。其次，某个使能变迁前集库所中token在该变迁激发后并不移除，符合实际推理情况。此外，通过输入/输出关联矩阵计算迭代，实现了并行推理。 最后，以一典型制造过程—埋弧自动焊接过程质量问题的预测和诊断为例，来说明模糊Petri网方法的实际应用。通过系统的实现，验证了相应方法是可行的。通过模糊Petri网的预测及诊断推理，便于实现质量异常的分析、预警、处理、过程控制及数字化管理，为生产策略的调整、纠正措施的采取提供了决策依据，加快了系统响应速度。 本文研究工作重点围绕模糊Petri网理论方法展开，虽以制造过程质量问题的预测与诊断为研究和应用背景，但并不局限于该领域，是属于具有一般性的共性方法。因此，所开展的方法研究工作具有良好的科研价值和广泛的应用前景。

Exploring alternatives for milk quality improvement and more efficient dairy production in a smallholder farming context – case study: Mantaro Valley (Peru)

Consumer demand is revolutionizing the way products are being produced, distributed and marketed. In relation to the dairy sector in developing countries, aspects of milk quality are receiving more attention from both society and the government. However, milk quality management needs to be better addressed in dairy production systems to guarantee the access of stakeholders, mainly small-holders, into dairy markets. The present study is focused on an analysis of the interaction of the upstream part of the dairy supply chain (farmers and dairies) in the Mantaro Valley (Peruvian central Andes), in order to understand possible constraints both stakeholders face implementing milk quality controls and practices; and evaluate “ex-ante” how different strategies suggested to improve milk quality could affect farmers and processors’ profits. The analysis is based on three complementary field studies conducted between 2012 and 2013. Our work has shown that the presence of a dual supply chain combining both formal and informal markets has a direct impact on dairy production at the technical and organizational levels, affecting small formal dairy processors’ possibilities to implement contracts, including agreements on milk quality standards. The analysis of milk quality management from farms to dairy plants highlighted the poor hygiene in the study area, even when average values of milk composition were usually high. Some husbandry practices evaluated at farm level demonstrated cost effectiveness and a big impact on hygienic quality; however, regular application of these practices was limited, since small-scale farmers do not receive a bonus for producing hygienic milk. On the basis of these two results, we co-designed with formal small-scale dairy processors a simulation tool to show prospective scenarios, in which they could select their best product portfolio but also design milk payment systems to reward farmers’ with high milk quality performances. This type of approach allowed dairy processors to realize the importance of including milk quality management in their collection and manufacturing processes, especially in a context of high competition for milk supply. We concluded that the improvement of milk quality in a smallholder farming context requires a more coordinated effort among stakeholders. Successful implementation of strategies will depend on the willingness of small-scale dairy processors to reward farmers producing high milk quality; but also on the support from the State to provide incentives to the stakeholders in the formal sector.

Infant milk formula manufacture: process and compositional interactions in high dry matter wet-mixes

Infant milk formula (IMF) is fortified milk with composition based on the nutrient content in human mother's milk, 0 to 6 months postpartum. Extensive medical and clinical research has led to advances in the nutritional quality of infant formula; however, relatively few studies have focused on interactions between nutrients and the manufacturing process. The objective of this research was to investigate the impact of composition and processing parameters on physical behaviour of high dry matter (DM) IMF systems with a view to designing more sustainable manufacturing processes. The study showed that commercial IMF, with similar compositions, manufactured by different processes, had markedly different physical properties in dehydrated or reconstituted state. Commercial products made with hydrolysed protein were more heat stable compared to products made with intact protein, however, emulsion quality was compromised. Heat-induced denaturation of whey proteins resulted in increased viscosity of wet-mixes, an effect that was dependant on both whey concentration and interactions with lactose and caseins. Expanding on fundamental laboratory studies, a novel high velocity steam injection process was developed whereby high DM (60%) wet-mixes with lower denaturation/viscosity compared to conventional processes could be achieved; powders produced using this process were of similar quality to those manufactured conventionally. Hydrolysed proteins were also shown to be an effective way of reducing viscosity in heat-treated high DM wet-mixes. In particular, using a whey protein concentrate whereby β-Lactoglobulin was selectively hydrolysed, i.e., α-Lactalbumin remained intact, reduced viscosity of wet-mixes during processing while still providing good emulsification. The thesis provides new insights into interactions between nutrients and/or processing which influence physical stability of IMF both in concentrated liquid and powdered form. The outcomes of the work have applications in such areas as; increasing the DM content of spray drier feeds in order to save energy, and, controlling final powder quality.

Risk mitigation framework for a robust design process

The increasing complexity of new manufacturing processes and the continuously growing range of fabrication options mean that critical decisions about the insertion of new technologies must be made as early as possible in the design process. Mitigating the technology risks under limited knowledge is a key factor and major requirement to secure a successful development of the new technologies. In order to address this challenge, a risk mitigation methodology that incorporates both qualitative and quantitative analysis is required. This paper outlines the methodology being developed under a major UK grand challenge project - 3D-Mintegration. The main focus is on identifying the risks through identification of the product key characteristics using a product breakdown approach. The assessment of the identified risks uses quantification and prioritisation techniques to evaluate and rank the risks. Traditional statistical process control based on process capability and six sigma concepts are applied to measure the process capability as a result of the risks that have been identified. This paper also details a numerical approach that can be used to undertake risk analysis. This methodology is based on computational framework where modelling and statistical techniques are integrated. Also, an example of modeling and simulation technique is given using focused ion beam which is among the investigated in the project manufacturing processes.

Textural properties of whey based edible films

The objective of this work was to study the textural properties of edible films made from sour (acid) whey for food wrapping application. Acid whey is often regarded as a waste product, obtained as a watery effluent in the manufacturing of cottage cheese. In general, owing to its high nutritional value, whey has gained importance as an additive in food manufacturing processes and in health drink formulations. In this work, fresh sour whey was used to make edible films. The proteins in the whey were concentrated by ultrafiltration to reduce the water content. Only natural ingredients such as acid whey and agar were used to form the film under controlled heating (650 W) in a microwave oven. The structural and surface characteristics of the films were tested by a texture analyser and scanning electron micrographs.

Performance enhancement of polymer nanocomposites via multiscale modelling of processing and properties

This paper provides an overview of research on modelling of the structure–property interactions of polymer nanocomposites in manufacturing processes (stretch blow moulding and thermoforming) involving large-strain biaxial stretching of relatively thin sheets, aimed at developing computer modelling tools to help producers of materials, product designers and manufacturers exploit these materials to the full, much more quickly than could be done by experimental methods alone. The exemplar systems studied are polypropylene and polyester terephalate, with nanoclays. These were compounded and extruded into 2mm thick sheet which was then biaxially stretched at 155°C for the PP and 90 to 100°C for the PET. Mechanical properties were determined for the unstretched and stretched materials, together with TEM and XRD studies of structure. Multi-scale modelling, using representative volume elements is used to model the properties of these products.

MIXED-MODEL PRODUCTION SYSTEM DESIGN FOR AIRCRAFT ASSEMBLY

With the advancement of flexible fixture and flexible tooling, mixed production has become possible for aircraft assembly as the manufacturing processes of different aircraft/sub-assembly models are similar. However, it is a great challenge to model the problem and provide a practical solution due to the low volume and complex constraints of aircraft assemblies. To tackle this problem, this work proposes a methodology for designing the mixed production system, and a new scheduling approach is proposed by combined backward and forward scheduling methods. These methods are validated through a real-life industrial case study. Simulation results show that the number of workstations and the cycle time for making a fuselage can be reduced by 50% and 39% respectively with the newly designed mixed-model system.

Implementation of a non-orthogonal model for the finite element simulation of textile composite draping

In the pursuit of producing high quality, low-cost composite aircraft structures, out-of-autoclave manufacturing processes for textile reinforcements are being simulated with increasing accuracy. This paper focuses on the continuum-based, finite element modelling of textile composites as they deform during the draping process. A non-orthogonal constitutive model tracks yarn orientations within a material subroutine developed for Abaqus/Explicit, resulting in the realistic determination of fabric shearing and material draw-in. Supplementary material characterisation was experimentally performed in order to define the tensile and non-linear shear behaviour accurately. The validity of the finite element model has been studied through comparison with similar research in the field and the experimental lay-up of carbon fibre textile reinforcement over a tool with double curvature geometry, showing good agreement.

Influence of temperature on the anisotropic cutting behaviour of single crystal silicon: A molecular dynamics simulation investigation

Using molecular dynamics (MD) simulation, this paper investigates anisotropic cutting behaviour of single crystal silicon in vacuum under a wide range of substrate temperatures (300 K, 500 K, 750 K, 850 K, 1173 K and 1500 K). Specific cutting energy, force ratio, stress in the cutting zone and cutting temperature were the indicators used to quantify the differences in the cutting behaviour of silicon. A key observation was that the specific cutting energy required to cut the (111) surface of silicon and the von Mises stress to yield the silicon reduces by 25% and 32%, respectively, at 1173 K compared to what is required at 300 K. The room temperature cutting anisotropy in the von Mises stress and the room temperature cutting anisotropy in the specific cutting energy (work done by the tool in removing unit volume of material) were obtained as 12% and 16% respectively. It was observed that this changes to 20% and 40%, respectively, when cutting was performed at 1500 K, signifying a very strong correlation between the anisotropy observed during cutting and the machining temperature. Furthermore, using the atomic strain criterion, the width of primary shear zone was found to vary with the orientation of workpiece surface and temperature i.e. it remains narrower while cutting the (111) surface of silicon or at higher machining temperatures. A major anecdote of the study based on the potential function employed in the study is that, irrespective of the cutting plane or the cutting temperature, the state of the cutting edge of the diamond tool did not show direct diamond to graphitic phase transformation.

Influence of impurities and catalyst surface characteristics on the oxygen charge transfer reaction in the Pt/YSZ system

Spillover processes (i.e. the migration of ionic species from the support to the catalyst and vice versa) are known to play a very important role in catalysis and electrocatalysis. These spillover processes can be influenced by impurities (pre-existing on the catalyst surface) and by the catalyst morphology that may differ as a result of the differences in catalyst manufacturing processes. This work investigates the influence of impurities present in three commercial platinum (Pt) precursors. The resulting platinum films studied here were supported on yttria-stabilised-zirconia (YSZ). It was found that the three different catalyst films contained a range of impurities (determined by ICP-OES) that appear to affect the oxygen charge transfer reaction as studied by cyclic voltammetry (CV). © 2012 Elsevier B.V.

Advances in the surface defect machining (SDM) of hard steels

This paper reports the realisation of precision surface finish (Ra 30 nm) on AISI 4340 steel using a conventional turret lathe by adapting and incorporating a surface defect machining (SDM) method [Wear, 302, 2013 (1124-1135)]. Conventional ways of machining materials are limited by the use of a critical feed rate, experimentally determined as 0.02 mm/rev, beyond which no appreciable improvement in the machined quality of the surface is obtained. However, in this research, the novel application of an SDM method was used to overcome this minimum feed rate limitation ultimately reducing it to 0.005 mm/rev and attaining an average machined surface roughness of 30 nm. From an application point of view, such a smooth finish is well within the values recommended in the ASTM standards for total knee joint prosthesis. Further analysis was done using SEM imaging, white light interferometry and numerical simulations to verify that adapting SDM method provides improved surface integrity by reducing the extent of side flow, microchips and weldments during the hard turning process.

Understanding aerospace composite components' supply chain carbon emissions

This paper examines a large structural component and its supply chain. The component is representative of that used in the production of civil transport aircraft and is manufactured from carbon fibre epoxy resin prepreg, using traditional hand layup and autoclave cure. Life cycle assessment (LCA) is used to predict the component’s production carbon emissions. The results determine the distribution of carbon emissions within the supply chain, identifying the dominant production processes as carbon fibre manufacture and composite part manufacture. The elevated temperature processes of material and part creation, and the associated electricity usage, have a significant impact on the overall production emissions footprint. The paper also demonstrates the calculation of emissions footprint sensitivity to the geographic location and associated energy sources of the supply chain. The results verify that the proposed methodology is capable of quantitatively linking component and supply chain specifics to manufacturing processes and thus identifying the design drivers for carbon emissions in the manufacturing life of the component.

Microneedle Technology

Microneedles (MNs) are minimally invasive devices consisting of numerous micron-sized projections amassed on a baseplate, designed to enhance transdermal drug delivery. When applied to the skin, the needles puncture the outermost layer, the stratum corneum, forming aqueous conduits through which drugs can diffuse to the dermal microcirculation. With an average length of 50-900 μm, MNs are short enough to avoid stimulation of dermal nerves and do not induce bleeding, yet gain access to the skin's rich microcirculation for drug delivery. MNs have been extensively investigated for drug and vaccine delivery, demonstrating their efficacy at increasing the number of compounds amenable to delivery through the skin. This chapter discusses the materials and fabrication methods involved in MN production, alongside the different types of MN arrays and their delivery capabilities. The field has expanded to consider novel applications of MNs including minimally invasive patient monitoring, ocular delivery and enhanced administration of cosmeceuticals. Patient usage and effects on the skin are also considered in terms of safety, efficacy and acceptability. The next steps in MN development are to focus on the scale-up of manufacturing processes, a challenge considering the number of small-scale methods detailed in the literature. Regulatory guidance is awaited to direct this, alongside provision of clearer patient instruction for safe and effective use of MN devices. MNs have tremendous potential to yield real benefits for patients and industry and with continued research in the key areas highlighted, this will begin to be realised over the next number of years.

Resource Efficient Configuration of an Aircraft Assembly Line

The design of efficient assembly systems can significantly contribute to the profitability of products and the competitiveness of manufacturing industries. The configuration of a an efficient assembly line can be supported by suitable methodologies and techniques, such as design for manufacture and assembly, assembly sequence planning, assembly line balancing, lean manufacturing and optimization techniques. In this paper, these methods are applied with reference to the industrial case study of the assembly line of a Skycar light aircraft. The assembly process sequence is identified taking into account the analysis of the assembly structure and the required precedence constraints, and diverse techniques are applied to optimize the assembly line performance. Different line configurations are verified through discrete event simulation to assess the potential increase of efficiency and throughput in a digital environment and propose the most suitable configuration of the assembly line.

Modelling and aerodynamic simulation of a vehicle and failure analysis of a laminated front fender

Master Thesis in Mechanical Engineering field of Maintenance and Production