Simulation optimisation of pull control policies for serial manufacturing lines and assembly manufacturing systems using genetic algorithms

Several efficient pull production control policies for serial lines implementing the lean/JIT manufacturing philosophy can be found in the production management literature. A recent development that is less well-studied than the serial line case is the application of pull-type policies to assembly systems where manufacturing operations take place both sequentially and in parallel. Systems of this type contain assembly stations where two or more parts from lower hierarchical manufacturing stations merge in order to produce a single part of the subsequent stage. In this paper we extend the application of the Base Stock, Kanban, CONWIP, CONWIP/Kanban Hybrid and Extended Kanban production control policies to assembly systems that produce final products of a single type. Discrete-event simulation is utilised in order to evaluate the performance of serial lines and assembly systems. It is essential to determine the best control parameters for each policy when operating in the same environment. The approach that we propose and probe for the problem of control parameter selection is that of a genetic algorithm with resampling, a technique used for the optimisation of stochastic objective functions. Finally, we report our findings from numerical experiments conducted for two serial line simulation scenarios and two assembly system simulation scenarios.     

First-train timing synchronisation using multi-objective optimisation in urban transit networks

Missed transfers affect urban transportation by increasing the travel times and decreasing the travel possibility, especially in the case of longer headways. A synchronised timetable can improve the transport efficiency of urban mobility and become an important consideration in the operation of urban transit networks (UTN). A mixed integer programming model is proposed to generate an optimal train timetable and minimise the total connection time, which includes smooth synchronisations for rail first-trains and the seamless synchronisation from rail first-trains to the bus service. Meanwhile, to characterise the characteristics of first-trains, binary variables are used to denote key transfer directions. Subsequently, the Sub-network Connection Method in conjunction with Genetic Algorithm is designed to obtain near-optimal solutions in an efficient way. Finally, a real-world case study, 16 rail lines and 41 transfer stations, based on the Beijing metro network and travel demand is conducted to validate the proposed timetabling model. Preliminary numerical results show that our approach improves the synchronisation substantially compared with the currently operated timetable.     

Optimization of principal-component-analysis-applied in situ spectroscopy data using neural networks and genetic algorithms

A new model of multidimensional in situ diagnostic data is presented. This was accomplished by combining a back-propagation neural network (BPNN), principal component analysis (PCA), and a genetic algorithm (GA). The PCA was used to reduce input dimensionality. The GA was applied to search for a set of optimized training factors involved in BPNN training. The presented technique was evaluated with optical emission spectroscopy (OES) data measured during the etching of oxide thin films in a CHF(3)-CF(4) inductively coupled plasma. For a systematic modeling, the etching process was characterized by a face-centered Box Wilson experiment. The etch responses to be modeled include oxide etch rate, oxide profile angle, and oxide etch rate non-uniformity. In PCA, three types of data variances were employed and the reduced input dimensionality corresponding to 100, 99, and 98% are 16, 8, and 5. The BPNN training factors to be optimized include the training tolerance, number of hidden neurons, magnitude of initial weight distribution, gradient of bipolar sigmoid function, and gradient of linear function. The prediction errors of GA-BPNN models are 249 A/min, 2.64 degrees, and 0.439% for the etch rate, profile angle, and etch rate non-uniformity, respectively. Compared to the conventional and previous full OES models, the presented models demonstrated a significantly improved prediction for all etch responses.     

Control point policy optimization using genetic algorithms

This paper describes the application of an integrated Genetic Algorithm (GA)/Discrete Event Simulation model for selecting optimum values for Critical Point Policy (CPP) hedging time and buffer size parameters. The CPP is shown to perform well, when compared with the Critical Ratio priority rule, in terms of improving service levels, particularly when subject to conditions where buffer sizes and Takt times are required to be small. The technique developed involves buffer sizes being chosen by a GA according to a constraint on the total storage space available within the system. A method is described for reducing the number of variables that the GA needs to deal with, hence, improving the efficiency of the GA optimization process. The development and application work reported also provides further understanding into how and when the CPP should be applied.     

Tools for intelligent control: fuzzy controllers,neural networks and genetic algorithms

The goal of this expository paper is to bring forth the basic current elements of soft computing (fuzzy logic, neural networks, genetic algorithms and genetic programming) and the current applications in intelligent control. Fuzzy sets and fuzzy logic and their applications to control systems have been documented. Other elements of soft computing, such as neural networks and genetic algorithms, are also treated for the novice reader. Each topic will have a number of relevant references of as many key contributors as possible.     

Optimization of composite stiffened panels under mechanical and hygrothermal loads using neural networks and genetic algorithms

The present work develops an optimization procedure for a geometric design of a composite material stiffened panel with conventional stacking sequence using static analysis and hygrothermal effects. The procedure is based on a global approach strategy, composed by two steps: first, the response of the panel is obtained by a neural network system using the results of finite element analyses and, in a second step, a multi-objective optimization problem is solved using a genetic algorithm. The neural network implemented in the first step uses a sub-problem approach which allows to consider different temperature ranges. The compression load and relative humidity of the air are assumed to be constants throughout the considered temperature range.     

A predictive model for slip resistance using artificial neural networks

This paper describes the formulation, building and validation of an artificial neural network model of the dynamic coefficient of friction (DCOF) as measured by a slip resistance testing device. The model predicts the DCOF as a function of six independent variables over a wide range of conditions. A grouped cross validation method is used to show the consistent performance of the model in predicting the DCOF for new values of the independent variables.     

Nonlinear constitutive models for FRP composites using artificial neural networks

This paper presents a new approach to generate nonlinear and multi-axial constitutive models for fiber reinforced polymeric (FRP) composites using artificial neural networks (ANNs). The new nonlinear ANN constitutive models are complete and have been integrated with displacement-based FE software for the nonlinear analysis of composite structures. The proposed ANN constitutive models are trained with experimental data obtained from off-axis tension/compression and pure shear (Arcan) tests. The proposed ANN constitutive model is generated for plane–stress states with assumed functional response in some parts of the multi-axial stress space with no experimental data. The ability of the trained ANN models to predict material response is examined directly and through FE analysis of a notched composite plate. The experimental part of this study involved coupon testing of thick-section pultruded FRP E-glass/polyester material. Nonlinear response was pronounced including in the fiber direction due to the relatively low overall fiber volume fraction (FVF). Notched composite plates were also tested to verify the FE, with ANN material models, to predict general non-homogeneous responses at the structural level.     

量子神经计算和量子遗传算法的理论分析和应用

经过比较研究发现,在量子计算与神经网络和遗传算法之间,不论在计算思想上还是模型表达上,都存在着许多相似之处,这些相似性启发人们去研究基于量子理论的神经网络和遗传算法模型,一方面探索神经网络和遗传算法在量子系统上的实现方法,另一方面研究量子理论启发下的新的神经网络与遗传算法模型。本文总结了本课题组近年来在量子计算与神经网络和遗传算法相结合领域的研究工作,包括量子系统实现神经计算的理论分析,量子神经网络物理模型的研究,基于量子概率表达的量子遗传算法及其应用研究等,并对今后的发展提出了展望。     

Processing wave-front-sensor slope measurements using artificial neural networks

For adaptive-optics systems to compensate for atmospheric turbulence effects, the wave-front perturbation must be measured with a wave front sensor (WPS), and key parameters of the atmosphere and the adaptive-optics system must be known. Two parameters of particular interest include the Fried coherence length r(0) and the WPS slope measurement error. Statistics-based optimal techniques, such as the minimum variance phase reconstructor, have been developed to improve the imaging performance of adaptive-optics systems. However, these statistics-based models rely on knowledge of the current state of the key parameters. Neural networks provide nonlinear solutions to adaptive-optics problems while offering the possibility of adapting to changing seeing conditions. We address the use of neural networks for three tasks: (l) to reduce the WPS slope measurement error, (2) to estimate the Fried coherence length r(0), and (3) to estimate the variance of the WPS slope measurement error. All of these tasks are accomplished by using only the noisy WPS measurements as input. Where appropriate, we compare our method with classical statistics-based methods to determine if neural networks offer true benefits in performance. Although a statistics-based method is found to perform better than a neural network in reducing WPS slope measurement error, neural networks perform better in estimating the variance of the WPS slope measurement error, and both methods perform well in estimating r(0).     

Optimization of laser welding process parameters for super austenitic stainless steel using artificial neural networks and genetic algorithm

The laser welding input parameters play a very significant role in determining the quality of a weld joint. The quality of the joint can be defined in terms of properties such as weld bead geometry, mechanical properties and distortion. In particular mechanical properties should be controlled to obtain good welded joints. In this study, the weld bead geometry such as depth of penetration (DP), bead width (BW) and tensile strength (TS) of the laser welded butt joints made of AISI 904L super austenitic stainless steel are investigated. Full factorial design is used to carry out the experimental design. Artificial neural networks (ANNs) program was developed in MatLab software to establish the relationship between the laser welding input parameters like beam power, travel speed and focal position and the three responses DP, BW and TS in three different shielding gases (argon, helium and nitrogen). The established models are used for optimizing the process parameters using genetic algorithm (GA). Optimum solutions for the three different gases and their respective responses are obtained. Confirmation experiment has also been conducted to validate the optimized parameters obtained from GA.     

Cyber-based design for additive manufacturing using artificial neural networks for Industry 4.0

Additive Manufacturing (AM) requires integrated networking, embedded controls and cloud computing technologies to increase their efficiency and resource utilisation. However, currently there is no readily applicable system that can be used for cloud-based AM. The objective of this research is to develop a framework for designing a cyber additive manufacturing system that integrates an expert system with Internet of Things (IoT). An Artificial Neural Network (ANN) based expert system was implemented to classify input part designs based on CAD data and user inputs. Three ANN algorithms were trained on a knowledge base to identify optimal AM processes for different part designs. A two-stage model was used to enhance the prediction accuracy above 90% by increasing the number of input factors and datasets. A cyber interface was developed to query AM machine availability and resource capability using a Node-RED IoT device simulator. The dynamic AM machine identification system developed using an application programme interface (API) that integrates inputs from the smart algorithm and IoT interface for real-time predictions. This research establishes a foundation for the development of a cyber additive design for manufacturing system which can dynamically allocate digital designs to different AM techniques over the cyber network.     

Design of artificial neural networks using a genetic algorithm to predict collection efficiency in venturi scrubbers

In this study, a new approach for the auto-design of neural networks, based on a genetic algorithm (GA), has been used to predict collection efficiency in venturi scrubbers. The experimental input data, including particle diameter, throat gas velocity, liquid to gas flow rate ratio, throat hydraulic diameter, pressure drop across the venturi scrubber and collection efficiency as an output, have been used to create a GA-artificial neural network (ANN) model. The testing results from the model are in good agreement with the experimental data. Comparison of the results of the GA optimized ANN model with the results from the trial-and-error calibrated ANN model indicates that the GA-ANN model is more efficient. Finally, the effects of operating parameters such as liquid to gas flow rate ratio, throat gas velocity, and particle diameter on collection efficiency were determined.     

基于CNN的点云图像融合目标检测

针对自动驾驶场景中目标检测存在尺度变化、光照变化和缺少距离信息等问题,提出一种极具鲁棒性的多模态数据融合目标检测方法,其主要思想是利用激光雷达提供的深度信息作为附加的特征来训练卷积神经网络(CNN)。首先利用滑动窗对输入数据进行切分匹配网络输入,然后采用两个CNN特征提取器提取RGB图像和点云深度图的特征,将其级联得到融合后的特征图,送入目标检测网络进行候选框的位置回归与分类,最后进行非极大值抑制(NMS)处理输出检测结果,包含目标的位置、类别、置信度和距离信息。在KITTI数据集上的实验结果表明,本文方法通过多模态数据的优势互补提高了在不同光照场景下的检测鲁棒性,附加滑动窗处理改善了小目标的检测效果。对比其他多种检测方法,本文方法具有检测精度与检测速度上的综合优势。     

Combining neural networks and genetic algorithms for optimising the parameter design of the inter-metal dielectric process

An inter-metal dielectric (IMD) is deposited between metal layers to provide isolation capability to a device and separate the different metal layers that are unnecessary in the conduction of electricity. Owing to the complicated input/response relationships of the IMD process, the void problem results in electric leakage and causes wafer scraping. In the current study, we combined neural networks, genetic algorithms (GAs) and the desirability function in order to optimise the parameter settings of the IMD process. Initially, a backpropagation (BP) neural network was developed to map the complex non-linear relationship between the process parameters and the corresponding responses. Moreover, the desirability function and GAs were employed to obtain the optimum operating parameters in respect to each response. The implementation of the proposed approach was carried out in a semiconductor manufacturing company in Taiwan, and the results illustrate the practicability of the proposed approach.     

Wavefront reduction using graphs,neural networks and genetic algorithm

This paper presents an algorithm for nodal numbering in order to obtain a small wavefront. Element clique graphs are employed as the mathematical models of finite element meshes. A priority function containing five vectors is used, which can be viewed as a generalization of Sloan's function. These vectors represent different connectivity properties of the graph models. Unlike Sloan's algorithm, which uses two fixed coefficients, here, five coefficients are employed, based on an evaluation by artificial neural networks. The networks weights are obtained using a simple genetic algorithm. Examples are included to illustrate the performance of the present hybrid method. Copyright © 2004 John Wiley & Sons, Ltd.     

Large-scale prediction of cationic metabolite identity and migration time in capillary electrophoresis mass spectrometry using artificial neural networks

We developed a computational technique to assist in the large-scale identification of charged metabolites. The electrophoretic mobility of metabolites in capillary electrophoresis-mass spectrometry (CE-MS) was predicted from their structure, using an ensemble of artificial neural networks (ANNs). Comparison between relative migration times of 241 various cations measured by CE-MS and predicted by a trained ANN ensemble produced a correlation coefficient of 0.931. When we used our technique to characterize all metabolites listed in the KEGG ligand database, the correct compounds among the top three candidates were predicted in 78.0% of cases. We suggest that this approach can be used for the prediction of the migration time of any cation and that it represents a powerful method for the identification of uncharacterized CE-MS peaks in metabolome analysis.     

Water distribution under trickle irrigation predicted using artificial neural networks

An artificial neural network (ANN) technology is presented as an alternative to physical-based modeling of subsurface water distribution from trickle emitters. Three options are explored to prepare input–output functional relations from a database created using a numerical model (HYDRUS-2D). From the database the feasibility and advantages of the three alternative options are evaluated: water-content at defined coordinates, moment analysis describing the shape of the plume, and coordinates of individual water-content contours. The best option is determined in a way by the application objectives, but results suggest that prediction using moment analyses is probably the most versatile and robust and gives an adequate picture of the subsurface distribution. Of the other two options, the direct determination of the individual water contours was subjectively judged to be more successful than predicting the water content at given coordinates, at least in terms of describing the subsurface distribution. The results can be used to estimate subsurface water distribution for essentially any soil properties, initial conditions or flow rates for trickle sources.     

Approximate physical modelling in dynamic PSA using artificial neural networks

The continuous increase in the computational power of modern computers allows us to consider the feasibility of extending the present PSA studies, based on the usual probabilistic approach, to those aspects connected with the plant's dynamics. Indeed, in many cases the evolution of the process variables strongly affects the safety characteristics of a plant and, therefore, it cannot be neglected. Such a dynamic analysis requires solving the mathematical models describing the plant's behaviour. The corresponding equations need to be integrated with time steps which are related to the time evolution of the physical processes and therefore much smaller than the characteristic times typical of the PSA analyses. This procedure leads, in general, to very large computer times, so that it is presently still prohibitive for real plants. Therefore many current investigations are concerned with the development of new methodologies currently tested on simple study cases.In the present paper we consider the application of a multilayered, supervised artificial neural network trained by the error back-propagation algorithm for the solution of the mathematical models related to a simple study case of a dynamic PSA. In the examined case, the results indicate a reduction in the computer time, while the inherent approximations do not exceed those resulting from the uncertainties in the input data. These advantages are expected to increase when the future parallel computers become available.