Parallel Optimization of Program Instructions Using Genetic Algorithms

This paper describes an efficient solution to parallelize software program instructions, regardless of the programming language in which they are written. We solve the problem of the optimal distribution of a set of instructions on available processors. We propose a genetic algorithm to parallelize computations, using evolution to search the solution space. The stages of our proposed genetic algorithm are: The choice of the initial population and its representation in chromosomes, the crossover, and the mutation operations customized to the problem being dealt with. In this paper, genetic algorithms are applied to the entire search space of the parallelization of the program instructions problem. This problem is NP-complete, so there are no polynomial algorithms that can scan the solution space and solve the problem. The genetic algorithm-based method is general and it is simple and efficient to implement because it can be scaled to a larger or smaller number of instructions that must be parallelized. The parallelization technique proposed in this paper was developed in the C# programming language, and our results confirm the effectiveness of our parallelization method. Experimental results obtained and presented for different working scenarios confirm the theoretical results, and they provide insight on how to improve the exploration of a search space that is too large to be searched exhaustively.     

微弧等离子喷涂AT13纳米涂层的工艺优化

根据微弧等离子喷涂在不同工艺参数下制备的AT13纳米涂层结合强度和显微硬度的测试结果, 运用遗传神经网络算法对喷涂工艺参数与涂层结合强度和显微硬度之间的非线性关系进行建模与仿真, 并应用遗传算法和多目标优化理论对AT13纳米涂层的性能进行了优化. 研究结果表明, 仿真值与实验值相一致, 相对误差<0.5%; 喷涂电流、Ar流量和Ar压力分别为150A、64m³/h和0.38MPa时, AT13纳米涂层综合性能最佳.     

Compounding Meta-Atoms into Metamolecules with Hybrid Artificial Intelligence Techniques

Molecules composed of atoms exhibit properties not inherent to their constituent atoms. Similarly, metamolecules consisting of multiple meta-atoms possess emerging features that the meta-atoms themselves do not possess. Metasurfaces composed of metamolecules with spatially variant building blocks, such as gradient metasurfaces, are drawing substantial attention due to their unconventional controllability of the amplitude, phase, and frequency of light. However, the intricate mechanisms and the large degrees of freedom of the multielement systems impede an effective strategy for the design and optimization of metamolecules. Here, a hybrid artificial-intelligence-based framework consolidating compositional pattern-producing networks and cooperative coevolution to resolve the inverse design of metamolecules in metasurfaces is proposed. The framework breaks the design of the metamolecules into separate designs of meta-atoms, and independently solves the smaller design tasks of the meta-atoms through deep learning and evolutionary algorithms. The proposed framework is leveraged to design metallic metamolecules for arbitrary manipulation of the polarization and wavefront of light. Moreover, the efficacy and reliability of the design strategy are confirmed through experimental validations. This framework reveals a promising candidate approach to expedite the design of large-scale metasurfaces in a labor-saving, systematic manner.     

现代优化计算方法在材料最优化设计中的应用

简要介绍三种现代优化算法 ,综合利用主成分降维技术、人工神经网络技术和遗传算法技术 ,在V PTC材料介电性能和五个影响因素之间建立神经网络模型 ,然后应用遗传算法搜索最高电阻值和相应的配方。结果表明 :现代优化算法在分析合理选择的样本数据 ,总结其中的数值规律 ,进而对材料性能进行优化设计方面 ,具有重要的应用价值。     

Hybrid neural network model for the design of beam subjected to bending and shear

There is no direct method for design of beams. In general the dimensions of the beam and reinforcement are initially assumed and then the interaction formula is used to verify the suitability of chosen dimensions. This approach necessitates few trials for coming up with an economical and safe design. This paper demonstrates the applicability of Artificial Neural Networks (ANN) and Genetic Algorithms (GA) for the design of beams subjected to moment and shear. A hybrid neural network model which combines the features of feed forward neural networks and genetic algorithms has been developed for the design of beam subjected to moment and shear. The network has been trained with design data obtained from design experts in the field. The hybrid neural network model learned the design of beam in just 1000 training cycles. After successful learning, the model predicted the depth of the beam, area of steel, spacing of stirrups required for new problems with accuracy satisfying all design constraints. The various stages involved in the development of a genetic algorithm based neural network model are addressed at length in this paper.     

大规模物流网络的组合遗传算法研究

在对遗传算法、生成树遗传算法和混合进化方法进行比较的基础上,提出组合遗传算法来解决大规模基本物流网络设计问题.该问题抽象为"选址-分派问题",并进一步分解,且将"分派"镶嵌在"选址"中."选址"和"分派"染色体分别使用二进制编码和矩阵编码、适值采用物流费用.运算结果显示该方法比混合进化方法结果更精确,且在大规模问题求解方面速度优于通常的遗传算法,因此是一种设计大规模物流网络的较好方法.     

Condition‐based Maintenance Optimization Using Neural Network‐based Health Condition Prediction

Artificial neural network (ANN)‐based methods have been extensively investigated for equipment health condition prediction. However, effective condition‐based maintenance (CBM) optimization methods utilizing ANN prediction information are currently not available due to two key challenges: (i) ANN prediction models typically only give a single remaining life prediction value, and it is hard to quantify the uncertainty associated with the predicted value; (ii) simulation methods are generally used for evaluating the cost of the CBM policies, while more accurate and efficient numerical methods are not available, which is critical for performing CBM optimization. In this paper, we propose a CBM optimization approach based on ANN remaining life prediction information, in which the above‐mentioned key challenges are addressed. The CBM policy is defined by a failure probability threshold value. The remaining life prediction uncertainty is estimated based on ANN lifetime prediction errors on the test set during the ANN training and testing processes. A numerical method is developed to evaluate the cost of the proposed CBM policy more accurately and efficiently. Optimization can be performed to find the optimal failure probability threshold value corresponding to the lowest maintenance cost. The effectiveness of the proposed CBM approach is demonstrated using two simulated degradation data sets and a real‐world condition monitoring data set collected from pump bearings. The proposed approach is also compared with benchmark maintenance policies and is found to outperform the benchmark policies. The proposed CBM approach can also be adapted to utilize information obtained using other prognostics methods. Copyright © 2012 John Wiley & Sons, Ltd.     

人工神经网络在弹体侵彻混凝土深度中的应用

针对弹体对混凝土材料侵彻深度问题,通过量纲分析和神经网络理论,建立了弹体侵彻深度h网络输出量与弹体长度lp、弹的长径比 lp/d、弹体形状系数ψ、弹体与混凝土的比强度σyt/σyp、弹体与混凝土的密度比ρp/ρt等13个网络输入量之间的非线性映射关系。并采用 RBF网络模型,通过Forrestal等文献的试验样本对网络模型训练,获得了弹体对混凝土材料侵彻深度的网络模型,输出结果满意。     

矿物质粉体对砂浆及混凝土Cl-渗透性的影响

研究了不同水胶比、不同矿物质粉体掺量的砂浆和混凝土，经标准养护至56天、90天时的导电量。在相同水胶比和相同矿物质粉体掺量下，混凝土的导电量远低于砂浆的导电量。含矿物质粉体的砂浆及混凝土的导电量均低于基准砂浆及混凝土的导电量。导电量随水胶比的降低而降低，也随龄期的增长而降低。     

基于逆问题求解的汽车操纵性能优化

给出一种基于逆问题求解的人-车-路闭环系统操纵性能优化的方法.利用径向基函数神经网络建立了汽车侧向位移与方向盘转角及其它响应之间的映射关系,由跟踪路径反求出方向盘转角及汽车的其它响应,进而计算闭环系统的操纵性能评价指标并进行优化.该方法是在不同汽车方案具有相同实际行驶路径的基础上对操纵性能进行分析并优化,从而得到的最优汽车方案在跟踪某一典型路径时具有最好的操纵性能.     

基于计算智能的包装件非线性特性识别研究进展

将人工神经网络方法,模糊自适应控制技术与进化计算应用于包装件缓冲垫层非线性特性的识别问题,可以给出一种为评价、论证与设计缓冲包装提供理论依据的新途径,概述了有关该问题的若干进展情况,讨论了存在的问题并指出了有关的研究发展方向。     

一种混合优化策略在工步排序中的应用

结合创成式ＣＡＰＰ系统中工步优化问题，介绍了由遗传算法（ＧｅｎｅｔｉｃＡｌｇｏｒｉｔｈｍｓ，ＧＡ）和模拟退火算法（ＳｉｍｕｌａｔｅｄＡｎｎｅａｌｉｎｇＡｌｇｏｒｉｔｈｍｓ，ＳＡ）构成的混合寻优策略。最后以一箱体的工艺规划过程为例，将基于混合寻优策略的工步排序融入以加工中心为主要加工设备的ＣＡＰＰ工艺决策过程。     

An Intelligence Computational Approach for the Fractional 4D Chaotic Financial Model

The main purpose of the study is to present a numerical approach to investigate the numerical performances of the fractional 4-D chaotic financial system using a stochastic procedure. The stochastic procedures mainly depend on the combination of the artificial neural network (ANNs) along with the Levenberg-Marquardt Backpropagation (LMB) i.e., ANNs-LMB technique. The fractional-order term is defined in the Caputo sense and three cases are solved using the proposed technique for different values of the fractional order α. The values of the fractional order derivatives to solve the fractional 4-D chaotic financial system are used between 0 and 1. The data proportion is applied as 73%, 15%, and 12% for training, testing, and certification to solve the chaotic fractional system. The acquired results are verified through the comparison of the reference solution, which indicates the proposed technique is efficient and robust. The 4-D chaotic model is numerically solved by using the ANNs-LMB technique to reduce the mean square error (MSE). To authenticate the exactness, and consistency of the technique, the obtained performances are plotted in the figures of correlation measures, error histograms, and regressions. From these figures, it can be witnessed that the provided technique is effective for solving such models to give some new insight into the physical behavior of the model.     

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。     

浅海环境参数宽带匹配场反演研究

宽带匹配场反演是获取海洋环境参数信息的有效方法 ,选择合适的反演目标函数对环境参数的估计结果十分重要。文章研究了浅海环境参数宽带匹配场反演的各种目标函数建立方法 ,同时结合亚洲海实验 (ASIAEX2 0 0 1)数据采用遗传全局寻优算法对该海域环境参数进行了反演 ,并对反演结果尤其是沉积层参数的合理性进行了分析。     

基于遗传算法的工程项目投资综合优化探讨

利用多目标优化理论、多属性效用函数理论建立了工程建设项目工期、成本和质量的综合均衡优化模型，在网络计划技术的基础上，用遗传算法对模型进行求解，可以得到最满意的决策方案和多个近似满意的备选方案，作为工程建设项目投资综合控制的目标。并应用实例对模型的可行性和实用性进行了验证。     

Energy Prediction in IoT Systems Using Machine Learning Models

The Internet of Things (IoT) technology has been developed for directing and maintaining the atmosphere in smart buildings in real time. In order to optimise the power generation sector and schedule routine maintenance, it is crucial to predict future energy demand. Electricity demand forecasting is difficult because of the complexity of the available demand patterns. Establishing a perfect prediction of energy consumption at the building’s level is vital and significant to efficiently managing the consumed energy by utilising a strong predictive model. Low forecast accuracy is just one of the reasons why energy consumption and prediction models have failed to advance. Therefore, the purpose of this study is to create an IoT-based energy prediction (IoT-EP) model that can reliably estimate the energy consumption of smart buildings. A real-world test case on power predictions is conducted on a local electricity grid to test the practicality of the approach. The proposed (IoT-EP) model selects the significant features as input neurons, the predictable data is selected as output nodes, and a multi-layer perceptron is constructed along with the features of the Convolution Neural Network (CNN) algorithm. The analysis of the proposed IoT-EP model has higher accuracy of 90%, correlation of 89%, and variance of 16% in less training time of 29.2 s, and with a higher prediction speed of 396 (observation/sec). When compared to existing models, the results showed that the proposed (IoT-EP) model outperforms with a satisfactory level of accuracy in predicting energy consumption in smart buildings.     

Impact Location and Quantification on a Composite Panel using Neural Networks and a Genetic Algorithm

Abstract: The problem of impact detection in composite panels using artificial neural networks is addressed in this paper. The data were taken from an experiment in which time dependent strain data were recorded on a network of surface-mounted piezoceramic sensors when the plate was impacted. Neural networks were trained to locate and quantify the impact event when presented with features extracted from the measured data. An important problem for detection systems like this is that of optimal sensor placement; this is solved here by means of a Genetic Algorithm. The study shows that a relatively small number of sensors can be used to detect reliably impacts on a composite plate.     

基于RBF神经网络的隧洞围岩变形预测方法

传统回归方法对于围岩变形时程曲线存在反弯点，这种情况的模拟具有难度。提出的基于RBF神经网络的隧道围岩变形预测方法，不仅能很好地描述复杂的围岩变形时程曲线     

Scalability of a Hybrid Extended Compact Genetic Algorithm for Ground State Optimization of Clusters

We analyze the utility and scalability of extended compact genetic algorithm (eCGA)—a genetic algorithm (GA) that automatically and adaptively mines the regularities of the fitness landscape using machine learning methods and information theoretic measures—for ground state optimization of clusters. In order to reduce the computational time requirements while retaining the high reliability of predicting near-optimal structures, we employ two efficiency-enhancement techniques: (1) hybridizing eCGA with a local search method, and (2) seeding the initial population with lowest energy structures of a smaller cluster. The proposed method is exemplified by optimizing silicon clusters with 4-20 atoms. The results indicate that the population size required to obtain near-optimal solutions with 98% probability scales sub linearly (as Θ(n^0.83)) with the cluster size. The total number of function evaluations (cluster energy calculations) scales sub-cubically (as Θ(n^2.45)), which is a significant improvement over exponential scaling of poorly designed evolutionary algorithms.