Modelling of structural response and optimization of structural control system using neural network and genetic algorithm

This paper proposes an integrated approach to the modelling and optimization of structural control systems in tall buildings. In this approach, an artificial neural network is applied to model the structural dynamic responses of tall buildings subjected to strong earthquakes, and a genetic algorithm is used to optimize the design problem of structural control systems, which constitutes a mixed‐discrete, nonlinear and multi‐modal optimization problem. The neural network model of the structural dynamic response analysis is included in the genetic algorithm and is used as a module of the structural analysis to estimate the dynamic responses of tall buildings. A numerical example is presented in which the general regression neural network is used to model the structural response analysis. The modelling method, procedure and the numerical results are discussed. Two Los Angeles earthquake records are adopted as earthquake excitations. Copyright © 2000 John Wiley & Sons, Ltd.     

A genetic-algorithms based evolutionary computational neural network for modelling spatial interaction dataNeural network for modelling spatial interaction data

Building a feedforward computational neural network model (CNN) involves two distinct tasks: determination of the network topology and weight estimation. The specification of a problem adequate network topology is a key issue and the primary focus of this contribution. Up to now, this issue has been either completely neglected in spatial application domains, or tackled by search heuristics (see Fischer and Gopal 1994). With the view of modelling interactions over geographic space, this paper considers this problem as a global optimization problem and proposes a novel approach that embeds backpropagation learning into the evolutionary paradigm of genetic algorithms. This is accomplished by interweaving a genetic search for finding an optimal CNN topology with gradient-based backpropagation learning for determining the network parameters. Thus, the model builder will be relieved of the burden of identifying appropriate CNN-topologies that will allow a problem to be solved with simple, but powerful learning mechanisms, such as backpropagation of gradient descent errors. The approach has been applied to the family of three inputs, single hidden layer, single output feedforward CNN models using interregional telecommunication traffic data for Austria, to illustrate its performance and to evaluate its robustness.     

QPSO-ILF-ANN-based optimization of TBM control parameters considering tunneling energy efficiency

In recent years, tunnel boring machines (TBMs) have been widely used in tunnel construction. However, the TBM control parameters set based on operator experience may not necessarily be suitable for certain geological conditions. Hence, a method to optimize TBM control parameters using an improved loss function-based artificial neural network (ILF-ANN) combined with quantum particle swarm optimization (QPSO) is proposed herein. The purpose of this method is to improve the TBM performance by optimizing the penetration and cutterhead rotation speeds. Inspired by the regularization technique, a custom artificial neural network (ANN) loss function based on the penetration rate and rock-breaking specific energy as TBM performance indicators is developed in the form of a penalty function to adjust the output of the network. In addition, to overcome the disadvantage of classical error backpropagation ANNs, i.e., the ease of falling into a local optimum, QPSO is adopted to train the ANN hyperparameters (weight and bias). Rock mass classes and tunneling parameters obtained in real time are used as the input of the QPSO-ILF-ANN, whereas the cutterhead rotation speed and penetration are specified as the output. The proposed method is validated using construction data from the Songhua River water conveyance tunnel project. Results show that, compared with the TBM operator and QPSO-ANN, the QPSO-ILF-ANN effectively increases the TBM penetration rate by 14.85% and 13.71%, respectively, and reduces the rock-breaking specific energy by 9.41% and 9.18%, respectively.     

水质综合评价的Hopfield网络模型

将人工神经网络技术运用于水质评价，建立了水质综合评价的Hopfield模型，并以重庆市南川凤嘴江水质资料为实例，对模型精确度进行了测试。结果表明：Hopfield网络用于评价水环境质量形象、客观、准确、合理,具有广阔的应用前景。     

Neural Network-Based Long-Term Hydropower Forecasting System

Artificial neural networks are alternatives to stochastic models even if the optimization of their architectures remains a tricky problem. Two different approaches in long-term forecasting of potential energy inflows using a feedforward neural network (FNN) and a recurrent neural network (RNN) are proposed. The problem of overfitting, particularly critical for limited hydrologic data records, is addressed using a new approach entitled optimal weight estimate procedure (OWEP). The efficiency of the two models using OWEP is assessed through multistep forecasts. The experiment results show that, in general, OWEP improves the models' performance and significantly reduces the training time on the order of 60 percent. The RNN outperforms the FNN but costs about a factor of 2 longer in training time. Furthermore, the neural network-based models provide more accurate forecasts than traditional stochastic models. Overall, the RNN appears to be the best suited for potential energy inflows forecasting and therefore for hydropower systems management and planning.     

结构优化设计中神经网络方法的应用

指出了传统优化设计方法的不足 ,概述了人工神经网络的基本特征和学习规则 ,提出了结构优化问题的神经网络方法设计原理和步骤 ,综述了对神经网络在结构优化中的应用。最后 ,研究了神经优化计算下一步需解决的难题     

Modelling hook times of mobile cranes using artificial neural networks

The hook times of mobile cranes are processes that are of non‐linear and discrete nature. Artificial neural network is a data processing technique that lends itself to this kind of problem. Three common artificial neural network architectures – multi‐layer feed‐forward (MLFF), group method of data handling (GMDH) and general regression neural network (GRNN) – are compared. The results show that the GRNN model aided with genetic algorithm (GA) is most promising in describing the non‐linear and discrete nature of the hook times. The MLFF model can also give a moderate level of accuracy in the estimation of hook travelling times of mobile cranes and is ranked second. The GMDH model is outperformed by the former two due to a less promising R‐square.     

Architectural layout optimization using annealed neural network

The facility layout problem is concerned with finding feasible locations for a set of interrelated objects that meet all design requirements and maximize design quality in terms of design preferences. The contribution of this paper is a new framework, named annealed neural network, for efficiently finding competitive solutions for the facility layout problem. This framework arises from the combination of Hopfield neural networks and simulated annealing. The first is a representation model of the layout problem and the second is a search algorithm for finding the optimum or near optimum solutions. The annealed neural network combines characteristics of the simulated annealing algorithm and the Hopfield neural network. Annealed neural network exhibits the rapid convergence of the neural network, while preserving the solution quality afforded by simulated annealing. Strategies for setting reasonable penalty factor in objective function and temperature in simulated annealing procedure were proposed. A case study of a hospital building with 28 facilities was employed to demonstrate that this model is rather efficient to solve the architectural layout problem, and it is amenable to fast computation for large layout problems.     

Crack detection in beam-like structures

Sensibility analysis of experimentally measured frequencies as a criterion for crack detection has been extensively used in the last decades due to its simplicity. However the inverse problem of the crack parameters (location and depth) determination is not straightforward. An efficient numerical technique is necessary to obtain significant results. Two approaches are herein presented: The solution of the inverse problem with a power series technique (PST) and the use of artificial neural networks (ANNs). Cracks in a cantilever Bernoulli–Euler (BE) beam and a rotating beam are detected by means of an algorithm that solves the governing vibration problem of the beam with the PST. The ANNs technique does not need a previous model, but a training set of data is required. It is applied to the crack detection in the cantilever beam with a transverse crack. The first methodology is very simple and straightforward, though no optimization is included. It yields relative small errors in both the location and depth detection. When using one network for the detection of the two parameters, the ANNs behave adequately. However better results are found when one ANN is used for each parameter. Finally, a combination between the two techniques is suggested.     

基于人工神经网络的爆炸冲击荷载参数识别方法

根据观测的爆破振动响应数据和有限元正演振动分析数据，建立了基于人工神经网络的爆炸冲击荷载参数识别方法，采用Levenberg-Marquardt优化方法修正网络的权值和阈值，大大提高了神经网络的收敛速度。研究了观测噪音对爆炸荷载参数识别结果的影响。数值计算结果表明，所建立的基于人工神经网络的爆炸冲击荷载参数识别方法具有良好的鲁棒性和抗观测噪声能力。     

Neural extension of experimental data to investigate using phosphogypsume in light brick production

In this study, usability of wastes produced in phosphoric acid plants in structural brick manufacture has been investigated. There are several parameters involved in using these wastes in brick production namely the rate of added waste, firing speed and firing temperature. The performance of these parameters can be measured by several criteria such as natural drying shortening, water absorption and weight loss. Therefore, so many experiments are needed to investigate the effects of these parameters on the bricks produced with these wastes. The result of a series of experiments were utilized to achieve this end. The results have shown that the industrial wastes considered improve the performance of the bricks in terms of the criteria mentioned above. However, the results have also shown that further investigations are needed to explore the effects of interim values on the performance of the bricks. To achieve that end, a neural experimental study is adopted. For this purpose, the results of the experiments conducted were used to construct an artificial neural network. The trained and tested network was then used to check the effects of 280 different combinations for each type of material mixtures mentioned. The outcome of these artificial tests have provided the optimal values for the waste addition rate, firing speed and firing temperature based on the four criteria mentioned previously.     

砂土液化预测的人工神经网络模型

本文根据人工神经网络的一典型模型-反向传播模型,以及地震荷载下的各项土的物理-力学参数,建立了土液化类型的神经网络数学模型。研究表明,人工神经网络方法性能良好,可望成为砂土液化预测的有效手段。     

A hybrid artificial neural network as a software sensor for optimal control of a wastewater treatment process

For control and automation of biological treatment processes, lack of reliable on-line sensors to measure water quality parameters is one of the most important problems to overcome. Many parameters cannot be measured directly with on-line sensors. The accuracy of existing hardware sensors is also not sufficient and maintenance problems such as electrode fouling often cause trouble. This paper deals with the development of software sensor techniques that estimate the target water quality parameter from other parameters using the correlation between water quality parameters. We focus our attention on the preprocessing of noisy data and the selection of the best model feasible to the situation. Problems of existing approaches are also discussed. We propose a hybrid neural network as a software sensor inferring wastewater quality parameter. Multivariate regression, artificial neural networks (ANN), and a hybrid technique that combines principal component analysis as a preprocessing stage are applied to data from industrial wastewater processes. The hybrid ANN technique shows an enhancement of prediction capability and reduces the overfitting problem of neural networks. The result shows that the hybrid ANN technique can be used to extract information from noisy data and to describe the nonlinearity of complex wastewater treatment processes.     

基于粒子群优化算法的城市日用水量预测模型

结合城市日用水量影响因素的特点和变化规律，建立了城市日用水量预测模型，采用粒子群优化算法优化BP人工神经网络的连接权值，以求解该预测模型。经优化后的BP人工神经网络运算速度快、泛化能力强、预测精度高。实例验证结果证明该日用水量预测模型和求解方法是可行的。     

Calibration of the descent local search algorithm parameters using orthogonal arrays

Solving optimization problems using heuristic algorithms requires the selection of its parameters. Traditionally, these parameters are selected by a trial and error process that cannot guarantee the quality of the results obtained because not all the potential combinations of parameters are checked. To fill this gap, this paper proposes the application of Taguchi's orthogonal arrays to calibrate the parameters of a heuristic optimization algorithm (the descent local search algorithm). This process is based on the study of the combinations of discrete values of the heuristic tool parameters and it enables optimization of the heuristic tool performance with a reduced computational effort. To check its efficiency, this methodology is applied to a technical challenge never studied before: the optimization of the tensioning process of cable‐stayed bridges. The statistical improvement of the heuristic tool performance is studied by the optimization of the tensioning process of a real cable‐stayed bridge. Results show that the proposed calibration technique provided robust values of the objective function (with lower minimum and mean values, and lower standard deviation) with reduced computational cost.     

Optimal design of double-layer barrel vaults using genetic and pattern search algorithms and optimized neural network as surrogate model

This paper presents a combined method based on optimized neural networks and optimization algorithms to solve structural optimization problems. The main idea is to utilize an optimized artificial neural network (OANN) as a surrogate model to reduce the number of computations for structural analysis. First, the OANN is trained appropriately. Subsequently, the main optimization problem is solved using the OANN and a population-based algorithm. The algorithms considered in this step are the arithmetic optimization algorithm (AOA) and genetic algorithm (GA). Finally, the abovementioned problem is solved using the optimal point obtained from the previous step and the pattern search (PS) algorithm. To evaluate the performance of the proposed method, two numerical examples are considered. In the first example, the performance of two algorithms, OANN + AOA + PS and OANN + GA + PS, is investigated. Using the GA reduces the elapsed time by approximately 50% compared with using the AOA. Results show that both the OANN + GA + PS and OANN + AOA + PS algorithms perform well in solving structural optimization problems and achieve the same optimal design. However, the OANN + GA + PS algorithm requires significantly fewer function evaluations to achieve the same accuracy as the OANN + AOA + PS algorithm.     

人工神经网络在建筑结构中的应用研究

对日益广泛应用于建筑结构的人工神经网络的基本原理与特征以及误差反向传播的多层感知器网络(BP网络)的多种改进算法进行了介绍,分析了人工神经网络在建筑结构的优化、控制以及损伤诊断等领域中的应用情况,对人工神经网络的推广应用具有一定指导意义。     

Artificial neural network optimized by differential evolution for predicting diameters of jet grouted columns

A novel and effective artificial neural network (ANN) optimized using differential evolution (DE) is first introduced to provide a robust and reliable forecasting of jet grouted column diameters. The proposed computational method adopts the DE algorithm to tackle the difficulties in the training and performance of neural networks and optimize the four quintessential hyper-parameters (i.e. the epoch size, the number of neurons in a hidden layer, the number of hidden layers, and the regularization parameter) that govern the neural network efficacy. This approach is further enhanced by a stochastic gradient optimization algorithm to allow ‘expensive’ computation efforts. The ANN-DE is first trained using a prepared jet grouting dataset, then verified and compared with the prevalent machine learning tools, i.e. neural networks and support vector machine (SVM). The results show that, the ANN-DE outperforms the existing methods for predicting the diameter of jet grouting columns since it well balances training efficiency and model performance. Specifically, the ANN-DE achieved root mean square error (RMSE) values of 0.90603 and 0.92813 for the training and testing phases, respectively. The corresponding values were 0.8905 and 0.9006 for the optimized ANN, then, 0.87569 and 0.89968 for the optimized SVM, respectively. The proposed paradigm is bound to be useful for solving various geotechnical engineering problems regardless of multi-dimension and nonlinearity.     

位移反分析的自适应神经模糊推理方法

现有各种位移反分析方法均存在着这种或那种不足之处：基于最优化理论的位移反分析方法，解的稳定性较差，易陷入局部极小，反演参数较多时收敛速度较慢，且难以搜索到最优解；基于人工神经网络的位移反分析方法，当解空间稍大时便难以收敛到所需要的精度，且训练结果不具有唯一性，因而很难获得与实际岩体相吻合的反演结果；基于遗传进化的位移反分析方法，需对搜索过程进行大量经验性干预才能搜索到最优解；基于遗传进化和神经网络的位移反分析方法，亦只在较小的解空间内才有效。针对这些不足之处，应用自适应神经模糊推理系统的原理，建立了位移反分析的自适应神经模糊推理方法，并应用该方法对所设定的某一标准弹塑性问题的力学参数进行了反演，反演结果表明，在较大的解空间内，这种位移反分析方法收敛速度快、解的稳定性好、反演结果精度高，是一种优异的位移反分析方法。