基于遗传算法的球度误差评定

首先对球度公差评定问题进行了综述.然后根据圆度公差的数学定义,引申提出球度公差最小区域条件下的评定模型,并给出遗传算法的适应度函数.随后给出算法实现中的中的关键问题.最后用实例对算法进行了检验,计算结果表明基于遗传算法的球度误差优化算法不仅符合最小区域的条件,而且易于理解和实现,能够获得全局最优解,保证了高精度、高效率.     

轧辊圆度误差智能优化评定研究

圆度误差是检验轧辊磨床轧棍加工质量的一个重要指标,圆度误差处理算法的选定对计算精度有着很大的影响,传统的评定方法均存在一定的局限性.通过分析神经网络和遗传算法相结合的现代智能优化算法在圆度误差评定中的可行性和优越性,采用常用的BP神经网络来训练轧辊圆度测量值,将权值矩阵在遗传算法中得到进化与改善,最后通过MATLAB进行仿真,并与传统的最小二乘法相比较,得出的结果显示,遗传讲经网络智能优化算法能够无限逼近真实值,提高轧辊圆度误差的计算精度.     

基于鞍点规划理论的圆度及球度误差最小区域法评定新解法

在形位公差的圆度和球度误差评定方法中,最小区域法是最符合国标中定义形位误差的方法之一。针对目前基于最小区域法建立的数学模型以及相应的求解方法存在局部收敛以及求解迂回等问题,建立了圆度和球度误差评定的鞍点规划模型,并基于鞍点规划理论的最小条件建立了鞍圆和鞍球面误差求解的新算法,通过简单几何分析和有限代数计算即可确定符合最小区域法评定原则的圆度、球度误差以及相应鞍圆、鞍球面的位置和参数。相比于传统优化算法,本文提出的方法避免了优化方法对初始值的依赖性,具有较高的求解稳定性。     

基于遗传算法的圆度公差评定法与采用最小二乘法评定的比较

根据提出的计算模型,对基于遗传算法的圆度误差评定和传统上采用最小二乘法的评定算法进行了比较分析,根据方法本身的特点和计算结果,分析了二者的不同点以及在工程应用中的适用场合.所构造的模型包括边界控制点和区域随机点,其中边界控制点模拟了由圆度误差最小区域条件所定义的最大内切圆和最小外切圆,而区域随机点模拟了实际情况下测试点的随机性和不确定性.计算结果表明基于遗传算法的圆度评定法精度较高,优于基于最小二乘法的评定算法.     

基于最小包容区域法的圆度误差评定方法

圆度误差的准确评定对轴和孔类零件的质量评判有很重要的意义。针对目前常用的圆度误差评定方法存在原理误差或模型误差的问题,提出一种完全符合最小包容区域法定义的圆度误差评定方法。该方法将区域搜索算法和圆度误差最小包容区域法评定的几何结构相结合,利用区域搜索算法确定准圆心,再根据准圆心位置和几何结构,对其进行判断和调整,最终找到准确的最小包容区域圆心,并给出最小包容区域圆度误差的精确解。构造多组仿真数据,利用此方法的评定结果与预设值相比较,证明了该方法的有效性和正确性;并利用该方法对其他文献中的数据进行评定与比较,数据处理的结果进一步显示了该方法的评定结果精确可靠,稳定性好,且效率高,可以有效地克服现有圆度误差评定方法难以找到准确最小包容区域圆心的缺陷。     

基于误差转换及图像域的圆度评定方法

针对工程应用中圆度误差评定方法存在理论深奥、计算复杂、检测效率低且不适用于大容量采样点的问题,提出了一种基于误差转换及图像域的圆度误差评定方法。该方法首先将图像域测量得到的原始圆度误差进行转换,使其满足误差评定的要求;然后以最小二乘圆为起始圆,寻求半径或半径差的“极大中的极小”,通过对最小二乘圆进行小尺度平移,并用遗传算法得到该平移规划坐标,从而获得平移后的理想圆并求得圆度误差值;最后对某型号零件进行试验,试验结果与用三坐标测量得到的结果相吻合,表明该方法可以有效、正确地进行圆度误差的评定。     

改进蜂群算法及其在圆度误差评定中的应用

针对基本人工蜂群算法(Artificial bee colony algorithm,ABC)的缺点,提出一种改进人工蜂群算法(Improved artificial bee colony algorithm,IABC),并应用于圆度误差最小区域评定中。该改进算法利用信息熵初始化种群,增强种群的多样性,并在引领蜂和跟随蜂搜索阶段,提出一种新的搜索策略,平衡算法的探索与开发能力。详细阐述IABC算法的基本原理与实现步骤,给出圆度误差满足最小包容区域条件的优化目标函数和收益度函数。通过基准测试函数验证IABC算法的有效性和准确性;通过对由三坐标机测得的多组测量数据进行圆度误差评定试验,结果表明IABC算法的评定精度优于最小二乘法、遗传算法以及粒子群算法等其他优化算法,且在求解质量和稳定性上优于ABC算法,验证了IABC算法不仅正确,而且适用于圆度误差的评定优化。     

圆度误差的神经网络评定及测量不确定度研究

为了更为准确的而又简便的评定圆度误差及其不确定度,根据最小二乘法建立圆度误差模型,基于BP神经网络算法优化目标函数的参数,阐述了BP神经网络优化算法的原理和实现方法。通过求解实例表明该方法对于圆度误差评定的非线性优化问题能得到最优解。采用传统的测量不确定度表示指南方法和蒙特卡洛方法计算得到圆度误差的不确定度,通过实例验证蒙特卡洛法的可靠性和准确性。该方法不需要求出数学模型中的传递系数,利用MATLAB操作简单,为圆度误差测量结果不确定度评定提供了更加简便的方法。     

基于海鸥算法的圆柱度误差评定方法

圆柱体零件的几何精度直接影响到机械设备的总体性能，而圆柱度误差是圆柱体零件的几何误差之一，对圆柱度误差进行精确测量和评估十分重要。针对最小区域圆柱度误差评定能否达到全局最优的问题，提出了一种基于新型元启发式海鸥算法(SOA)的圆柱度误差评定方法。首先，对圆柱体轮廓要素的提取进行了阐述，并建立了基于最小区域法的圆柱度误差评定模型；然后，介绍了海鸥优化算法中海鸥的位置更新原理、算法的优化准则和算法流程；最后，用Talyrond 585LT圆柱度仪提取了6个圆柱零件的轮廓数据，并进行了评定，通过对比不同种群数的优化结果，找到了最佳种群数，同时将其所得结果与采用遗传算法(GA)所得结果进行了对比。研究结果表明：种群数的选择对海鸥算法的优化结果影响较大，在种群数为30时能达到最优解，其精度比遗传算法高，其运行时间随着种群数的增加而增加。海鸥算法优化过程稳定，在评定最小区域圆柱度误差(MZC)方面有较好的适应性。     

基于改进果蝇优化的机械圆度误差评定研究

针对机械零件圆度误差的最小区域圆法评定问题,提出一种基于改进果蝇优化算法的评定圆度误差新方法。首先根据最小区域圆法圆度误差数学描述的优化模型设计了果蝇种群个体的编码方式以及新型味道浓度判定函数;其次在保持基本果蝇优化算法典型流程的基础上,引入增强搜索和交互学习机制来增强算法的学习效率以及保持种群的多样性;最后采用3个典型的圆度误差评定问题来验证算法性能。实例分析计算表明该方法是可行有效的,其结果具有较高精度且优于传统的评定方法,适用于求解圆度误差的评定优化问题。     

FORM ERROR EVALUATION OF CIRCLES BASED ON A FINELY-DESIGNED GENETIC ALGORITHM

A genetic algorithm(GA)-based new method is designed to evaluate the circularity error of mechanical parts. The method uses the capability of nonlinear optimization of GA to search for the optimal solution of circularity error. The finely-designed GA (FDGA) characterized dynamical bisexual recombination and Gaussian mutation. The mathematical model of the nonlinear problem is given. The implementation details in FDGA are described such as the crossover or recombination mechanism which utilized a bisexual reproduction scheme and the elitist reservation method; and the adaptive mutation which used the Gaussian probability distribution to determine the values of the offspring produced by mutation mechanism. The examples are provided to verify the designed FDGA. The computation results indicate that the FDGA works very well in the field of form error evaluation such as circularity evaluation.     

基于实数编码遗传算法的平面度评定

将基于实数编码的遗传算法应用于平面度的评定.根据尺寸和公差的数学定义,建立完全符合最小区域条件的平面度评定的数学模型,并在此基础上给出遗传算法的适应度函数.随后详细地介绍了算法的实现步骤,在基于实数编码的基础上,遗传选择操作采用一种正比选择策略--转轮法,遗传交叉操作采用简单算术交叉法,而遗传变异操作是随机均匀实数变异操作.最后对文献[5]的实验数据进行了评定,仿真结果表明该算法不仅合理,而且效率高、精度高,优于其它算法.     

Circularity error evaluation: Theory and algorithm

Many procedures for the evaluation of circularity error based on different criteria have been developed. The procedures that are based on the minimum radial separation criterion are either too complex or lack an algorithmic approach to find optimal solution. This paper presents an optimization-based technique to find the value of circularity error based on the minimum radial separation criterion. The problem is formulated as a nonlinear optimization problem. Based on the developed necessary and sufficient conditions a generalized nonlinear optimization procedure is presented. The performance of the developed procedure is analyzed for different size problems generated using a simulation program. Results indicate that the procedure is accurate and very efficient in solving large size real life problems.     

工件圆度误差测量不确定度评定

为了实现工件圆度误差的不确定度评定,对基于三坐标测量机的工件圆度轮廓数据的采样策略、圆度评定方法及不确定度评定方法进行研究。首先,根据工件圆度轮廓特征进行实验测量,获取不同工件的多个样本。接着,基于最小二乘法和微分进化优化算法对样本的圆度误差进行了误差评定。然后,在分析比较误差大小的基础上,说明了采用的采样策略和微分进化评定算法。最后,基于圆度误差评定结果运用了测量不确定度表示指南(GUM)和蒙特卡洛方法(MCM)进行不确定度评定。实验结果表明:微分进化算法与最小二乘法相比均值差最大达到1.1μm, MCM方法比GUM方法得到的标准不确定度均值小0.02μm。合理的采样点数、微分进化算法及MCM不确定度评定方法可以得到更稳定可靠、精度高的评定结果。     

基于BP网络和遗传算法的铝型材挤压模工作带优化模型

建立了改进的BP神经网络、数值仿真和遗传算法相结合的铝型材挤压模工作带长度优化模型。将型材截面划分单元 ,由正交试验法得到单元工作带长度值作为网络训练样本的输入值 ,模型目标值为变形后质点速度均方差。采用基于有限体积法的数值仿真技术获得样本目标值。模型的全局优化解由遗传算法求得。计算实例给出了模型实现和求解的完整过程     

Precision modeling of form errors for cylindricity evaluation using genetic algorithms

A heuristic approach is proposed in this paper to model form errors for cylindricity evaluation using genetic algorithms (GAs). The proposed GAs method shows good flexibility and excellent performance in evaluating the engineering surfaces via measurement data involved with randomness and uncertainty. The numerical-oriented genetic operator is used as a basic representation for error modeling in the paper. The theoretical basis for the proposed Gas-based cylindricity evaluation algorithms is first presented. The performance of the method under various combinations of parameters and the precision improvement on the evaluation of cylindricity are carefully analyzed. One numerical example is presented to illustrate the effectiveness of the proposed method and to compare the Gas-based modeling results with those obtained by the least-squares method. Numerical results indicate that the proposed GAs method does provide better accuracy on cylindricity evaluation. The method can also be extended for solving difficult form error minimization and profile evaluation problems of various geometric parts in engineering metrology.     

RESEARCH ON THE MINIMUM ZONE CYLINDRICITY EVALUATION BASED ON GENETIC ALGORITHMS

A genetic algorithm (GA)-based method is proposed to solve the nonlinear optimization problem of minimum zone cylindricity evaluation. First, the background of the problem is introduced. Then the mathematical model and the fitness function are derived from the mathematical definition of dimensioning and tolerancing principles. Thirdly with the least squares solution as the initial values, the whole implementation process of the algorithm is realized in which some key techniques, for example, variables representing, population initializing and such basic operations as selection, crossover and mutation, are discussed in detail. Finally, examples are quoted to verify the proposed algorithm. The computation results indicate that the GA-based optimization method performs well on cylindricity evaluation. The outstanding advantages conclude high accuracy, high efficiency and capabilities of solving complicated nonlinear and large space problems.     

Flatness error evaluation and verification based on new generation geometrical product specification (GPS)

New generation geometrical product specification (GPS) links the whole course of a geometrical product from the research, development, design, manufacturing and verification to its release, utilization, and maintenance. Measurement process is one of the most important part of verification/inspection in the new generation GPS. With the knowledge-intensive and globalization trend of the economy, unifying the evaluation and verification of form errors will play a vital role in international trade and technical communication. Considering the plane feature is one of the most basic geometric primitives which contribute significantly to fundamental mechanical products such as guide way of machine tool to achieve intended functionalities, the mathematical model of flatness error minimum zone solution is formulated and an improved genetic algorithm (IGA) is proposed to implement flatness error minimum zone evaluation. Then, two evaluation methods of flatness error uncertainty are proposed, which are based on the Guide to the Expression of Uncertainty in Measurement (GUM) and a Monte Carlo Method (MCM). The calculating formula and the propagation coefficients of each element and correlation coefficients based on GUM and the procedures based on MCM are developed. Finally, two examples are listed to prove the effectiveness of the proposed method. An investigation into the source and effects of different uncertainty contributors for practical measurement on CMM is carried out and the uncertainty contributors significant are analyzed for flatness error verification. Compared with conventional methods, the proposed method not only has the advantages of simple algorithm, good flexibility, more efficiency and accuracy, but also guarantees the minimum zone solution specified in the ISO/1101 standard. Furthermore, it accords with the requirement of the new generation GPS standard which the measurement uncertainty characterizing the reliability of the results is given together. And it is also extended to other form errors evaluation and verification.