基于遗传算法的超精密切削表面粗糙度预测模型参数辨识及切削用量优化

建立易于分析各切削用量对粗糙度影响关系的表面粗糙度预测模型和最优的切削用量组合,是超精密切削加工技术的不断发展的需要。针对最小二乘法和传统优化方法的不足,提出了将遗传算法用于超精密切削表面粗糙度预测模型的参数辨识,并用于求解最优切削用量,给出了金刚石刀具超精密切削铝合金的表面粗糙度预测数学模型和切削用量优化结果,进行了遗传算法和常规优化算法的比较,结果表明遗传算法较最小二乘法和传统的优化方法更适合于粗糙度预测模型的参数辨识及保证切削用量的最优。     

神经网络与遗传算法在拉延筋参数反求中的应用

以某车型前地板角支撑板的拉延工序为例,讨论BP神经网络技术与遗传算法在拉延筋几何参数反求中的综合应用问题,建立能描述反映成形效果的三个参数与半圆形拉延筋几何参数之间非线性映射关系的神经网络模型,并运用遗传算法对神经网络结构进行了优化。提出逐次局部密化样本点的样本点设计方法。该方法有助于加快神经网络的设计进程,提高神经网络的模拟精度。当训练样本数据可通过有限元法自动获得时,使用该方案则更为便利。     

基于遗传算法和神经网络的冷挤压工艺参数模糊优化设计

将模糊优化思想引入冷挤压工艺参数优化设计,以冷挤压模具的四个关键工艺参数为设计变量,以最终挤压件的各处材料的损伤值为目标函数,建立了冷挤压工艺参数模糊优化模型。提出了利用神经网络进行材料损伤近似计算的策略,从而形成了基于遗传算法和神经网络的冷挤压工艺参数模糊优化方法。对冷挤压工艺参数模糊优化模型进行了求解,优化结果表明模糊优化思想能提高冷挤压工艺设计质量。     

STUDY ON INJECTION AND IGNITION CONTROL OF GASOLINE ENGINE BASED ON BP NEURAL NETWORK

According to advantages of neural network and characteristics of operating procedures ofengine, a new strategy is rapresented on the control of fuel injection and ignition timing of gasolineengine based on improved BP network algorithm. The optimum ignition advance angle and fuelinjection pulse band of engine under different speed and load are tested for the samples trainingnetwork, focusing on the study of the design method and procedure of BP neural network in engine     

基于神经网络的立体视觉摄像机标定

摄像机标定是立体视觉研究的重要组成部分。根据神经网络与摄像机标定的特点,提出了基于神经网络的双目立体视觉摄像机标定方法,构建了三种不同的神经网络,并在试验中利用BP算法和实际数据对各种方法进行了分析和比较。     

NEURAL NETWORK APPROACH TO SURFACE BLENDING BASED ON DI GITIZED POINTS

0 INTRODUCTIONIn practical application of industrial area, thefreefOrm surface is often comprised of a lot of surfacepatches (pieces) and the generation of only a singlesurface is difficult to meet design requirements.Hence, we can say, the surface patches are the basicelements from which the surfaces are constructed,and they can be represented by mathematicalequations. Their form and accuracy depend on thefOllowing three factors: ro The wire-frame geometryform which the surfaces are con…     

一种基于小波网络的切削刀具故障监测

提出了一种基于小波神经网络的切削刀具故障监测方法，即提取反映刀具磨损状态的多源特征参数，利用小波神经网络的非线性模型和学习机制，实现在线状态监测；同时针对故障诊断的多输入输出问题带来的网络规模增大、收敛速度慢等问题，提出一种网络优化算法，即采用尺度参数的自适应调整法及平移参数的寻优搜索法，寻找最优小波基元，从而简化小波网络并加快收敛，仿真实例证明了该方法的有效性。     

复合材料加筋结构的神经网络响应面优化设计

针对复合材料加筋结构优化设计的复杂性,提出利用人工神经网络结构近似分析响应面来反映结构设计输入与结构响应输出的全局映射关系的优化方法。通过正交试验设计选取合适的结构有限元分析样本点,进行神经网络响应面的构建和训练;将神经网络响应面作为目标函数或者约束条件,汇同其他常规约束条件完成优化模型的建立,并应用遗传算法(GA)进行优化,从而形成一套适应性强的的高效优化方法。以复合材料翼身融合体帽型加筋板的质量优化为实例,建立加筋板模型的重量响应面目标函数、强度和翘曲稳定性响应面约束条件;通过PATRAN/NASTRAN有限元软件进行有限元计算,获取用于响应面训练的样本点数值。算例结果表明,该方法能以很少的有限元分析次数取得高精度的响应面近似模型,并且使优化计算耗时大为减少,优化效率大大提高。     

基于神经网络响应面的疲劳裂纹扩展寿命的可靠性分析

当失效形式的极限状态方程中随机变量个数较多或非线性较高时,其形式很复杂,因此传统的计算失效概率的方法不再适用。针对疲劳裂纹扩展寿命失效概率计算的复杂性,提出基于神经网络响应面的可靠性分析方法。首先建立神经网络响应面模拟疲劳裂纹扩展寿命的极限状态方程,然后使用遗传算法(GA)计算可靠性指标。数值试验表明,本方法可以快速、精确地模拟疲劳裂纹扩展寿命的极限状态函数,进而计算出失效概率和可靠性指标。同其他模拟技术相比,在精度相同的情况下,神经网络响应面法可以大大减少模拟时间。     

DECOUPLING CONTROL OF TWO MO-TORS SYSTEM BASED ON NEURAL NETWORK INVERSE SYSTEM

In accordance with the characteristics of two motors system, the united mathematic model of two-motors inverter system with v/f variable frequency speed-regulating is given. Two-motor inverter system can be decoupled by the neural network invert system, and changed into a sub-system of speed and a sub-system of tension. Multiple controllers are designed, and good results are obtained. The system has good static and dynamic performances and high anti-disturbance of load.     

基于遗传算法的模糊振动主动控制器的参数编码设计和约束处理

在采用模糊控制的压电材料智能结构中,利用遗传算法对模糊控制隶属函数进行优化时,遗传编码的设计将影响优化结果,从振动主动控制要求出发,设计了振幅输入模糊集的隶属函数参数编码,并对模糊子集约束条件进行处理。结果表明新的编码方式可以产生适合振动主动控制的隶属函数,且设计的约束修补算法简单有效。     

机器人焊缝跟踪神经网络控制的研究

研究神经网络技术在弧焊机器人焊缝跟踪过程中的应用,通过神经网络在笛卡尔空间轨迹的补偿作用,确定出基于笛卡尔空间参考轨迹控制的机器人焊缝跟踪神经网络控制器。与传统的关节计算力矩法相比,所设计的神经网络控制器具有良好的控制特性及较强的鲁棒性,焊缝跟踪精度得到了显著的提高。     

SURFACE ROUGHNESS PREDICTION USING HYBRID NEURAL NETWORKS

Surface roughness is an important outcome in the machining process and it forms a major part in the manufacturing system. Surface roughness depends on different machining parameters and its prediction and control is a challenge to the researchers. There is a need to predict surface roughness prior to machining to attain higher productivity levels. Owing to advances in computing power there is an increase in the demand for the use of intelligent techniques. Recent research is directed towards hybridization of intelligent techniques to make the best out of each technique. This article proposes the development of a novel hybrid Neural Network (NN) trained with Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) for the prediction of surface roughness. The proposed hybrid neural network is found to be competent in terms of computational speed and efficiency over the neural network model.     

SURFACE ROUGHNESS PREDICTION USING HYBRID NEURAL NETWORKS

Surface roughness is an important outcome in the machining process and it forms a major part in the manufacturing system. Surface roughness depends on different machining parameters and its prediction and control is a challenge to the researchers. There is a need to predict surface roughness prior to machining to attain higher productivity levels. Owing to advances in computing power there is an increase in the demand for the use of intelligent techniques. Recent research is directed towards hybridization of intelligent techniques to make the best out of each technique. This article proposes the development of a novel hybrid Neural Network (NN) trained with Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) for the prediction of surface roughness. The proposed hybrid neural network is found to be competent in terms of computational speed and efficiency over the neural network model.     

基于神经网络的动力学反解算法及其应用研究

论述了使用神经网络求解结构动力学反问题的基本原理和方法,并针对BP网络在使用中存在的主要问题,提出了网络快速学习算法,以及训练样本正交化处理方法和反解结果精度的自适应修正等一系列解决方法。应用上述理论进行了设备动态诊断的数值仿真和试验研究,结果令人满意。     

神经网络模型参考自适应控制及其在带材板形控制系统中的应用

通过对某轧机液压弯辊系统特性的分析,提出了神经网络模型参考自适应控制策略,兼顾了系统的动态特性和静态特性,成功地解决了未知、不确定、非线性系统的辨识与控制问题;并将其用于液压弯辊系统仿真,结果表明该系统的性能良好,将会明显提高产品质量。     

GA-BASED PID NEURAL NETWORK CONTROL FOR MAGNETIC BEARING SYSTEMS

In order to overcome the system non-linearity and uncertainty inherent in magnetic bear-ing systems,a GA(genetic algorithm)-based PID neural network controller is designed and trained to emulate the operation of a complete system (magnetic bearing,controller,and power amplifiers). The feasibility of using a neural network to control nonlinear magnetic bearing systems with un-known dynamics is demonstrated. The key concept of the control scheme is to use GA to evaluate the candidate solutions (chromosomes),increase the generalization ability of PID neural network and avoid suffering from the local minima problem in network learning due to the use of gradient descent learning method. The simulation results show that the proposed architecture provides well robust performance and better reinforcement learning capability in controlling magnetic bearing systems.     

基于遗传神经网络的机器人腕力传感器动态建模与补偿方法

介绍用于MotomamV3X机器人上的新型多维腕力传感器,比较遗传算法与人工神经网络的特点,将遗传算法的交叉和变异操作进行改进,提出一种融合改进遗传算法(Genetic algorithm, GA)的函数连接型人工神经网络(Functional link artificial neural network, FLANN),并将其用于所介绍的新型机器人腕力传感器动态建模与动态性能补偿中。介绍动态建模与动态补偿原理及改进遗传神经网络算法,给出该传感器的动态模型和动态补偿模型。该方法利用腕力传感器的动态标定数据,采用改进遗传神经网络搜索和优化模型参数,保留了遗传算法的全局搜索能力和FLANN结构简单,鲁棒性好,且具备自学习能力的特点,克服了FLANN容易陷入局部极小的缺陷,具有快的网络训练速度及高的动态建模精度。理论分析和试验结果都证实了所提出的动态建模与动态补偿方法的有效性。     

基于柔度矩阵和神经网络的结构损伤识别法

提出一种分步识别结构损伤的方法。首先利用测量模态参数建立结构柔度矩阵来确定结构损伤的大体位置，然后应用神经网络技术和结构的加速度响应对确定的损伤范围进行参数识别，根据识别的刚度值判别结构的损伤程度。通过一个8自由度结构的仿真计算表明，该方法稳定性好，计算精度高，对噪声具有很高的鲁棒性，在10％噪声情况下，应用神经网络技术能较精确地得到结构的损伤程度，显示了该方法对大型复杂结构进行损伤诊断的潜力。     

SOLVING INVERSE KINEMATICS OF REDUNDANT MANIPULATOR BASED ON NEURAL NETWORK

For the redundant manipulators, neural network is used to tackle the velocity inverse kinematics of robot manipulators. The neural networks utlized are multi-layered perceptions with a back-propagation training algorithm. The weight table is used to save the weights solving the inverse kinematics based on the different optimization performance criteria. Simulations verify the effectiveness of using neural network.