Prediction of chloride permeability of concretes containing ground pozzolans by artificial neural networks

This research was to study the chloride penetration resistance of normal (W/B of 0.80, 0.62, 0.48) and high (W/B of 0.41, 0.35, 0.30) strength concretes containing ground pozzolans such as fly ash, bottom ash and rice husk ash using the rapid chloride penetration test and the immersion test methods. Furthermore, on the basis of this experimental data, an artificial neural network technique is carried out to derive an explicit artificial neural network formulation for the prediction of chloride permeability as a function of six input parameters: water to binder ratio, percent replacement, testing ages, pozzolans types, aggregate to cement ratio and the actual compressive strength. To verify the model, linear and non-linear regression equations are carried out and compared with the proposed artificial neural network prediction model. The results indicate that the incorporation of ground fly ash, ground bottom ash and ground rice husk ash substantially improve the workability and chloride permeability. The artificial neural network models have more accurate and precise prediction than linear and non-linear regression technique. Moreover, it is concluded that the artificial neural network models have a strong prediction capability of chloride penetration of concrete and can be easily expanded for the new additional database to re-train the network.     

神经网络在快速反射镜控制系统中的应用仿真

根据快速反射系统设计了模糊控制器，介绍了用神经网络实现模糊推量的方法，然后给出该网络模型用于快速反射镜跟踪控制系统的仿真结果。最合给出本文的结论。     

模糊神经网络在高层建筑横风向振动控制中的应用研究

提出了模糊神经网络方法控制高层建筑横风向风振反应。通过观测部分楼层加速度和控制力输出，建立了模糊神经网络控制器，解决了传统控制中有限的传感器数目对系统振动状态估计的困难．利用模糊神经网络控制器预测结构的控制行为，消除了闭环控制系统中存在的时滞。利用模糊神经网络控制器的自学习能力来确定模糊规则和语言变量隶属函数，解决了土木工程复杂结构模糊控制中，难于依据专家的主观经验来确定模糊控制规则和语言变量隶属函数等困难。模糊神经网络方法的优势在于算法自身的鲁棒性，处理结构非线性、参数不确定性及时变等问题的能力。通过对基准建筑的刚度不确定性分析，讨论了模糊神经网络控制器的鲁棒性。仿真分析表明，模糊神经网络控制策略能有效地抑制高层建筑的横风向风振反应，控制效果略优于LQG控制，而拥有LQG控制不具备的诸多优点。     

Microconcrete with partial replacement of Portland cement by fly ash and hydrated lime addition

The reduction in Portland cement consumption means lower CO₂ emissions. Partial replacement of Portland cement by pozzolans such as fly ash has its limitations due to the quantity of calcium hydroxide generated in the mix. In this work we have studied the contribution of the addition of hydrated lime to Portland cement + fly ash systems. We have also studied several levels of cement replacement, ranging from 15% to 75%.The best mechanical results were obtained replacing 50% of Portland cement by the same amount of fly ash plus the addition of hydrated lime (20% respect to the amount of fly ash). In these systems, an acid-base self-neutralization of the matrix has occurred through a pozzolanic reaction of fly ash with portlandite liberated in the hydration of Portland cement and the added hydrated lime. It has been identified for these mixtures a significant amount of hydrated gehlenite, typical reaction product from rich-alumina pozzolans.     

气辅注塑成型注气压力的模糊神经网络控制研究

针对气辅注塑成形的注气压力精确控制要求,设计了具有５层结构的模糊神经网络控制器和控制算 法,利用神经网络的学习能力实现对模糊逻辑规则的优化,改善了系统的适应性。对系统３段压力控制的仿真 分析,验证了模糊神经网络控制模型的可行性,控制效果良好。     

Properties and performance of silane: blended cement systems

The paper presents the results of a study dealing with the performance of water repellents on hardened blended cement pastes. Since on the European market Portland cement does not play the dominant role anymore and due to the new national and European policies concerning Greenhouse Gases and sustainability, cement manufacturers produce more and more blended cements (CEM II–CEM V). Nevertheless, the majority of experience concerning the efficacy of water repellents is gained from Portland cement; therefore knowledge in regard to the interactions of blended cement with water repellent agent is minimal. Two silane-based products were applied on ‘fresh’ and carbonated cement substrates containing limestone, fly ash, slag and trass, and were investigated in terms of their functionality. The evaluation of the treatments’ performance and effectiveness were assessed using various laboratory measurements. Hydrophobicity, water absorption, colour changes and the penetration depth of silanes into the substrate were evaluated before and after artificial aging experiments. Moreover, the outdoor weathering test was performed to shed light on treated surface appearance in a ‘real’ outdoor environment. The results showed that surface wettability was independent on water ingress or colour variations, especially for cement specimens artificially aged by accelerated carbonation. Cement pastes containing slag and trass seemed to more distinctly affect the water repellents’ surface performance.     

Abstracts

Abstract

In this paper, we present an algorithm for diagnosis of acute lymphoblastic leukaemia (ALL) in developing countries like Mexico. The proposed method uses a robust fuzzy logic algorithm to classify the ALL using microscopic cell images obtained by smears of bone marrow aspirates. We improve the results of fuzzy method using a radial basis function neural network to realise the diagnosis.     

基于自适应模糊逻辑和神经网络的双足机器人控制研究

在双足机器人行走控制中,为了改善系统的行走性能,提出了一种基于RBF神经网络前馈控制的力矩补偿控制方法。该方法将自适应模糊控制和神经网络逆模控制有效地结合起来,利用神经网络来逼近系统的逆动力学模型,提高系统了的控制性能,改善了机器人的行走特性。     

Joint Channel and Multi-User Detection Empowered with Machine Learning

The numbers of multimedia applications and their users increase with each passing day. Different multi-carrier systems have been developed along with varying techniques of space-time coding to address the demand of the future generation of network systems. In this article, a fuzzy logic empowered adaptive backpropagation neural network (FLeABPNN) algorithm is proposed for joint channel and multi-user detection (CMD). FLeABPNN has two stages. The first stage estimates the channel parameters, and the second performs multi-user detection. The proposed approach capitalizes on a neuro-fuzzy hybrid system that combines the competencies of both fuzzy logic and neural networks. This study analyzes the results of using FLeABPNN based on a multiple-input and multiple-output (MIMO) receiver with conventional partial opposite mutant particle swarm optimization (POMPSO), total-OMPSO (TOMPSO), fuzzy logic empowered POMPSO (FL-POMPSO), and FL-TOMPSO-based MIMO receivers. The FLeABPNN-based receiver renders better results than other techniques in terms of minimum mean square error, minimum mean channel error, and bit error rate.     

递归神经网络的模糊随机学习算法

提出了一种新的基于模糊逻辑的Alopex学习算法（FLA）。FLA算法利用模糊逻辑推理实时获得适应于学习过程的适当的算法修正值，克服了Alopex算法中修正值固定不变的缺点，使得随机学习过程在速度、精度和稳定性之间获得平衡。将该算法应用于神经网络的训练，可以无需神经网络的梯度信息和结构信息，因此可以用于具有各种结构特性的递归神经网络对动态系统的学习过程。实验结果表明了FLA算法的有效性。     

Speed control of SR motor by self-tuning fuzzy PI controller with artificial neural network

In this work, the dynamic model, flux-current-rotor position and torque-current-rotor position values of the switched reluctance motor (SRM) are obtained in MATLAB/Simulink. Motor control speed is achieved by self-tuning fuzzy PI (Proportional Integral) controller with artificial neural network tuning (NSTFPI). Performance of NSTFPI controller is compared with performance of fuzzy logic (FL) and fuzzy logic PI (FLPI) controllers in respect of rise time, settling time, overshoot and steady state error.     

Prediction of debonding strength for masonry elements retrofitted with FRP composites using neuro fuzzy and neural network approaches

This paper proposes application of neuro fuzzy and neural network for predicting debonding strength of retrofitted masonry elements. In order to achieve high-fidelity model, this study uses extensive experimental databases for bond test results between Fiber Reinforced Polymer (FRP) and masonry elements by collecting existing bond test subassemblage tests from the literature. Various influential parameters that affect debonding resistance including thickness of the FRP strip, width of the FRP strip, elastics modulus of the FRP, bonded length, tensile strength of the masonry block and width of the masonry block are considered as input parameters to the artificial neural network (ANN) and adaptive neuro fuzzy inference system (ANFIS). Test results of the ANN and ANFIS models were compared with multiple nonlinear regression, multiple linear regression and existing bond strength models. The accuracy of the optimal MNLR model was increased by 39% and 23% with respect to RMSE and MAE criteria using ANFIS. The comparison results indicated that the ANN and ANFIS models performed better than the other models and could be successfully used for prediction of debonding strength of retrofitted masonry elements.     

Prediction and optimization of a ceramic casting process using a hierarchical hybrid system of neural networks and fuzzy logic

This paper is a case study that describes a hybrid system integrating fuzzy logic, neural networks and algorithmic optimization for use in the ceramics industry. A prediction module estimates two quality metrics of slip-cast pieces through the simultaneous execution of two neural networks. A process improvement algorithm optimizes controllable process settings using the neural network prediction module in the objective function. An expert system module contains a hierarchy of two fuzzy logic rule bases. The rule bases prescribe processing times customized to individual production lines given ambient conditions, mold characteristics and the neural network predictions. This paper demonstrates the applicability of newer computational techniques to a very traditional manufacturing process and the system has been implemented at a major US plant.     

Calcium sulphate anhydrite based composite binders; effect of Portland cement and four pozzolans on the hydration and strength

The strength, hydration products, microstructure and heat of early hydration were investigated on alternative hydraulic green cements based on calcium sulphate anhydrite partially blended with Portland cement and pozzolans. Four pozzolans of different physical and chemical nature, namely a geothermal waste, silica fume, metakaolin and pulverized fuel ash were characterized. The composite binders showed hydraulic behavior. The use of Portland cement favoured the strength, which was also higher with the incorporation of siliceous nanometric pozzolans compared to the micrometric silicoaluminate pozzolans. The nanoparticles enhanced the early hydration and changed the gypsum morphology promoting denser matrices of hydration products. The geothermal waste pozzolan was the most effective, while one of the composites with metakaolin showed formation of ettringite and strength losses. The heat of hydration of the composites was considerably lower than that of the neat Portland cement.     

模糊神经技术在刀具故障检测中的应用

提出了一种刀具故障检测方法，该方法把模糊逻辑和神经网络结合起来，并用神经网络分解技术，建立了一个刀具状态识别网络。该网络适用于多传感器对刀具复杂状态进行识别和分类，具有训练时间短，执行速度快，可靠性高，抗噪能力强等特点。     

基于模糊神经网络的间接测量建模方法研究

在神经网络理论和模糊逻辑方法的基础上,将二者结合起来,讨论了一种基于模糊C均值聚类和径向基函数(RBF)的模糊神经网络,并将其应用于间接测量过程的非参量建模中。该方法尤其适用于非线性模型的构造,能够有效地提高测量的准确度和可靠性。     

An intelligent control for linear ultrasonic motor using interval type-2 fuzzy neural network

An interval type-2 fuzzy neural network (IT2FNN) is developed for the position control of a thetas-axis motion-control stage using a linear ultrasonic motor to confront the uncertainties of the motion-control stage. A T2FNN consists of a type-2 fuzzy linguistic process as the antecedent part and a three-layer interval neural network as the consequent part. A general T2FNN is computationally intensive due to the complexity of reducing type 2 to type 1. Therefore an IT2FNN is adopted to simplify the computational process. Moreover, the developed IT2FNN combines the merits of an interval type-2 fuzzy logic system and a neural network. Furthermore, the parameter-learning of the IT2FNN, which is based on the supervised gradient decent method using a delta adaptation law, is performed on line. Experimental results show that the dynamic behaviours of the proposed IT2FNN control system are more effective and robust with regard to uncertainties than the type-1 FNN control system.     

Modeling of constitutive relationship of Ti–25V–15Cr–0.2Si alloy during hot deformation process by fuzzy-neural network

In this paper, an adaptive fuzzy-neural network model has been established to model the constitutive relationship of Ti–25V–15Cr–0.2Si alloy during high temperature deformation. The network integrates the fuzzy inference system with a back-propagation learning algorithm of neural network. The experimental results were obtained at deformation temperatures of 900–1100 °C, strain rates of 0.01–10 s⁻¹, and height reduction of 50%. After the training process, the fuzzy membership functions and the weight coefficient of the network can be optimized. It has shown that the predicted values are in satisfactory agreement with the experimental results and the maximum relative error is less than 10%. It proved that the fuzzy-neural network was an easy and practical method to optimize deformation process parameters.     

智能预报模式与水文中长期智能预报方法

建立了以模糊优选、BP神经网络及遗传算法有机结合的智能预报模式与方法。在应用该方法进行中长期水文智能预报时，首先选取训练样本的数量，根据预报因子与预报对象的相关关系得到相对隶属度矩阵;再将其作为BP神经网络输入值以训练连接权重；最后将得到的连接权重值用于预报检验。计算结果表明，智能预报模式与方法的运行速度、精度及稳定性都达到了实际应用的要求。     

Application of a neural network approach to the electrophoretic deposition of PEEK-alumina composite coatings

Nano-size Al₂O₃-polyetheretherketone(PEEK) composite thick films have been prepared on stainless steel substrates from non-aqueous colloidal suspensions by electrophoretic deposition (EPD). The effects on the deposition efficiency of process parameters, such as the deposition time, the difference of potential applied and their interactions were studied using a neural network approach to develop a quantitative understanding of the system. Furthermore the use of the neural network was optimized in the number of epochs, hidden layers and artificial neurons in each hidden layer by a design of experiment (DOE) analysis, demonstrating that these two methods can work together improving the final results of the neural network approach. Afterwards, a MonteCarlo analysis based on a simulation of 100,000 virtual depositions has permitted to deeply investigate the effect of independent variables (e.g. deposition time and difference of potential applied) on the deposition yield (dependent variable).