遗传算法改进的BP神经网络在GPS高程拟合中的应用

针对BP神经网络收敛速度慢和易于陷入局部极小值的问题,采用将遗传算法全局寻优和BP神经网络局部寻优相结合的方法,优化神经网络各层之间的连接权和阈值,提高了BP神经网络的计算精度、收敛速度和泛化能力。本文论述了遗传算法的基本思想、实现过程,并对高程拟合算例进行训练检验,实验结果表明,遗传算法改进的BP神经网络进行GPS高程拟合是可行的,能够有效地提高BP神经网络的拟合精度。     

Neural Network Approaches to Aid Simple Truss Design Problems

Abstract: Diverse problems in engineering may be solved accurately with computers. In structural engineering, many solution techniques exist. Over the past few years, neural networks have evolved as a new computing paradigm, and many engineering applications have been studied. This paper describes configuring and training of a neural network for a truss design application and explores the possible roles for neural networks in structural design problems. The specific problem considered is a simple truss design where, given a geometry and a loading, economical cross-sectional areas of all the members are to be selected. For this problem, a two-layer neural network is trained using the back-propagation algorithm with patterns representing optimal designs for diverse loading conditions. The performance of the trained neural network is evaluated with a sample problem.     

优化神经网络在多传感器火灾探测中的应用

针对BP神经网络在拟合过程中探测精度低、容易陷入局部最优的问题,提出一种基于遗传算法（GA）和模拟退火算法（SA）共同改进的BP神经网络模型,该网络模型可以有效提高火灾识别准确率,同时避免网络过拟合现象,使预测结果跳出局部最优从而达到全局最优。首先,通过GA改进隐藏层结构部分,然后通过SA改进连接权重部分,最后利用优化后的GA-SA-BP模型对火灾实验数据进行信息融合实现火灾探测。实验研究表明,对比单一BP神经网络,经GA和SA改进后的BP神经网络能够有效改善网络拟合能力,并提升火灾探测精度至98.91%。     

Multi-objective optimal design of braced frames using hybrid genetic and ant colony optimization

In this article, multi-objective optimization of braced frames is investigated using a novel hybrid algorithm. Initially, the applied evolutionary algorithms, ant colony optimization (ACO) and genetic algorithm (GA) are reviewed, followed by developing the hybrid method. A dynamic hybridization of GA and ACO is proposed as a novel hybrid method which does not appear in the literature for optimal design of steel braced frames. Not only the cross section of the beams, columns and braces are considered to be the design variables, but also the topologies of the braces are taken into account as additional design variables. The hybrid algorithm explores the whole design space for optimum solutions. Weight and maximum displacement of the structure are employed as the objective functions for multi-objective optimal design. Subsequently, using the weighted sum method (WSM), the two objective problem are converted to a single objective optimization problem and the proposed hybrid genetic ant colony algorithm (HGAC) is developed for optimal design. Assuming different combination for weight coefficients, a trade-off between the two objectives are obtained in the numerical example section. To make the final decision easier for designers, related constraint is applied to obtain practical topologies. The achieved results show the capability of HGAC to find optimal topologies and sections for the elements.     

基于MATLAB的桁架结构优化设计

介绍了基于BP神经网络的全局性结构近似分析方法 ,解决了结构优化设计问题中变量的非线性映射问题。在此基础上 ,利用改进的遗传算法 ,对桁架结构在满足应力约束条件下进行结构最轻优化设计。利用Matlab的神经网络工具箱 ,编程求解了三杆桁架优化问题。     

A novel approach for optimal chiller loading using particle swarm optimization

This study employs two new methods to solve optimal chiller loading (OCL) problem. These methods are continuous genetic algorithm (GA) and particle swarm optimization (PSO). Because of continuous nature of variables in OCL problem, continuous GA and PSO easily overcome deficiencies in other conventional optimization methods. Partial load ratio (PLR) of the chiller is chosen as the variable to be optimized and consumption power of the chiller is considered as fitness function. Both of these methods find the optimal solution while the equality constraint is exactly satisfied. Some of the major advantages of proposed approaches over other conventional methods can be mentioned as fast convergence, escaping from getting into local optima, simple implementation as well as independency of the solution from the problem. Abilities of proposed methods are examined with reference to an example system. To demonstrate these abilities, results are compared with binary genetic algorithm method. The proposed approaches can be perfectly applied to air-conditioning systems.     

A comparative study on optimization of constrained layer damping treatment for structural vibration control

This paper presents a comparison of optimization algorithms for constrained layer damping (CLD) patches’ layout with an objective to minimize the maximum vibration response of the odd modes, which constitutes the dominant acoustic radiation, of a simply supported beam excited by a harmonic transverse force. An analytical model developed for relating the displacement response of the beam with bonded CLD patches and their layout is adopted to formulate the optimization problem. Four different nonlinear optimization methods/algorithms, sub-problem approximation method, the first-order method, sequential quadratic programming (SQP) and genetic algorithm (GA), are then, respectively, used to optimize the CLD patches’ locations and lengths with the aim of minimum displacement amplitude at the middle beam. The efficiency of each considered optimization method is evaluated and also compared in terms of obtained optimal beam displacement and the added weight owing to damping treatment. The results show that GA is most efficient in obtaining the best optimum for this optimization problem in spite of highest computation efforts required to improve its stability.     

Algorithms for Optimal Design of Smart Structural Systems

Simultaneous optimal design of smart structural systems is considered in this article. A major augmentation to this multidisciplinary optimization problem is optimal structural layout design in terms of shape and topology. Four algorithms, namely, guided random search techniques, sequential mathematical programming, and their mixtures, are presented and used for the problem solving. A method to prevent singularities when updating structural layout is also suggested. Design variable linking schemes are used to reduce the number of design variables and give reasonable optimal results. Two examples are solved. Comparison of the algorithms on the basis of achieved improvements in system energy, control effort, robustness, and controllability is made.     

Optimal placement of metallic dampers for seismic upgrading of multistory buildings based on a cost‐effectiveness criterion using genetic algorithm

In this paper, an optimal placement methodology for metallic dampers is proposed to upgrade the seismic performance of multistory buildings. Most previous studies on optimal damper placement (ODP) problems have been focused on minimizing the seismic responses, whereas the present study aims to utilize the minimum total cost of dampers to achieve a prescribed level of seismic response. To this end, the optimization objective is constructed based on a cost‐effectiveness criterion, and the optimization constraint is defined based on a desired level of seismic response. An improved integer‐coded genetic algorithm is presented for solving the ODP problem. A 16‐story shear building is illustrated to verify the proposed optimal placement methodology. It is shown that the proposed methodology can be used to achieve the predetermined performance level while minimizing the retrofitting cost. Moreover, different algorithms, objective functions, and levels of accuracy on the optimization are also compared. Finally, a two‐step optimization approach is proposed for achieving better placement schemes with less computational efforts.     

Optimal tuning of fuzzy parameters for structural motion control using multiverse optimizer

In recent years, there is an increasing interest in optimization of structural control algorithms. Fuzzy logic controller is one of the most common and versatile control algorithms that is generally formulated based on the human knowledge and expert. Human knowledge and experience do not yield optimal control responses for a given structure, and tuning of the fuzzy parameters is necessary. This paper focuses on the optimization of a fuzzy controller applied to a seismically excited nonlinear building. In the majority of cases, this problem is formulated based on the linear behavior of the structure; however, in this paper, objective functions and the evaluation criteria are considered with respect to the nonlinear responses of the structures. Multiverse optimizer is a novel nature‐inspired optimization algorithm that is based on the three concepts of cosmology as white hole, black hole, and wormhole. This algorithm has fast convergence rate and can be utilized in continuous and discrete optimization problems. In this paper, the multiverse optimizer is considered as the optimization algorithm for optimization of the fuzzy controller. The performance of the selected algorithm is compared with eight different optimization algorithms. The results prove that the selected algorithm is able to provide very competitive results.     

基于改进GA-SVR算法的隧道工程三维弹塑性模型参数的智能辨识

将支持向量回归（SVR）算法引入岩土工程数值计算模型参数的辨识中可以充分发挥SVR算法的小样本、泛化性好和全局最优化的优点。但现阶段标准的SVR算法只能解决一维输出变量的回归问题，这就使其在反分析领域的应用受到限制。引入一种改进的SVR算法，这种算法通过将多维输出变量回归转化为多层标准一维输出变量回归来解决这个问题，并与十进制编码的遗传算法相结合，形成改进的GA-SVR算法，用遗传算法搜索最优的SVR模型参数以建立最优的待辨识参数与位移之间的非线性映射关系，然后用遗传算法进行待辨识参数的最优辨识。为对比这种改进GA-SVR算法的效果，将遗传算法与BP神经网络相结合，形成GA-BP算法且编制相应的计算程序。将这两种算法运用于同样的隧道工程三维弹塑性模型参数的智能辨识，数值算例表明改进的GA-SVR算法较GA-BP算法可以取得更高的辨识精度和更好的计算效率，可运用于类似岩土工程计算参数的辨识。     

基于遗传算法解决附有不等式约束的最小二乘平差问题的研究

测量数据处理中经常有些先验信息可以利用,这些先验信息可以总结成等式或不等式。附等式约束的平差理论目前已经十分成熟,因而如果是等式约束,则可用附等式约束的间接平差方法来处理。但如果是不等式约束,则计算相对困难。Holland等人提出的遗传算法在最优化计算中取得了非常好的效果,本文尝试将遗传算法引入附不等式约束的平差计算中。本文首先介绍了附不等式约束的最小二乘平差模型,分析了基于遗传算法解决该问题的理论依据,进而通过用内罚函法将不等式约束平差转化为无约束平差,以方便运用遗传算法,最终调用Matlab遗传工具箱来求解平差结果。通过实例分析,该算法同其它常用的算法进行比较,证明该方法具有快速的收敛性,求解结果良好。     

Reliability-based optimization in structural engineering

In this paper reliability-based optimization problems in structural engineering are formulated on the basis of the classical decision theory. Several formulations are presented: Reliability-based optimal design of structural systems with component or systems reliability constraints, reliability-based optimal inspection planning and reliability-based experiment planning. It is explained how these optimization problems can be solved by application of similar techniques. The reliability estimation is limited to first order reliability methods (FORM) for both component and systems reliability evaluation. The solution strategies applying first order non-linear optimization algorithms are described in detail with special attention to sensitivity analysis and stability of the optimization process. Furthermore, several practical aspects are treated as: Development of the reliability-based optimization model, inclusion of the finite element method as the response evaluation tool and how the size of the problem can be made practicable. Finally, the important task of model evaluation and sensitivity analysis of the optimal solution is treated including a strategy for model-making with both pre and post-analysis.     

遗传算法和高斯牛顿法联合反演地下水渗流模型参数

充分利用遗传算法善于进行全局搜索和高斯牛顿法善于进行局部搜索的优点,克服了两种方法各自的不足,用改进的遗传算法和高斯牛顿法联合反演地下水数值模型参数,首先用遗传算法求出地下水模型参数的初值,然后利用这组初值用高斯牛顿法进行数值模型参数的反演,并以非均质各向同性承压二维非稳定流动模型,结合有限元法讨论了用遗传算法和高斯牛顿法联合反演地下水数值模型参数的过程。计算结果表明,联合参数反演方法,具有收敛速度快、解的精度高的特点。     

The optimization of architectural shape based on Genetic Algorithm

Genetic Algorithm (GA) is widely adopted in optimization and the improvement of its optimization performance is attracting many researchers’ attentions. In solving practical problems in the process of architectural design, the ways of converting design problems into mathematical models that can be addressed by GA are of great significance in achieving final optimal results. However, no such rule that can be applied to such conversion has been developed so far. In general, problems which can be addressed by GA can be divided into combinatorial problems and numerical problems. In this paper, by means of attempting to disintegrate a complicated architectural problem into combinatorial and numerical problems, the author discusses feasibility and practicality of solving these two types of problems simultaneously utilizing GA and discloses both advantages and disadvantages of GA by comparing with other algorithms.     

基于大跨空间网架结构主动控制的新型超磁致伸缩材料作动器优化配置研究

首先基于新型超磁致伸缩材料(GMM)设计了一种新型作动器,并对其进行了试验研究,获得了其本构方程.利用结构振动模态方程提出了遗传优化算法中的主要参数——优化性能指标;利用MATLB所提供的遗传算法工具箱并结合前述结论确定了作动器最优布置位置.最后利用所编制的LQR主动控制程序对一平板网架结构进行了不布置及随机布置与优化布置该作动器下的地震响应对比分析.结果表明:优化布置该新型作动器后结构地震响应明显减弱,从而验证了该新型作动器在结构主动控制中应用的有效性及遗传优化算法对于此类问题进行研究的可行性.     

Application of evolutionary global optimization techniques in the design of RC water tanks

This paper presents evolutionary-based optimization procedures for designing conical reinforced concrete water tanks. The material cost of the tank that includes concrete, reinforcement, and formwork required for walls and floor was chosen as the objective function in the nonlinear optimization problem formulation. The wall thicknesses (at the bottom and at the top), base thickness, depth of water tank, and wall inclination were considered as design variables.Three advanced optimization techniques to solve the nonlinear constrained structural optimization problems were investigated. These methods are: (1) Shuffled Complex Evolution (SCE), (2) Simulated Annealing (SA) and Genetic Algorithm (GA). Several tests were performed to illustrate the robustness of these techniques and results were encouraging for SCE Method. The SCE method proved to be superior to the SA and GA methods in obtaining the best discovered solutions. The paper concludes that the robust search capability of SCE algorithm technique is well suited for solving the structural problem in hand.     

粒子群优化算法在桁架优化设计中的应用

粒子群优化（PSO）算法是近年来发展起来的一种基于群智能的随机优化算法,具有概念简单、易于实现、占用资源低等优点。为了解决有应力约束和位移约束的桁架的尺寸优化问题,将PSO算法应用于桁架结构的尺寸优化设计。首先介绍了原始的PSO算法的基本原理,然后引入压缩因子改进了PSO算法,并提出合理的参数设置值。对几个经典问题进行了求解,并与传统的优化算法和遗传算法进行了比较。数值结果表明,改进的PSO算法具有良好的收敛性和稳定性,可以有效地进行桁架结构的尺寸优化设计。     

基于模式-遗传-神经网络的流变参数反演

介绍了一种岩石流变多参数反演的智能方法。该方法把遗传算法和神经网络有机结合起来,并在遗传算法中嵌入模式搜索加速优化进程;该方法基于均匀设计获得的样本进行神经网络学习,用模式–遗传–神经网络进行岩体流变参数的最优辩识。该方法用经过最佳预测学习算法训练的神经网络来表达岩体流变参数和位移之间的映射关系,除具有一般遗传算法的优点外,还提高了参数反演的精度,节省了参数反演的计算时间,使得某些原来用传统优化方法在时间上几乎无法进行的参数反演如今变为可能,并用工程实例验证了此方法的可行性与优越性。     

Design optimization of 3D steel structures: Genetic algorithms vs. classical techniques

In this paper an elitist genetic algorithm (GA) developed by the authors is compared with common commercial solutions for complex structural optimization. After its prior validation on two-dimensional (2D) structures, the GA was tuned and improved in order to obtain structural elements with minimum weights that satisfy the ultimate limit states of the applicable building code. Subsequently, the same spatial structures were optimized using a commercial structural analysis program. Finally, the cost and weight improvements obtained using the GA, although at a higher computational cost, are discussed.