Development of a Methodology for Determination and Analysis of Thermal Displacements of Machine Tools Using Finite Elements Method and Artificial Neural Network

In the processes of manufacturing, MT （machine tools） plays an important role in the manufacture of work pieces with complex and high dimensional and geometric accuracy. Much of the errors of a machine tool are those which are thermally induced which are from internal and external heat sources acting on the machine. In this paper, a methodology for determining and analyzing the thermal deformation of machine tools using FEM （finite element method） and ANN （artificial neural networks） is presented. After modeling the machine using FEM is defined the location of the heat sources, it is possible to obtain the temperature gradient and the corresponding thermal deformation at predetermined periods. Results obtained with simulations using the software NX.7.5 showed that this methodology is an effective tool in determining the thermal deformation of the machine, correlating the temperature reading at strategic points with volumetric deformation at the tool tip. Therefore, the thermal analysis of the errors in the pair tool part can be established. After training and validation process, the network will be able to make the prediction of thermal errors just stating the temperature values of specific points of each heat source, providing a way for compensation of thermally induced errors.     

A Wind Tunnel Two-Dimensional Parametric Investigation of Biplane Configurations

This paper presents an experimental and systematic investigation about how geometric parameters on a biplane configuration have an influence on aerodynamic parameters. This experimental investigation has been developed in a two-dimensional approach. Theoretical studies about biplanes configurations have been developed in the past, but there is not enough information about experimental wind tunnel data at low Reynolds number. This two-dimensional study is a first step to further tridimensional investigations about the box wing configuration. The main objective of the study is to find the relationships between the geometrical parameters which present the best aerodynamic behavior： the highest lift, the lowest drag and the lowest slope of the pitching moment. A tridimensional wing-box model will be designed following the pattern of the two dimensional study conclusions. It will respond to the geometrical relationships that have been considered to show the better aerodynamic behavior. This box-wing model will be studied in the aim of comparing the advantages and disadvantages between this biplane configuration and the plane configuration, looking for implementing the box-wing in the UAV＇s field. Although the box wing configuration has been used in a small number of existing UAV, prestigious researchers have found it as a field of high aerodynamic and structural potential.     

Acoustic Resonance in Exhaust Gas Economizer Tube Bundles Prediction and Suppression of Resonance

The tendency for air column resonance generation in structures with a constant volume behind a tube array like that of an exhaust gas economizer has been reported, but many points remain unclear with respect to the mechanism and conditions that generate acoustical resonance. When acoustical resonance is generated, in reality, prevention and suppression measures are implemented by inserting a baffle plate into the ducts through a process of trial and error. The purpose of this study is to clarify the mechanism of generation of acoustical resonance, and to establish an appropriate measure to prevent such resonance. Noise generated in an exhaust gas economizer was correlated with the flow inside the tube array and experimentally analyzed, and the mechanism for resonance generation was considered. In addition, the effectiveness of a baffle plate positioned in order to prevent resonance was investigated. We have successfully employed a single baffle plate to suppress resonance.     

Position control optimization of aerodynamic brake device for high-speed trains

The aerodynamic braking is a clean and non-adhesion braking, and can be used to provide extra braking force during high-speed emergency braking. The research of aerodynamic braking has attracted more and more attentions in recent years. However, most researchers in this field focus on aerodynamic effects and seldom on issues of position control of the aerodynamic braking board. The purpose of this paper is to explore position control optimization of the braking board in an aerodynamic braking prototype. The mathematical models of the hydraulic drive unit in the aerodynamic braking system are analyzed in detail, and the simulation models are established. Three control functions--constant, linear, and quadratic--are explored. Two kinds of criteria, including the position steady-state error and the acceleration of the piston rod, are used to evaluate system performance. Simulation results show that the position steady state-error is reduced from around 12-2 mm by applying a linear instead of a constant function, while the acceleration is reduced from 25,71-3.70 m/s2 with a quadratic control function. Use of the quadratic control function is shown to improve system performance. Experimental results obtained by measuring the position response of the piston rod on a test-bench also suggest a reduced position error and smooth movement of the piston rod. This implies that the acceleration is smaller when using the quadratic function, thus verifying the effectiveness of control schemes to improve to system performance. This paper proposes an effective and easily implemented control scheme that improves the position response of hydraulic cylinders during position control.     

基于BP神经网络的压缩机性能预测模型的建立

利用BP网络,以全年运行的空气源热泵冷热水机组性能预测为目标,对半封闭往复式压缩机建立了性能预测模型。以网络的泛化能力作为衡量建模效果的标准,同时指出模型的建立过程就是网络最佳结构参数的搜索过程。计算实例显示,网络的预测效果与训练效果吻合较好,训练代价适中,所建模型可用于压缩机全年运行的性能预测与故障诊断中。     

Electromagnetic Torque Estimation of a Three-Phase Induction Motor for Setting the Feedrate of a Milling Machine

In this paper, it presents a project of a fuzzy controller and a neural estimator to control a coordinate table powered by three-phase induction motor, aiming to implement an intelligent milling system. The position/speed control is performed using vector techniques of three-phase induction machines. The estimation of the motor electromagnetic torque is used for setting the feedrate of the table. The speed control is developed using TS （Takagi-Sugeno） fuzzy logic model and electromagnetic torque estimation using neural network type LMS （least mean square） algorithm. The induction motor is powered by a frequency inverter driven by a DSP （digital signal processor）. Control strategies are implemented in DSP. Simulation results are presented for evaluating the performance of the system.     

Design and Fabrication of Chitosan for Application of Artificial Photosynthesis

Artificial photosynthesis is a new approach to generate sustainable energy. In order to constrain reaction solution in a solid state structure and increase the reaction efficiency in artificial photosynthesis reactions, we presented two methods to fabricate the chitosan scaffold with interconnected micro channels as construction structure of a novel artificial photosynthesis device. We built 3D chitosan structure with a home-made heterogeneous 3D rapid prototyping machine, and we used lyophilization method to generate the micron-scale pores inside the chitosan scaffold. Chitosan in acetic acid could achieve different viscosities. We found a proper chitosan recipe to construct 3D scaffold by our own rapid prototyping machine. Optional support material sodium bicarbonate was used in printing 3D scaffold for holding the printed structures, and the results images indicate that this method can make the scaffold stronger and more stable.     

Prediction of Wing Aerodynamic Performance in Rain Using Neural Net

A new method for prediction of wing aerodynamic performance in rain condition was presented.Three-and four-layer artificial neural networks based on improved algorithm for error Back Propagation(BP)network were respectively built.Detailed approaches to determine the optical parameters for network model were introduced and the specific steps for applying BP network model to predict wing aerodynamic performance in rain were given.On this basis,the established optimal three-and four-layer BP network model was used for this prediction.Results indicate that both of the network models are appropriate for predicting wing aerodynamic performance in rain.The sum of square error level produced by two models is less than 0.2%,and the prediction accuracy by four-layer network model is higher than that of three-layer network.     

人工神经网络——阳极溶出伏安标准曲面法测定水体中元素的化学形态分布

建立了人工神经网络阳极溶出伏安标准曲面法,并将其成功地用于Pb~(2+)-Cd~(2+)-OH~--Cl~-混合体系中铅和镉的累积稳定常数和化学形态分布计算。     

Reconstruction of Turbulent Swirling Flow in a Dump Combustor

Experimental data of the continuous evolution of fluid flow characteristics in a dump combustor is very useful and essential for better and optimum designs of gas turbine combustors and ramjet engines. Unfortunately, experimental techniques such as 2D and/or 3D LDV （Laser Doppler Velocimetry） measurements provide only limited discrete information at given points; especially, for the cases of complex flows such as dump combustor swirling flows. For this type of flows, usual numerical interpolating schemes appear to be unsuitable. Recently, neural networks have emerged as viable means of expanding a finite data set of experimental measurements to enhance better understanding of a particular complex phenomenon. This study showed that generalized feed forward network is suitable for the prediction of turbulent swirling flow characteristics in a model dump combustor. These techniques are proposed for optimum designs of dump combustors and ramjet engines.     

水电机组振动故障诊断的人工神经网络选择研究

引起水电机组振动的原因很复杂，而且水电机组的振动故障往往是多故障同时发生，使得故障诊断很困难，目前主要是应用基于模式识别的神经网络来进行故障分类，但如何选择故障诊断的神经网络一直是个难点。文章研究了3种人工神经网络，即反向传播网络（BPN）、概率神经网络（PNN）和学习矢量量化网络（LVQ）对水电机组振动故障诊断性能的影响。结果表明，人工神经网络的结构和算法，包括相关训练参数的选择对故障诊断性能有着重要影响。比较而言，学习矢量量化网络和概率神经网络在分类能力方面要比反向传播网络好一些，概率神经网络在计算负载方面比学习矢量量化网络要更胜一筹。     

Economic Assessment for Optimum Wind Farm Investment Connected to a Distribution System

Decarbonization of electricity industry for the goal of sustainability success has resulted in large investment in alternative energy sources such as wind, solar, biomass. Although these energy resources are sustainable and have the potential of reducing the world carbon foot print, there are costs associated with its utilization. In recent time, electricity from alternative energy sources like wind and solar are not cost competitive with electricity from the conventional power plant. This paper is aimed at investigating the optimum investment in a typical wind farm project using a TSA （time series analysis） alongside simple economic tool, AAP （annual annuity payment） model. This study involves a year round analysis of （8,760h） at different wind farm capacity connected to a 132/33kV DS （distribution system）. It also focused on digressing from the technical and environmental benefits to financial assessment of increasing wind generation capacity in the DS. Indeed, this development presents a risk of investment to the stakeholders which necessitates proper scrutiny and to ensure profitability of the venture. The level of capital cost along with operation and maintenance （OM） costs are either financed by private or public sectors on wind farm with the sole aim of achieving the ROI （return-on-investment）. The results obtained from this study shows the possible ROI is not proportional to the wind capacity invested. Also, a sensitivity analysis conducted revealed the profit derived from wind farm is more responsive to the investment/capital cost and the price at which the electricity is being sold.     

基于人工智能的配网网络发令系统及应用

提出了基于人工智能的配网网络发令系统。首先分析配网调度指令的特点,说明网络下令与传统下令的区别。针对人工智能技术在数据处理、数据训练等方面在调度的应用,以调度语音识别为对象,分析人工神经网络在语音特征提取方面的流程。建立了配网网络发令系统,提出系统构架,说明系统功能,并以倒闸操作为例说明网络下令的流转过程以及智能决策应用。     

级环境下斜流压气机叶片扩压器气动优化设计

以某斜流压气机的叶片扩压器为研究对象,采用人工神经网络和遗传算法相结合对其进行气动优化设计.优化后,斜流级总压比提高了3.85％,效率提高了2.07％.叶片扩压器静压恢复系数和总压恢复系数也分别大幅提高至0.7和0.95的水平.与原方案相比,扩压器叶片最大负荷点从10％弦长后移至25％弦长.扩压器叶栅通道靠近机匣区域的...     

Multi-disciplinary design optimization and performance evaluation of a single stage transonic axial compressor

The multidisciplinary design optimization method, which integrates aerodynamic performance and structural stability, was utilized in the development of a single-stage transonic axial compressor. An approximation model was created using artificial neural network for global optimization within given ranges of variables and several design constraints. The genetic algorithm was used for the exploration of the Pareto front to find the maximum objective function value. The final design was chosen after a second stage gradient-based optimization process to improve the accuracy of the optimization. To validate the design procedure, numerical simulations and compressor tests were carried out to evaluate the aerodynamic performance and safety factor of the optimized compressor. Comparison between numerical optimal results and experimental data are well matched. The optimum shape of the compressor blade is obtained and compared to the baseline design. The proposed optimization framework improves the aerodynamic efficiency and the safety factor.     

基于人工神经网络的最大充放电功率预测方法研究

针对混合动力汽车中镍氢电池组模块,通过人工神经网络算法预测出其在下一时刻的最大充放电功率值.首先,通过查匹配表获得目标值与SOC(剩余电荷量)、电压、温度的对应关系,再选取与目标值相关联的SOC、电压、电流、温度,加上上一时刻预测出的最大充放电功率值作为该人工神经网络的输入变量,通过实验对人工神经网络结构算法进行设计优化.最终得到最大充放电功率预测用人工神经网络模型,经实际数据测试分析,其误差小于8％.此预测对于混合动力汽车在启动和爬坡控制策略有重要的实用意义.     

GENERALISED TRANSFER FUNCTIONS OF NEURAL NETWORKS

When artificial neural networks are used to model non-linear dynamical systems, the system structure which can be extremely useful for analysis and design, is buried within the network architecture. In this paper, explicit expressions for the frequency response or generalised transfer functions of both feedforward and recurrent neural networks are derived in terms of the network weights. The derivation of the algorithm is established on the basis of the Taylor series expansion of the activation functions used in a particular neural network. This leads to a representation which is equivalent to the non-linear recursive polynomial model and enables the derivation of the transfer functions to be based on the harmonic expansion method. By mapping the neural network into the frequency domain information about the structure of the underlying non-linear system can be recovered. Numerical examples are included to demonstrate the application of the new algorithm. These examples show that the frequency response functions appear to be highly sensitive to the network topology and training, and that the time domain properties fail to reveal deficiencies in the trained network structure.     

运载火箭气动外形CAD/CFD集成设计系统研究

运载火箭模块化设计带来了"组合爆炸"的难题,为了提高气动外形设计和仿真的效率,开展了数字化集成设计技术研究。首先,规划了系统总体架构,设计了开放的四层软件体系架构,并定制了软件工作流程。然后,对气动设计参数进行分类,建立了设计模板库。对Pro/E进行了二次开发,实现了几何建模参数化和模型干涉检查等功能。基于Python语言,编写了参数化网格生成脚本,综合应用CFD-GEOM、Gambit、Fluent等工具软件,实现了仿真分析自动化。基于VC++6.0开发了图形用户界面,实现了设计过程、研制数据和应用工具的系统集成。最后,通过某型运载火箭的应用实例,验证了系统的正确性和有效性。     

嵌入式神经网络加速器及 SoC 芯片

为了提高人工智能加速器的运算效率和功耗效率,提出了一种新的卷积神经网络(CNN)加速器结构,并实现了神经网络存算一体的方法。首先,设计出一种神经网络架构,其具有高度并行计算以及乘加器(MAC)单元高效运行的特性。其次,为了降低功耗和面积,采用了对称的静态随机存储器(SRAM)阵列和可调数据流向结构,实现多层网络在SRAM中高效计算,减少了访问外部存储器次数,降低了功耗,提高运算效率。通过中芯国际40 nm工艺,完成了系统芯片(SoC)设计、流片与测试。结果表明运算速度在500 MHz下,算力可达288 GOPS;全速运行功耗89.4 mW;面积1.514 mm~2;算力功耗比3.22 TOPS/W;40 nm算力面积比为95.1 GOPS/mm~2。与已有文献的相比,算力功耗至少提升4.54%,算力面积至少提升134%,对于嵌入式场景应用较适合。