Fault diagnosis of rotating machinery based on multiple ANFIS combination with GAs

This paper presents a novel method for fault diagnosis based on empirical mode decomposition (EMD), an improved distance evaluation technique and the combination of multiple adaptive neuro-fuzzy inference systems (ANFISs). The method consists of three stages. First, prior to feature extraction, some preprocessing techniques, like filtration, demodulation and EMD are performed on vibration signals to acquire more fault characteristic information. Then, six feature sets, including time- and frequency-domain statistical features of both the raw and preprocessed signals, are extracted. Second, an improved distance evaluation technique is proposed, and with it, six salient feature sets are selected from the six original feature sets, respectively. Finally, the six salient feature sets are input into the multiple ANFIS combination with genetic algorithms (GAs) to identify different abnormal cases. The proposed method is applied to the fault diagnosis of rolling element bearings, and testing results show that the multiple ANFIS combination can reliably recognise different fault categories and severities, which has a better classification performance compared to the individual classifiers based on ANFIS. Moreover, the effectiveness of the proposed feature selection method based on the improved distance evaluation technique is also demonstrated by the testing results.     

Artificial evolutionary approaches to produce smoother surface in magnetic abrasive finishing of hardened AISI 52100 steel

In this work, two models of feed forward back-propagation neural network (FFBP-NN) and adaptive neuro-fuzzy inference system (ANFIS) have been developed to predict the performance of magnetic abrasive finishing process, based on experimental data of literature [7]. Input parameters of process are electromagnet’s voltage, mesh number of abrasive particles, poles rotational speed and weight percent of abrasive particles, and also the output is percentage of surface roughness variation. In order to select the best model, a comparison between developed models has been done based on their mean absolute error (MAE) and root mean square error (RMSE). Moreover, optimization methods based on simulated annealing (SA) and particle swarm optimization (PSO) algorithms were used to maximize the percent of surface roughness variation and select the optimal process parameters. Results indicated that the models based on artificial intelligence predict much more precise values with respect to predictive regression model developed in main literature [7]. Also, the ANFIS model had a lowest value of MAE and RMSE with respect to others. So it was used as an objective function to maximize the surface roughness variation by using SA and PSO. Comparison between the obtained optimal solutions and analysis of results in main literature indicated that SA and PSO could find the optimal answers logically and precisely.     

基于ANFIS的温度传感器非线性校正方法

介绍了用神经网络进行传感器非线性误差校正的原理与方法,分析了自适应神经模糊推理系统(ANFIS)的基本原理。通过模糊聚类和混合学习算法,ANFIS可以逼近高阶输入输出非线性系统,将该算法用于两个典型非线性系统建模,均能获得满意结果。之后,将ANFIS算法用于温度传感器非线性校正中,试验结果表明该方法与基于CMAC网络和BP网络的校正方法相比,校正的精度高于以上两种校正方法。     

声表面波气体传感器发展概况

声表面波气体传感器包括采用敏感膜和结合气相色谱两种方式.比较而言,采用敏感膜的声表面波气体传感器体积小、功耗低,但可检测的气体种类少、灵敏度低、存在交叉干扰问题;声表面波与气相色谱联用的气体分析仪灵敏度高、可检测气体种类多、很好地解决交叉干扰问题,特别适合于复杂大气背景条件下的气体成分分析.文中介绍了两类声表面波气体传感器的发展概况.     

Application of ANFIS to predict crop yield based on different energy inputs

In this paper, adaptive neuro-fuzzy inference system (ANFIS) was used to predict the grain yield of irrigated wheat in Abyek town of Ghazvin province, Iran. Due to large number of inputs (eight inputs) for ANFIS, the input vector was clustered into two groups and two networks were trained. Inputs for ANFIS 1 were diesel fuel, fertilizer and electricity energies and for ANFIS 2 were human labor, machinery, chemicals, water for irrigation and seed energies. The RMSE and R² values were found 0.013 and 0.996 for ANFIS 1 and 0.018 and 0.992 for ANFIS 2, respectively. These results showed that ANFIS 1 and ANFIS 2 could well predict the yield. Finally, the predicted values of the two networks were used as inputs to the third ANFIS. The results indicated that the energy inputs in ANFIS 1 have a greater impact on the final yield production than other energy inputs. Also, the RMSE and R² values for ANFIS 3 were 0.013 and 0.996, respectively. These results showed that ANFIS 1 and the combined network (ANFIS 3) could both predict the grain yield with good accuracy.     

Parylene增强型声表面波传感器及其温度响应

针对声表面波(SAW)传感器对品质因数、寿命和成本的要求,研制了Parylene增强型SAW传感器。根据金属剥离工艺要求,利用LOR剥离胶和AZ5214光刻胶双层胶旋涂工艺制作了梯形结构;在传统光学光刻条件下制作了2μm的超细叉指电极。传感器制作过程利用了MEMS工艺,不仅实现了传感器的微型化,还可以批量化生产,得到的以石英为基底的传感器谐振频率达到249.077 953 MHz。最后在传感器的表面镀制Parylene聚合物薄膜以提高基底温度灵敏度。实验对比了未增强型(未镀Parylene)和增强型SAW传感器(镀Parylene)的温度灵敏度。结果显示:未增强型SAW传感器温度灵敏度为2.048kHz/℃,Parylene增强型SAW传感器温度灵敏度为2.855kHz/℃,比前者提高了0.807kHz/℃,且镀Parylene之后谐振频率变化量与温度具有较好的线性度,线性相关系数达到0.996 15。实验证明,Parlene增强型SAW传感器的性能优于未增强的SAW传感器。     

Adaptive-Network-Based Fuzzy Inference System Application to Estimate the Flashover Voltage on Insulator

Abstract

This paper presents application of adaptive network based fuzzy inference system (ANFIS) to estimate critical flashover voltage on polluted insulators. Diameter, height, creepage distance, form factor and equivalent salt deposit density were used as input variables for ANFIS, and critical flashover voltage was estimated. In order to train the network and to test its performance, the data sets are derived from experimental results obtained from the literature and a mathematical model. Obtained results were given in both tabulated and graphical form for various cases studies, separately. Satisfactory and more accurate results obtained by using ANFIS to estimate the critical flashover voltage for the considered conditions compared with the previous works. Both test and validation stages were explained in detail and it is observed that estimated results rather close to experimental results.     

Navigation of autonomous mobile robot using adaptive network based fuzzy inference system

Intelligent soft computing techniques such as fuzzy inference system (FIS), artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) are proven to be efficient and suitable when applied to a variety of engineering systems. The hallmark of this paper investigates the application of an adaptive neuro-fuzzy inference system (ANFIS) to path generation and obstacle avoidance for an autonomous mobile robot in a real world environment. ANFIS has also taken the advantages of both learning capability of artificial neural network and reasoning ability of fuzzy inference system. In this present design model different sensor based information such as front obstacle distance (FOD), right obstacle distance (ROD), left obstacle distance (LOD) and target angle (TA) are given input to the adaptive fuzzy controller and output from the controller is steering angle (SA) for mobile robot. Using ANFIS tool box, the obtained mean of squared error (MSE) for training data set in the current paper is 0.031. The real time experimental results also verified with simulation results, showing that ANFIS consistently perform better results to navigate the mobile robot safely in a terrain populated by variety obstacles.     

用于图像检索的三种分类器方法及其性能评价

为了研究基于不同分类器的基于内容图像检索(CBIR)方法检索结果之间的关系,针对3种基于不同分类器的CBIR方法—基于解析特征相似性的k近邻方法、基于学习特征相似性的BP神经网络方法和基于信息论的互信息方法,分析研究了它们各自的检索性能以及它们之间检索结果的相关度和权重相关度(相关度描述不同CBIR方法检索到相同ROI占返回ROI总数中的比例信息,权重相关度则描述这些相同的ROI在各自检索结果中的不同排序位置信息)。实验结果表明,K-NN,BP-ANN和MI之间检索结果相关度较差,当返回15个ROI时,平均查准率分别为72.6%7,0.7%和68.9%,K-NN与MI,K-NN与BP-ANN以及MI与BP-ANN之间检索结果相关度分别为7.09%,9.60%和14.37%,权重相关度分别为0.011,0.023和0.039。这表明,由于基于不同分类器,不同CBIR方法可能会检索到视觉上和排列顺序上非常"不同的"相似图像。     

Management of the wind speed data using adaptive neuro-fuzzy methodology

In this paper, the accuracy of the Weibull model of wind speed is evaluated using an adaptive neuro-fuzzy inference system (ANFIS) based on wind data. The wind data comprises of wind speed measurements in the city of Nis in Serbia at different heights of 10 m, 30 m and 40 m for duration of one year. The ANFIS results are compared with the experimental results and Weibull model using root-mean-square error (RMSE), coefficient of determination, and Pearson coefficient. The effectiveness of the proposed unified strategy is verified based on the simulation results.     

基于自适应模糊神经网络的噪声抵消器

讨论了基于自适应模糊神经网络的噪声抵消器的设计方法。自适应模糊神经网络系统具有非线性映射和自学习能力，能够用于噪声信号的非线性建模。它不仅能够获取信号的最佳估计，并互能够克服信号处理中存在的模型和噪声的不确定性、不完备性。该法设计的滤波器效果良好，并可以用于多路信号和复杂信号的噪声消除。     

交通环境振动测试数据中暗振动去除的ANFIS法

轨道交通引起的环境振动测试数据中混杂着暗振动的成分。提出了一种去除暗振动的自适应神经模糊推理系统(adaptive neuro-fuzzy inference system,简称ANFIS)法,阐述了其基本原理,给出了该法的具体实现步骤。通过一条列车引起的地面振动加速度时程与一条暗振动加速度时程叠加得到现场实测振动加速度时程,采用提出的ANFIS法及其他几种已有方法对该算例进行了去除暗振动的计算,并进行了对比分析。几种方法计算的时程均方根误差分别为:谱幅值修正法0.414mm/s~2,自功率谱法0.363mm/s~2,自互功率谱法0.261mm/s~2,ANFIS法0.074mm/s~2,可见,ANFIS法均方根误差最小;几种方法计算的加权振级VLz分别为:振动级修正法63.842dB,谱幅值修正法62.894dB,自功率谱法63.859dB,自互功率谱法63.802dB,ANFIS法63.805dB,ANFIS法计算结果与真实交通振动值63.815dB最接近。结果表明,在时程、傅里叶谱、功率谱密度及振动级的计算上,ANFIS法计算结果都与真实交通振动值非常接近,产生的误差比其他已有方法更小。     

基于声表面波谐振器和延迟线的编码型无源无线阵列传感器

提出了一种新型延迟编码式谐振声表面波无源无线传感系统。该传感单元由单端口谐振器和延迟线组成，激励信号采用间歇正弦脉冲串信号。响应为一变化的振荡波形，该信号的频率与声表面波器件固有频率相等。该传感阵列系统利用不同延迟线构成编码器不仅有谐振式无源无线传感器距离远的优点，而且，还具有延迟型大规模编码的优势。该方法提高了传感系统遥感测量的距离、灵敏度和信噪化，也为该声表面波阵列传感器的广泛应用提供了保证。     

Fault diagnosis of the polypropylene production process (UNIPOL PP) using ANFIS

The performance of a chemical process plant can gradually degrade due to deterioration of the process equipment and unpermitted deviation of the characteristic variables of the system. Hence, advanced supervision is required for early detection, isolation and correction of abnormal conditions. This work presents the use of an adaptive neuro–fuzzy inference system (ANFIS) for online fault diagnosis of a gas-phase polypropylene production process with emphasis on fast and accurate diagnosis, multiple fault identification and adaptability. The most influential inputs are selected from the raw measured data sets and fed to multiple ANFIS classifiers to identify faults occurring in the process, eliminating the requirement of a detailed process model. Simulation results illustrated that the proposed method effectively diagnosed different fault types and severities, and that it has a better performance compared to a conventional multivariate statistical approach based on principal component analysis (PCA). The proposed method is shown to be simple to apply, robust to measurement noise and able to rapidly discriminate between multiple faults occurring simultaneously. This method is applicable for plant-wide monitoring and can serve as an early warning system to identify process upsets that could threaten the process operation ahead of time.     

基于改进自适应模糊推理系统的YG3硬质合金精密外圆磨削表面质量预测

为提高硬质合金材料精密外圆磨削的表面完整性和加工质量,研究其表面质量的预测技术,建立了基于自适应模糊推理系统(ANFIS)的YG3硬质合金精密外圆磨削表面粗糙度预测模型,并引入混合田口遗传算法(HTGA)对预测模型进行了改进。采用工艺试验中所用的磨削参数及相应条件下测得的表面粗糙度数据作为训练样本和测试样本,通过对BP神经网络模型、传统ANFIS预测模型及改进ANFIS预测模型的预测结果进行对比分析,对三种模型的有效性和预测精度进行了验证。结果表明,所提出的改进ANFIS预测模型从预测值相对误差Er的分布及均方根相对误差EMSRE的大小来看,均优于其他两种预测模型,预测精度较高,是一种有效的表面质量预测方法。      

Using adaptive neuro-fuzzy inference system (ANFIS) for proton exchange membrane fuel cell (PEMFC) performance modeling

In this paper, an adaptive neuro-fuzzy inference system (ANFIS) is used for modeling proton exchange membrane fuel cell (PEMFC) performance using some numerically investigated and compared with those to experimental results for training and test data. In this way, current density I (A/cm²) is modeled to the variation of pressure at the cathode side P_C (atm), voltage V (V), membrane thickness (mm), Anode transfer coefficient α_an, relative humidity of inlet fuel RH_a and relative humidity of inlet air RH_c which are defined as input (design) variables. Then, we divided these data into train and test sections to do modeling. We instructed ANFIS network by 80% of numerical-validated data. 20% of primary data which had been considered for testing the appropriateness of the models was entered ANFIS network models and results were compared by three statistical criterions. Considering the results, it is obvious that our proposed modeling by ANFIS is efficient and valid and it can be expanded for more general states.     

基于混合智能新模型的故障诊断

为了解决机械设备中早期故障和复合故障识别的难题,提高故障诊断的准确率,利用经验模式分解(Empirical mode decomposition,EMD)、改进的距离评估技术、自适应神经模糊推理系统(Adaptive neuro-fuzzy inference system, ANFIS)和遗传算法(Genetic algorithm, GA)等技术,提出一种综合运用多征兆域特征集和多个分类器的混合智能诊断模型。该模型在特征提取之前,利用滤波、EMD、解调等预处理技术挖掘潜藏在动态信号中的故障信息;从原始振动信号和预处理信号中,分别提取从不同侧面表征设备运行状态的时域和频域统计特征,得到6个特征集。采用提出的一种改进的距离评估技术选择特征,从6个原始特征集中相应地筛选出6个敏感特征集。将6个敏感特征集输入到基于GA组合的多个ANFIS分类器以得到最终的诊断结果。该模型在电力机车轮对轴承的故障诊断中实现了轴承不同故障类型、不同故障程度,以及复合故障的可靠识别,获得了满意的诊断结果。应用结果也验证了基于改进的距离评估技术的特征选择方法的有效性。     

磁梯度张量系统传感器阵列的快速旋转校准

为有效消除磁梯度张量系统传感器阵列间非对准误差和传感器系统误差对测量精度造成的影响,提出了一种只需绕系统任意轴旋转一周便可理论上实现所有磁传感器与参考平台正交系间精确校准方法。利用两组无数学简化的非线性转换构建传感器系统误差线性校正模型,仅需同一旋转周期的10组测量数据便能得到参考平台与各传感器的理想正交输出。通过构建磁传感器三轴横倾、俯仰、方位转换的旋转矩阵,得到传感器空间任意姿态的非对准误差校正模型并对旋转角进行最小二乘估计,仅需同一旋转周期的3组测量数据便能对准张量系统。仿真和实测结果表明:在理想情况下仿真参数估计准确率接近100%,实验校正后各传感器输出具有较高重合与同轴性,张量分量RMSE(均方根误差)小于30nT/m。能以较简单步骤和较少采样数据高效提高差分法磁梯度张量系统测量精度。     

Feature selection for automatic visual inspection of solder joints

In this paper, three successive feature reduction methods are employed to select good features for the automatic visual inspection of solder joints. This reduction strategy includes (1) a stability test to remove the features with unstable performance, (2) a separability examination to select the features with good classification capabilities, (3) a correlation analysis to delete the redundant features. Three sets of features are implemented in this feature reduction work: (1) a circular sub-area feature set is related to the intensity conditions within distinct areas in the joint image, (2) a moment of inertia feature set is based upon the intensity of pixels and their relative position in the image plane, (3) a surface curvature feature set analyses the three-dimensional joint topology. Initially 50 features are formulated based on the above strategy. The reduction technique deletes 39 features from this set because of instability, poor performance, and high correlation with other features. Finally, the remaining 11 feature are tested and shown to be superior to state-of-the-art identification methods.     

Optimization of correlated multiple responses of ultrasonic machining (USM) process

Ultrasonic machining (USM) process has several important performance measures (responses), some of which are correlated. For example, material removal rate and tool wear rate are highly correlated. Although in the recent past several methods have been proposed in the literature to resolve the multi-response optimization problems, only a few of them take care of the possible correlation between the responses. All these methods primarily make use of principal component analysis (PCA) to consider the possible correlation between the responses. Process engineers may face the difficulty of selecting the appropriate method because the relative optimization performances of these methods are unknown. In this paper, two sets of past experimental data on USM process are analysed using three methods dealing with the multiple correlated responses, and the optimization performances of these three methods are subsequently compared. It is observed that both the weighted principal component (WPC) and PCA-based TOPSIS methods result in a better optimization performance than the PCA-based grey relational analysis method. However, the WPC method is preferable because of its simpler computational procedure.