Prediction of die casting process parameters by using an artificial neural network model for zinc alloys

Pressure die casting is an important production process. In pressure die casting, the first setting of process parameters is established through guess work. Experts use their previous experience and knowledge to develop a solution for a new application. Due to rapid expansion in the die casting process to produce better quality products in a short period of time, there is ever increasing demand to replace the time-consuming and expert-reliant traditional trial and error methods of establishing process parameters. A neural network system is developed to generate the process parameters for the pressure die casting process. The system aims to replace the existing high-cost, time-consuming and expertdependent trial and error approach for determining the process parameters. The scope of this work includes analysing a physical model of the pressure die casting filling stage based on governing equations of die cavity filling and the collection of feasible casting data for the training of the network. The training data were generated by using ZN-DA3 material on a hot chamber die casting machine with a plunger diameter of 60 mm. The present network was developed using the MATLAB application toolbox. In this work, the neural network was developed by comparing three different training algorithms: i.e. error backpropagation algorithm; momentum and adaptive learning algorithm; and Levenberg-Marquardt approximation algorithm. It was found that the Levenberg-Marquardt approximation algorithm was the preferred method for this application as it reduced the sum-squared error to a small value. The accuracy of the developed network was tested by comparing the data generated from the network with those of an expert from a local die casting industry. It was established that by using this network the selection of process parameters becomes much easier, so that it can be used by a novice user without prior knowledge of the die casting process or optimization techniques.     

记忆径向基神经网络在冷轧液压自动位置系统的优化控制

针对冷轧液压自动位置控制系统多变量、强耦合、高阶次和时变性等特点,提出一种引入记忆因子的径向基函数神经网络在线自适应调节PID参数的系统。为提高网络精度,利用改进的混洗蛙跳算法离线全优化记忆径向基神经网络,在获得网络结构的同时得到初始参数,避免网络模型训练的繁琐,并利用测试函数证明优化后的网络具有良好的逼近能力。然后利用优化后记忆径向基神经网络的自校正功能在线细调PID参数,仿真结果表明,该控制系统跟踪快、超调小、适应性强,控制品质优于传统PID和普通径向基神经网络PID控制方法。     

A neural-network approach to describe the scatter of cyclic stress–strain curves

In order to predict a product’s durability in the early phases of development it is necessary to know the stress–strain behaviour of the material, its resistance to fatigue and the loading states in the material. These parameters, however, tend to exhibit a considerable degree of uncertainty. Due to a lack of knowledge of the actual circumstances in which the product is used, during the early development phase, simulations based on statistical methods are used. The results of the experiments show that the cyclic stress–strain curves demonstrate not only a large amount of scatter, but also a dependence on the temperature, the size of the cross-section, the content of alloying elements, the loading rate, etc.This article presents a method for modelling cyclic stress–strain curve scatter using a hybrid neural network for an arbitrary selection of the influencing factors. In an example of the measured data for a high pressure die-cast aluminium alloy it is clear that the suggested method is suitable for describing cyclic stress–strain curves. The main advantage of a hybrid neural network in comparison with a conventional method is the neural network’s ability to precisely describe the influence of various factors, and their combinations, based on the form and scatter of the cyclic stress–strain curve families. Defining the model parameters, i.e., training the neural network, is a procedure that does not require any additional user interventions; however, it enables us to gather knowledge that would otherwise require a lot of research. Thus, the trained neural network is a robust tool that can be used to predict cyclic stress–strain curves for random values of influencing factors. The capabilities of the presented method are only limited by the quantity of the measured data used for the neural-network training.     

Neural networks and modelling in vacuum science

The paper is an attempt to describe how neural networks may be used as an approximation-modelling tool. A brief survey of the evolution of the approximation theory and neural networks is presented. Practical applications are based on modelling of vacuum science problems, especially the modelling of a cold cathode pressure gauge. The problem of approximation of wide range functions, that are one of the characteristics of vacuum science problems, is introduced. Parameters such as pressure or cathode current span over several decades and neural networks are not suitable for any approximation of such functions; therefore, two strategies need to be introduced, and these are described. The approximation made by the neural network is obtained by the training process. The models obtained by several independent repetitions of training processes performed on the same training set lead to slightly different results. Therefore the definition of training stability is introduced and described. Finally, some practical hints regarding the neural network synthesis (design) are given.     

基于神经网络的组合预报模型及其应用

时间序列预报是神经网络的一个应用领域,多数研究集中在神经网络直接预报方面。本文从故障预报角度研究了神经网络组合预报模型,由神经网络给出常规预报方法的最佳组合。首先从函数逼近角度阐述这种模型的理论依据,在此基础上给出了模型的评价指标和神经网络的有效训练算法,最后给出在空间推进系统上的应用实例。     

矿渣微晶玻璃材料设计神经网络模型

研究了人工神经网络在矿渣微晶玻璃材料设计中的应用。采用基于变尺度法的新学习算法建立了三层前馈型神经网络,发现当网络结构为M-2M-1,取一定范围内的学习误差时,网络具有很好的学习效果。研究证明,建立的人工神经网络模型学习速度快,收敛稳定,强壮性好,能根据较少的实验样本有效抽取矿渣微晶玻璃组成、工艺和性能之间的内在规律,是进行微晶玻璃材料设计的有力工具。     

Determination of deformation and failure properties of ductile materials by means of the small punch test and neural networks

This paper describes an approach to identify plastic deformation and failure properties of ductile materials. The experimental method of the small punch test is used to determine the material response under loading. The resulting load displacement curve is transferred to a neural network, which was trained using load displacement curves generated by finite element simulations of the small punch test and the corresponding material parameters. The simulated material behavior of the specimen is based on the ductile elastoplastic damage theory of Gurson, Tvergaard and Needleman. During a training process the neural network generates an approximated function for the inverse problem relating the material parameters to the shape of the load displacement curve of the small punch test. This technique was tested for three different materials (ductile steels). The identified parameters are verified by testing and simulating notched tensile specimens.     

A neural network approach for the prediction of the refractive index based on experimental data

This article presents a systematic approach for correlating the refractive index of different material kinds and forms with experimentally measured inputs like wavelength, temperature, and concentration. The correlation is accomplished using neural network models, which can deal effectively with the nonlinear nature of the problem without requiring a predefined form of equation, while taking into account all the parameters affecting the refractive index. The proposed methodology employs the powerful radial basis function network architecture and the neural network training procedure is accomplished using an innovative algorithm, which provides results with increased prediction accuracy. The methodology is applied to two cases, involving the estimation of the refractive index of semiconductor material crystals and an ethanol–water mixture and the results show that the refractive index predictions are accurate approximately to the same number of decimal places as the real measurements. Comparisons with other neural network training methods, but also with empirical forms like the Sellmeier equation, highlight the superiority of the proposed approach.     

Neural Network Mapping of Corrosion Induced Chemical Elements Degradation in Aircraft Aluminum

A neural network (NN) model is developed for the analysis and prediction of the mapping between degradation of chemical elements and electrochemical parameters during the corrosion process. The input parameters to the neural network model are alloy composition, electrochemical parameters, and corrosion time. The output parameters are the degradation of chemical elements in AA 2024-T3 material. The NN is trained with the data obtained from Energy Dispersive X-ray Spectrometry (EDS) on corroded specimens. A very good performance of the neural network is achieved after training and validation with the experimental data. After validating the NN model, simulations were carried out to obtain the trends in element degradation with varying pH values, and the results showed correct trends. The preliminary results obtained demonstrate that through a comprehensive study, a better corrosion resistant material can be designed by controlling the degradation of the chemical elements during the corrosion process through neural network methods.     

基于GA-BP神经网络的结构损伤位置识别

针对传统BP神经网络训练中存在的一些问题，提出了一种基于遗传算法(GA)-BP神经网络混合技术识别结构损伤位置的方法。该方法利用基因实数编码的遗传算法优化BP网络的结构及初始参数，从而大大提高了神经网络的训练精度。运用GA-BP网络与传统BP网络技术分别对两个算例进行了结构损伤定位的识别仿真，结果表明遗传BP稳定性好，精度高，对噪声有很好的鲁棒性，便于工程应用。     

可调激活函数神经元参数选取方法研究

隐层神经元采用相同的Sigmoid激活函数会限制神经网络的非线性能力,对Sigmoid函数引入两个参数可改善其响应特性,增强其非线性逼近能力.本文给出了一种可调Sigmoid激活函数,分析了可调激活函数中参数所表示的几何意义;给出提升网络维数的可调激活函数中参数的快速选取方法和理论基础.这为人们在采用可调Sigmoid激活函数解决实际问题时,如何快速选取激活函数中的参数提供了一种可行方法.     

基于神经网络的三角形网格智能重建

探讨了曲面密集三维散乱点数据的三角网格智能重建方法。建立了基于自组织特征映射神经网络的三角网格构建模型。该模型利用神经元对曲面散乱点的学习和训练来模拟曲面上的点与点之间的内在关系,结点连接权矢量集作为对散乱点集的工程近似化并重构曲面样本点的内在拓扑关系,实现曲面密集三维散乱点数据的自组织压缩。按六角形阵列侧抑制邻区训练调整网络神经元权重矢量,使网络输出层结点呈六角形阵列分布,可实现测量点集压缩后的Delaunay三角逼近剖分。计算机仿真实验表明,所建神经网络模型可以实现期望规模和精度的三角网格剖分并有效保持原数据点集的拓扑特征。     

旋转机械故障诊断的神经网络方法研究

BP神经网络具有较好的非线性映射能力,可以描述频率特征和故障之间的关系,而概率神经网络学习规则简单、训练速度快、避免局部极小和反复训练的问题。根据两种神经网络的原理选择合适的参数建立两个旋转机械故障诊断模型,并利用模型对某旋转机械的故障数据进行处理,结果显示两种网络在故障诊断方面的实用价值。通过对故障数据的结果对比可以看到PNN网络比BP网络具有更好的容错能力。     

Robustness analysis of radial base function and multi-layered feed-forward neural network models

In this paper, two popular types of neural network models (radial base function (RBF) and multi-layered feed-forward (MLF) networks) trained by the generalized delta rule, are tested on their robustness to random errors in input space. A method is proposed to estimate the sensitivity of network outputs to the amplitude of random errors in the input space, sampled from known normal distributions. An additional parameter can be extracted to give a general indication about the bias on the network predictions. The modelling performances of MLF and RBF neural networks have been tested on a variety of simulated function approximation problems. Since the results of the proposed validation method strongly depend on the configuration of the networks and the data used, little can be said about robustness as an intrinsic quality of the neural network model. However, given a data set where ‘pure’ errors from input and output space are specified, the method can be applied to select a neural network model which optimally approximates the nonlinear relations between objects in input and output space. The proposed method has been applied to a nonlinear modelling problem from industrial chemical practice. Since MLF and RBF networks are based on different concepts from biological neural processes, a brief theoretical introduction is given.     

Einfluss der Netztopologie der künstlich neuronalen Netze auf das Ergebnis der Abschätzung zyklischer Kennwerte

Several estimation methods have been developed to estimate the cyclic material parameters out of the static material properties. Most of these methods are based on empirical equations. Increasing numbers of input‐ and influencing parameters lead to an rising effort for determining these equations and the accuracy decreases. For this reason new suitable methods are sought to estimate the cyclic material behaviour. A very promising approach is the application of the artificial neural networks, which can derive self‐depended a relationship between in‐ and output parameters. Static parameters such as yield strength, tensile strength …? etc., which can rapidly be determined used as input parameters. The output parameters are the cyclic material parameters of the strain‐life curve and stress‐strain curve according to the Manson‐Coffin‐Basquin‐ and Ramberg‐Osgood curve. Many different artificial neural networks with different structures and complexity can be applied. In this paper the influence of the topology of an artificial neural network on the estimation accuracy will be investigated. Based on the results of a reference artificial neural network it will be shown, that more complex topologies in the network do not lead inevitably to better estimations.     

小波神经网络在超声检测信号中降噪的研究

摘 要: 为了最大限度的消除粗晶材料超声检测时,晶粒散射波对有用信号的严重干扰,提高接收信号的信噪比,将小波神经网络引入粗晶材料超声检测信号处理领域中。在训练小波神经网络时,采用了改进的梯度下降算法。该网络有一个动态的权值,它随误差变化而调整。结果表明,小波神经网络应用在粗晶材料超声检测信号的降噪时,能够达到较理想的降噪效果。     

Neural modeling of dynamic systems with nonmeasurable statevariables

The paper studies the ability possessed by recurrent neural networks to model dynamic systems when some relevant state variables are not measurable. Neural architectures based on virtual states-which naturally arise from a space state representation-are introduced and compared with the more traditional neural output error ones. Despite the evident potential model ability possessed by virtual state architectures we experimented that their performances strongly depend on the training efficiency. A novel validation criterion for neural output error architectures is suggested which allows to assess the neural network not only in terms of its approximation accuracy but also with respect to stability issues     

基于神经网络的贮仓结构参数识别

利用神经网络技术，提出了识别结构物理参数的一种方法。用单元刚度矩阵基本值和模态应变能来选择基本模态，用修正的Latin超立方采样技术和模态准入准则来产生网络的输入数据。贮仓在动载作用下的自振频率和模态作为网络的输入，子矩阵参与系数作为网络的输出，用Levenberg-Marquardt算法训练网络。仿真计算表明，方法是可行的。     

基于卷积神经网络的水下多目标方位估计方法

针对水下多目标方位估计问题,提出了一种利用卷积神经网络模型对目标声源进行方位估计的方法。该方法使用不等强度的声源数据进行训练并使用焦点损失函数作为训练损失函数。通过对阵列接收到的信号进行特征提取,使用焦点损失函数指导卷积神经网络训练,最终利用训练好的卷积神经网络模型进行目标方位估计。对不同模型参数的训练进行对比,结果表明所训练的卷积神经网络模型在较低信噪比条件下也能正确估计弱目标的方位。试验结果表明,与采用二元交叉熵损失函数的卷积神经网络模型相比,该方法对弱目标的方位估计能力更强,提高了方位估计的准确率。     

NEW REGULARIZATION BY TRANSFORMATION FOR NEURAL NETWORK BASED INVERSE ANALYSES AND ITS APPLICATION TO STRUCTURE IDENTIFICATION

The present authors have been developing an inverse analysis approach using the multilayer neural network and the computational mechanics. This approach basically consists of the following three subprocesses. First, parametrically varying model parameters of a system, their corresponding responses of the system are calculated through computational mechanics simulations such as the finite element analyses, each of which is an ordinary direct analysis. Each data pair of model parameters vs. system responses is called training pattern. Second, a neural network is iteratively trained using a number of training patterns. Here the system responses are given to the input units of the network, while the model parameters to be identified are shown to the network as teacher data. Finally, some system responses measured are given to the well-trained network, which immediately outputs appropriate model parameters even for untrained patterns. This is an inverse analysis. This paper proposes a new regularization method suitable for the inverse analysis approach mentioned above. This method named the Generalized-Space-Lattice (GSL) transformation transforms original input and/or output data points of all training patterns onto uniformly spaced lattice points over a multi-dimensional space. The topological relationships among all the data points are maintained through this transformation. The neural network is then trained using the GSL-transformed training patterns. Since this method significantly remedies localization of training patterns caused due to strong nonlinearity of problem, the neural network can learn the training patterns efficiently as well as accurately. Fundamental performances of the present inverse analysis approach combined with the GSL transformation are examined in detail through the identification of a vibrating non-uniform beam in Young's modulus based on the observation of its multiple eigenfrequencies and eigenmodes.