Expert system based controller for the high-accuracy automatic assembly of vehicle headlamps

A vehicle headlamp is a high-sophisticated device basically composed of three elements: housing, lighting system and lens. Nowadays, they are assembled with no dimensional control, obtaining a final product with high physical deviations. Since the car manufacturer is requiring more and more precision, the number of faulty headlamps is increasing, and thus the cost to the headlamp producer. To solve this problem, a new automatic assembly machine was built. It integrated different technologies such as sensor systems, linear actuators and computational capacities.This work presents the conception, design and development of a fuzzy control algorithm for an automatic machine capable of correcting the dimensional tolerances of vehicle headlamps during its assembly process. Particularly, this paper describes the problem of modeling and control an industrial system, with a high complex dynamics and unknown model, by a rule-based model constructed from measured data.To validate the proposed control system, different experimental tests were carried out in both the lab and the assembly line in the factory. The obtained results show how the prototype, together with the proposed fuzzy software, solve the problem of assembling vehicle headlamps obtaining a final product with minimum dimensional errors.     

Optimum spacing design of grillage systems using a genetic algorithm

In this study a genetic algorithm based method is developed for the optimum design of grillage systems. The algorithm not only selects the optimum sections for the grillage elements from a set of standard universal beam sections, but also finds the optimum spacing required for the grillage system. Deflection limitations and allowable stress constraints are considered in the formulation of the design problem. Due to the fact that grillage elements are thin walled sections, warping plays an important role in their design, particularly, when they are subjected to significant torsional loading. The algorithm developed has the flexibility of including or excluding the effect of warping in the design process. Grillage structures are designed for uniformly distributed loading. The optimum spacings are determined both considering and not taking into account the effect of warping in the design. The comparison of the results shows that inclusion of warping in the design process has a significant effect on the optimum spacing.     

子空间结构保持的多层极限学习机自编码器

处理高维复杂数据的聚类问题,通常需先降维后聚类,但常用的降维方法未考虑数据的同类聚集性和样本间相关关系,难以保证降维方法与聚类算法相匹配,从而导致聚类信息损失.非线性无监督降维方法极限学习机自编码器(Ex-treme learning machine,ELM-AE)因其学习速度快、泛化性能好,近年来被广泛应用于降维及去...     

一种基于几何概率的聚类有效性函数

聚类有效性是聚类分析中尚未解决的基本问题,最佳聚类数的确定是聚类有效性问题中的主要研究内容。以几何概率为理论依据,针对2维数据集提出了一种新的聚类有效性函数,用于确定最佳聚类数。该函数利用2维数据集与2维离散点集之间存在的对应关系,以2维离散点集在特征空间中的分布特征为依据,测度对应数据集的聚类结构,思路直观、容易理解。测度过程中,将点集中的点两两相连生成一个线段集合保存点集的结构信息,通过比较线段集合中线段方向取值与完全随机条件下线段方向取值的相对大小,构造聚类有效性函数。实验结果表明,针对给定的样本数据集,生成该函数的曲线,再根据曲线的形态能够有效地确定2维数据集的最佳聚类数,指导聚类算法设计。     

Structural optimization of the cross-beam of a gantry machine tool based on grey relational analysis

Crossbeam structural design of gantry machine tool is a multi-level, multi-index and multi-scheme decision-making problem. In order to solve the above problem, the optimum seeking model of crossbeam structure was built through using the grey relational analysis and Analytic Hierarchy Process. The finite element analysis of the static and dynamic performance parameters for four kinds of crossbeam structural schemes designed had been done, and the optimal design scheme was selected by using the optimum seeking model. After conducting sensitivity analysis for the optimal crossbeam selected, the reasonable design variables were obtained, and the dynamic optimization design model of crossbeam was established. Six groups of non-inferior solutions were obtained after solving the optimization design model. The optimal solution was selected from the non-inferior solution set through using the crossbeam structural optimization method based on grey relational analysis again, which makes the crossbeam’s dynamic performance improving greatly. The dynamic experiments on the crossbeams before and after optimization design were conducted, then the experimental results show that the first four order natural frequencies of the crossbeam increase 17.56 %, 19.36 %, 17.04 % and 19.58 % respectively, which proves that the structural optimization design method based on grey relational analysis proposed in this paper is reasonable and practicable.     

机器学习方法在蛋白质折叠结构预测中的应用研究

蛋白质的生物功能是由它们的空间折叠结构决定的,理解蛋白质的折叠过程是生物信息学领域中极具挑战性的问题之一。近年来,各种优化方法用于蛋白质空间折叠结构预测。这些方法仍存在着不足,算法在变量数目增大时,难以收敛到全局最优解,并容易产生早熟收敛,从而影响求解精度和效率。针对蛋白质结构预测模型多变量多极值的特点,文章采用机器学习方法对蛋白质进行折叠结构预测。     

Topology optimization of frame structures with flexible joints

A method for structural topology optimization of frame structures with flexible joints is presented. A typical frame structure is a set of beams and joints assembled to carry an applied load. The problem considered in this paper is to find the stiffest frame for a given mass. By introducing design variables for beams and joints, a mass distribution for optimal structural stiffness can be found. Each beam can have several design variables connected to its cross section. One of these is an area-type design variable which is used to represent the global size of the beam. The other design variables are of length ratio type, controlling the cross section of the beam. Joints are flexible elements connecting the beams in the structure. Each joint has stiffness properties and a mass. A framework for modelling these stiffnesses is presented and design variables for joints are introduced. We prove a theorem which can be interpreted as the fact that the removal of structural elements, e.g. joints or beams, can be modelled by a small strictly positive material amount assigned to the element. This is needed for the computations of sensitivities used in the applied gradient based iterative method. Both two and three dimensional problems, as well as multiple load cases and multiple mass constraints, are treated.     

基于拓扑优化的筏形基础传力途径仿真

文章主要介绍了通过拓扑优化对筏形基础的传力途径进行仿真分析的基本原理和所涉及的关键问题.结合大型有限元软件ANSYS,建立了筏板-地基土共同作用体系的全三维有限元模型.并在此基础上,采用变密度法建立拓扑优化模型,选取序列线性规划法对某高层建筑的筏形基础结构进行了仿真分析.仿真结果较为清楚地反映了筏板内部应力分布的宏观规律.表明,应用有限元法进行筏形基础结构拓扑优化设计是一种有效的优化方法,可为高层建筑筏形基础设计提供重要的概念化设计参考.     

Optimization of airplane wing structures under gust loads

A methodology is presented for the optimum design of aircraft wing structures subjected to gust loads. The equations of motion, in the form of coupled integro-differential equations, are solved numerically and the stresses in the aircraft wing structure are found for a discrete gust encounter. The gust is assumed to be one minus cosine type and uniform along the span of the wing. In order to find the behavior of the wing structure under gust loads and also to obtain a physical insight into the nature of the optimum solution, the design of the typical section (symmetric double wedge airfoil) is studied by using a graphical procedure. Then a more realistic wing optimization problem is formulated as a constrained nonlinear programming problem based on finite element modeling and the optimum solution is found by using the interior penalty function method. A sensitivity analysis is conducted to find the effects of changes in design variables about the optimum point on the response quantities of the wing structure.     

Optimal design of truss structures using parallel computing

Parallel design optimization of large structural systems calls for a multilevel approach to the optimization problem. The general optimization problem is decomposed into a number of non-interacting suboptimization problems on the first level. They are controlled from the second level through coordination variables. Thus, the solutions of the independent first-level subsystems are directed towards the overall system optimum. In the present paper, optimal design of truss structures using parallel computing technique is described. In this method, optimization of a large truss structure has been carried out by decomposing the structure into sub-domains and suboptimization tasks. Each sub-domain has independent design variables and a small number of behaviour constraints. The two-level sub-domain optimum design approach is summarized by several numerical examples with speedups and efficiencies of algorithms on message passing systems. It has been noticed that the efficiency of the algorithm for design optimization increases with the size of the structure.     

基于支持向量机的数据库学习算法

文中介绍了一个利用数据库中的大量数据进行决策的方法。对于仅涉及数据库中部分数据的问题,对数据库中与当前问题相关的数据采用具有强泛化能力的支持向量机方法学习分类规则和回归函数,完成对当前问题的分类和估值。支持向量机算法用非线性映射把数据映射到一个高维特征空间,在高维特征空间进行线性分类和线性回归,将原问题转化为一个凸次优化问题。上述算法实现了一个隧道工程支护设计系统,并取得了较好的效果。     

Research on quality performance conceptual design based on SPEA2+

In order to solve the multi-objective performance optimal problems, SPEA2+ is used to realize the performance design of injection molding machine. The optimization objectives are constructed to maximize mould control power, maximize injection quantity and minimize injection power. The mathematical model is found to optimize the problem. A solution is extracted to eliminate the imprecise nature of preference through the Pareto optimal set based on fuzzy set theory. Compared with NSGA-II and SPEA2, SPEA2+ could acquire the Pareto front with better distribution and smaller distance with the optimum solutions. Finally, the case illustration of HTG1000X3Y injection molding machine is taken as an example to demonstrate that such method is effective and practical. Effective references could be provided to decision makers for objectives tradeoff at the performance conceptual design stage of injection molding machine.     

Multilevel redundancy allocation using two dimensional arrays encoding and hybrid genetic algorithm

With the popularity of multilevel design in large scale systems, reliability redundancy allocation on multilevel systems is becoming attractive to researchers. Multilevel redundancy allocation problem (MLRAP) is not only NP-hard, but also qualifies as hierarchy optimization problem. Exact method could not tackle MLRAP very well, so heuristic and meta-heuristic methods are often used to solve it. To improve the effectiveness of current algorithms on MLRAP, this paper proposes a hybrid genetic algorithm (HGA) based on the two dimensional redundancy encoding mechanism. Instead of hierarchical genotype representation, a two dimensional array is used to represent the solutions to MLRAP. Each row of the array contains the redundancy information of a certain unit in the system and each element in one row stands for the redundancy value of one element of that unit. The number of rows of this array is fixed and equals to the number of distinct units in the system. Each row of the array is an unfixed-length vector whose length depends on the redundancy of all elements of its parent unit. On top of this two dimensional arrays, a local search operator employing simulated annealing strategy is used to generate new population for the next generation instead of the traditional genetic operators. Experimental results have shown that our two dimensional arrays based HGA outperforms the state-of-the-art approaches using two kinds of multilevel system structure.     

Hysteretic damping of structures vibrating at resonance: An iterative complex eigensolution method based on damping-stress relation

In this paper the damping is examined as an engineering property used in analysis and design of structures and machines. The design engineer needs to know not only the stresses of his structure or machine, under steady state conditions but also the stresses under resonance conditions. Then the material damping, as a function of the stress of the structure, has an important role to play and ignoring the damping the calculated stresses are far from reality. The nonlinearity here is due to the dependence of the hysteretic damping on the stress of the structure. Specifically here two problems are investigated in the following way:Firstly the direct problem is solved. The direct problem is to find the maximum bending stress at the resonance when the relation of the dissipating energy (or of the hysteretic damping) vs. the bending stress is known in advance. To perform this calculation, a useful tool for the design engineer, the structure is modelled using the continuum mechanics analytical approach or the finite elements (FE) method. Then the eigenvalues are calculated and using an iterative procedure the real stress. The procedure presented here is called iterative complex eigensolution method (ICEM). Secondly the inverse problem is solved. The inverse problem is to find the relation between the hysteretic damping and the bending stress. For this purpose the logarithmic decrement is experimentally measured, the eigenvalues and the maximum bending stress of the structure, excited at the eigenvalue, when the damping is the same as the measured one, are computed using the finite elements method. Once the bending stresses are found in each discrete element of the structure, then the mathematical expression of the relation of the dissipating energy and the stresses can be specified by minimizing a suitably formed objective function.     

基于作业模式的提梁机有限元仿真分析

既有提梁机设计和结构分析多参考相关规范进行,未考虑走行轨道不平顺、大车 走行不同步等作业工况对整机结构的影响,不利于提梁机作业安全。针对这一问题,基于提梁 机作业模式分析了影响提梁机结构安全的5 种因素,并假定每种影响因素具有3 种取值方式, 采用正交试验设计方法,通过L9(34)正交表合理确定考虑影响因素的工况组合。结合试验结果 对MG450 型提梁机进行了有限元仿真分析,与传统方法进行了计算结果对比分析。研究表明： 作业模式对结构分析结果影响明显,基于作业模式的提梁机有限元分析结果更能够真实地反映 结构受力情况。     

Optimization of airplane wing structures under landing loads

A methodology is presented for the optimum design of aircraft wing structures subjected to landing loads. The stresses developed in the wing during landing are computed by considering the interaction between the landing gear and the flexible airplane structure. The landing gear is assumed to have nonlinear characteristics typical of conventional gears, namely, velocity squared damping, polytropic air-compression springing and exponential tire force-deflection characteristics. The coupled nonlinear differential equations of motion that arise in the landing analysis are solved by using a step-by-step numerical integration technique. In order to find the behavior of the wing structure under landing loads and also to obtain a physical insight into the nature of the optimum solution, the design of the typical section (symmetric double-wedge airfoil) is studied by using a graphical procedure. Then a more realistic wing optimization problem is formulated as a constrained nonlinear programming problem based on finite element modeling. The optimum solutions are found by using the interior penalty function method. A sensitivity analysis is conducted to find the effect of changes in design variables about the optimum point on the various response parameters on the wing structure.     

Algorithms for point-wise state variable constraints in structural optimization

This paper presents algorithms to treat the point-wise state variable constraints in finite dimensional and distributed parameter structural optimization problems. The idea is to impose such constraints at all local maximum points, or over a small region around the maximum points. Therefore, methods for design sensitivity analysis to handle the constraints at some particular point for the distributed parameter problem are presented. The direct differentiation and adjoint variable methods are employed to derive the design sensitivity expressions. For the finite dimensional problem the new idea is easily carried out in optimization process. Simply supported and clamped beams are optimized using the new approach. These are modeled with nonuniform beam elements. Comparisons of finite dimensional and distributed parameter problems are also made.     

Automatic design synthesis with artificial intelligence techniques

Design synthesis represents a highly complex task in the field of industrial design. The main difficulty in automating it is the definition of the design and performance spaces, in a way that a computer can generate optimum solutions. Following a different line from the machine learning, and knowledge-based methods that have been proposed, our approach considers design synthesis as an optimization problem. From this outlook, neural networks and genetic algorithms can be used to implement the fitness function and the search method needed to achieve optimum design. The proposed method has been tested in designing a telephone handset. Although the objective of this application is based on esthetic and ergonomic cues (subjective information), the algorithm successfully converges to good solutions.     

A short and versatile finite element multiscale code for homogenization problems

We describe a multiscale finite element (FE) solver for elliptic or parabolic problems with highly oscillating coefficients. Based on recent developments of the so-called heterogeneous multiscale method (HMM), the algorithm relies on coupled macro- and microsolvers. The framework of the HMM allows to design a code whose structure follows the classical finite elements implementation at the macro level. To account for the fine scales of the problem, elementwise numerical integration is replaced by micro FE methods on sampling domains. We discuss a short and flexible FE implementation of the multiscale algorithm, which can accommodate simplicial or quadrilateral FE and various coupling conditions for the constrained micro simulations. Extensive numerical examples including three dimensional and time dependent problems are presented illustrating the efficiency and the versatility of the computational strategy.     

Optimum design of space structures with active and passive damping

This paper investigates the effects of passive damping on the optimum structural design with active controllers. An integrated design of structure and active control system is performed. The minimum weight design is obtained by imposing constraints on the closed-loop damping and the imaginary part of the closed-loop eigenvalues of the active control system. The mathematical optimization problem is solved by using the NEWSUMT-A program, which is based on quadratic extended interior penalty function method. The ACOSS-FOUR model is selected for the numerical studies. The active control effort, performance index, and optimum weights are presented as a function of the passive damping.