Pattern Recognition Methods Research Articles

Forecasting the atmospheric air pollution caused by nitrogen dioxide in the City of Odesa

In most of Ukrainian large cities with a high concentration of population the air quality often does not meet the requirements. The air quality is also a critical component for Odesa, a recreational and resort center of Ukraine. Moreover, Odesa is a city with a million inhabitants and a large number of vehicles, freight traffic resulting from the presence of the seaport. And nitrogen dioxide is one of the main pollutants. It is emitted to the air basin mainly by motor vehicles.
 The paper presents the results of testing the certain methods of short-term forecasting of the atmospheric air nitrogen dioxide-related pollution level in Odesa. The study is based on the data taken from the materials of the National Report on the State of the Environment in Ukraine and the observations of air pollution using the network of stationary observation points.
 The obtained results of the forecast for the winter period using the UkrSIHMI's method indicated that the dynamics of changes in the actual values of the indicator Q coincides with the prognostic values of such indicator. In summer the actual value of Q in almost all cases coincides with the prognostic indicators. The accuracy of the forecast according to the UkrSIHMI's method for the winter period averaged 96.6 %, and for the summer period – 98.3 %. According to the method of pattern recognition for the winter and summer periods the percentage of accuracy of predictions was only 67 %.
 The UkrSIHMI's method proved to be more reliable than the method of pattern recognition. The low percentage of reliability of the forecasting results obtained via the method of pattern recognition may be due to modeling problems and small variance of pollution classes in the studied sample, as well as to the focus of the methodology on determining a pollution class rather than a level of pollution. The disadvantage of using the UkrSIHMI's method is the impossibility of using it in autumn and spring.
 Forecasting the air pollution is one of the main ways to solve air quality problems in both industrial and urban agglomerations. Development of an effective prognostic scheme for forecasting the air pollution in our city can prevent further air quality deterioration.

Visible and Near-Infrared Spectroscopy Combined With Bayes Classifier Based on Wavelength Model Optimization Applied to Wine Multibrand Identification.

The identification of high-quality wine brands can avoid adulteration and fraud and protect the rights and interests of producers and consumers. Since the main components of wine are roughly the same, the characteristic components that can distinguish wine brands are usually trace amounts and not unique. The conventional quantitative detection method for brand identification is complicated and difficult. The naive Bayes (NB) classifier is an algorithm based on probability distribution, which is simple and particularly suitable for multiclass discriminant analysis. However, the absorbance probability between spectral wavelengths is not necessarily strongly independent, which limits the application of Bayes method in spectral pattern recognition. This research proposed a Bayes classifier algorithm based on wavelength optimization. First, a large-scale wavelength screening for equidistant combination (EC) was performed, and then wavelength step-by-step phase-out (WSP) was carried out to reduce the correlation between wavelengths and improve the accuracy of Bayes discrimination. The proposed EC-WSP-Bayes method was applied to the 5-category discriminant analysis of wine brand identification based on visible and near-infrared (Vis-NIR) spectroscopy. Among them, four types of wine brands were collected from regular sales channels as identification brands. The fifth type of samples was composed of 21 other commercial brand wines and home-brewed wines from various sources, as the interference brand. The optimal EC-WSP-Bayes model was selected, the corresponding wavelength combination was 404, 600, 992, 2,070, 2,266, and 2,462 nm located in the visible light, shortwave NIR, and combination frequency regions. In modeling and independent validation, the total recognition accuracy rate (RARTotal) reached 98.1 and 97.6%, respectively. The technology is quick and easy, which is of great significance to regulate the alcohol market. The proposed model of less-wavelength and high-efficiency (N = 6) can provide a valuable reference for small special instruments. The proposed integrated chemometric method can reduce the correlation between wavelengths, improve the recognition accuracy, and improve the applicability of the Bayesian method.

A multiscale fuzzy dual-domain attention network for urban remote sensing image segmentation

ABSTRACT Semantic segmentation of high-resolution remote sensing images plays an important role in the remote sensing community. However, many indistinguishable objects are prevalent within urban remote sensing images, and some objects belonging to the same class are different and many objects that do not belong to the same class are similar. These tricky objects make the images exhibit low-interclass variance and high-intraclass variance, which significantly limits segmentation performance. Therefore, a fresh insight was presented to alleviate this issue by incorporating the fuzzy pattern recognition method and deep-learning method. Specifically, we proposed a multiscale fuzzy dual-domain attention network (MFDAN). In MFDAN, a two-dimensional Gaussian fuzzy learning module is proposed to eliminate those factors that influence the intraclass and interclass variance. In addition, a dual-domain attention module is proposed to derive more informative semantic representations in the channel and spatial domains, respectively. These two modules will be integrated in a multiscale perspective. Extensive experiments on the benchmark datasets illustrate qualitatively and quantitatively that the proposed MFDAN is competitive.

Numerical methodologies and tools for efficient and flexible unmixing of single-sample grain-size distributions: Application to late Quaternary aeolian sediments from the desert-loess transition zone of the Tengger Desert

Parametric curve-fitting techniques are routinely adopted to derive the components of individual grain-size distributions of sediment samples. However, due to the lack of methodologies and calculation platforms to efficiently generate reproducible unmixing solutions, and appropriate statistical techniques to analyze and visualize unmixing results characterized by complex and variable structures, the technique has seldom been extended to the systematic analysis of massive grain-size distributions in sedimentological and paleoenvironmental research. In this study, we use numerical techniques to develop a novel strategy enabling the efficient and flexible unmixing of single-sample grain-size distributions. The method was applied to a large set of grain-size distributions of late Quaternary aeolian sediments from the desert-loess transition zone around the Tengger Desert in North China. The potential structures of the pooled unmixed subpopulations were derived using a finite mixture model, taking variances into account, together with a pattern recognition (clustering) method. Five unimodal endmembers were identified by the finite mixture model and four prevalent structural patterns were recognized by the clustering method. The significances of the endmembers and structural patterns are discussed in detail. We demonstrate that the single-sample unmixing method can be used to complement the widely applied end-member modelling to reveal potential provenances of large datasets of grain-size distributions, and that the component structures of grain-size distributions may contain valuable information that can be used to constrain the associated paleoenvironmental conditions.

Analyzing Human Muscle State with Flexible Sensors

Analyzing human muscle states has attracted extensive attention. EMG (electromyography) pattern recognition methods based on these works have been proposed for many years. However, uncomfortable wearing and high prices make it inconvenient for motion tracking and muscle analysis by using robotic arms and inertial sensors in daily life. In this study, we propose to use smart clothes integrated with flexible sensors to collect arm motion data, estimate the kinematic information of continuous arm motion, and predict the EMG signal of each arm muscle. Firstly, the neural network regression model integrated with the LSTM (long short-term memory) module is used to continuously estimate the sensor resistances collected by the smart clothes and the angles collected by Kinect. Then, six types of shoulder and elbow movements’ angles and the corresponding EMG signals of 5 subjects are preprocessed and aligned. The stacked regression model based on extremely randomized trees (extra trees) is used for regression. Our experimental results show that the average estimation absolute error from the sensor resistances to the joint angle is 3.45 degrees, and the absolute percentage error from the joint angle to the EMG signal is only 1.82%.

The Detection Method of Potato Foliage Diseases in Complex Background Based on Instance Segmentation and Semantic Segmentation.

Potato early blight and late blight are devastating diseases that affect potato planting and production. Thus, precise diagnosis of the diseases is critical in treatment application and management of potato farm. However, traditional computer vision technology and pattern recognition methods have certain limitations in the detection of crop diseases. In recent years, the development of deep learning technology and convolutional neural networks has provided new solutions for the rapid and accurate detection of crop diseases. In this study, an integrated framework that combines instance segmentation model, classification model, and semantic segmentation model was devised to realize the segmentation and detection of potato foliage diseases in complex backgrounds. In the first stage, Mask R-CNN was adopted to segment potato leaves in complex backgrounds. In the second stage, VGG16, ResNet50, and InceptionV3 classification models were employed to classify potato leaves. In the third stage, UNet, PSPNet, and DeepLabV3+ semantic segmentation models were applied to divide potato leaves. Finally, the three-stage models were combined to segment and detect the potato leaf diseases. According to the experimental results, the average precision (AP) obtained by the Mask R-CNN network in the first stage was 81.87%, and the precision was 97.13%. At the same time, the accuracy of the classification model in the second stage was 95.33%. The mean intersection over union (MIoU) of the semantic segmentation model in the third stage was 89.91%, and the mean pixel accuracy (MPA) was 94.24%. In short, it not only provides a new model framework for the identification and detection of potato foliage diseases in natural environment, but also lays a theoretical basis for potato disease assessment and classification.

Tracing sources of oilfield wastewater based on excitation-emission matrix fluorescence spectroscopy coupled with chemical pattern recognition techniques

In order to prevent the illegal discharge of oilfield wastewater, this work proposed excitation-emission matrix fluorescence (EEMF) spectroscopy coupled with two kinds of chemical pattern recognition methods for tracing the sources of oilfield wastewater. The first pattern recognition method was built from the relative concentrations extracted by alternating trilinear decomposition (ATLD) based on partial least squares-discriminant analysis (PLS-DA) algorithm, and the other one was modeled based on strictly multi-way partial least squares-discriminant analysis (N-PLS-DA). Both methods showed good discrimination abilities for oilfield wastewater samples from three different sources. The total recognition rates of the training and prediction sets are 100%, the values of sensitivity and selectivity are 1. This study showed that EEMF spectroscopy combined with chemical pattern recognition techniques could be used as a potential tool for tracing the sources of oilfield wastewater.

Characterization of the volatile flavor profiles of Zhenjiang aromatic vinegar combining a novel nanocomposite colorimetric sensor array with HS-SPME-GC/MS

In this work, the feasibility of combining headspace solid phase microextraction-gas chromatography-mass spectrometry (HS-SPME-GC/MS) with colorimetric sensor array (CSA) for flavor characterization of Zhenjiang aromatic vinegar (ZAV) from different grades was evaluated. Firstly, a new type of nanocomposite CSA that combines porous nanomaterials including hollow zeolitic imidazolate framework-8 (H-ZIF-8) and Universitetet i Oslo-66-NO2 (UiO-66-NO2) with chemically responsive dyes was successfully constructed. Then, the nanocomposite CSA was applied to effectively discriminate ZAV of different grades and further quantitively predict the characteristic aroma components by using multivariate data analysis. Compared with other pattern recognition methods, support vector machine (SVM) model achieved the highest recognition rate both for training set (100%) and prediction set (94.44%). Furthermore, a good performance of quantitative prediction of characteristic aroma components of ZAV including acetic acid, total volatile acids, furfural, aldehyde ketones, ethyl acetate and esters combining CSA with partial least square (PLS) regression was achieved with all the correlation coefficients being over 0.80 for training and prediction sets. Therefore, the nanocomposite CSA combined with chemometrics could be an effective tool for the rapid and nondestructive assessment of flavor and quality of vinegar.

Automated Classification of Fish Using Machine Learning and Pattern Recognition

Automation in classification is a ubiquitous process in every modern industry. This process is especially critical in the seafood industry of raw fish, where nature produces magnitude of complex forms and patterns. This paper is presenting the various methods of pattern recognition to the fish classification problem, discusses the processing of images, decision-making methods for classification of patterns (Nearest Neighbor Algorithm and Bayesian decision theory), optimization of the classifiers for pattern recognition, ways of analysis of patterns by groups, and the basic concepts of K-means clustering algorithm.

Dynamic subwindow matching: A new similarity measure for seismic facies analysis

Dynamic subwindow matching: A new similarity measure for seismic facies analysis

Characterization of unifloral Italian (Piedmont region) honeys by headspace solid phase microextraction coupled to gas chromatography–mass spectrometry

AbstractBackgroundThe characterization of five different types of Italian unifloral honeys was performed by means of headspace solid‐phase microextraction coupled to gas chromatography–mass spectrometry. The main volatile compounds of the honeys were identified and semi‐quantified. Pattern recognition methods such as principal component analysis (PCA) and k‐nearest neighbors (k‐NN) were performed to classify honey samples with different botanical origins, and identify the most discriminating parameters.ResultsMore than 100 of compounds were unambiguously identified in the honeys sampled. Methyl salicylate was present in honeydew honeys only, and this permitted to differentiate this kind of honeys from the others. Dandelion honeys featured high levels of nitriles, while lime tree honeys displayed the presence of carvacrol and of the terpenes cis and trans‐p‐mentha‐2,8‐dien‐1‐ol. Acacia honeys did not display potential chemical markers but were characterized by low levels of benzaldehyde and high levels of cis‐linalool oxide. Chestnut and rhododendron honeys did not show a volatile profile markedly different from those of other honeys. Application of PCA and k‐NN methods did not allow a complete classification of honeys of different botanical origin on the basis of their volatile profile.ConclusionUnifloral honeys from Piedmont (Italy) displays a rich and variegate profile of volatile compounds, some of which are candidates as markers. Nonetheless, to classify honeys, further studies are necessary likely comprising a larger set of samples and the determination of different chemical parameters in conjunction with volatile compounds.

Performance Analysis and Architecture of a Clustering Hybrid Algorithm Called FA+GA-DBSCAN Using Artificial Datasets.

Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is a widely used algorithm for exploratory clustering applications. Despite the DBSCAN algorithm being considered an unsupervised pattern recognition method, it has two parameters that must be tuned prior to the clustering process in order to reduce uncertainties, the minimum number of points in a clustering segmentation MinPts, and the radii around selected points from a specific dataset Eps. This article presents the performance of a clustering hybrid algorithm for automatically grouping datasets into a two-dimensional space using the well-known algorithm DBSCAN. Here, the function nearest neighbor and a genetic algorithm were used for the automation of parameters MinPts and Eps. Furthermore, the Factor Analysis (FA) method was defined for pre-processing through a dimensionality reduction of high-dimensional datasets with dimensions greater than two. Finally, the performance of the clustering algorithm called FA+GA-DBSCAN was evaluated using artificial datasets. In addition, the precision and Entropy of the clustering hybrid algorithm were measured, which showed there was less probability of error in clustering the most condensed datasets.

LASSO-Based Pattern Recognition for Replenished Items With Graded Responses in Multidimensional Computerized Adaptive Testing.

As a branch of statistical latent variable modeling, multidimensional item response theory (MIRT) plays an important role in psychometrics. Multidimensional graded response model (MGRM) is a key model for the development of multidimensional computerized adaptive testing (MCAT) with graded-response data and multiple traits. This paper explores how to automatically identify the item-trait patterns of replenished items based on the MGRM in MCAT. The problem is solved by developing an exploratory pattern recognition method for graded-response items based on the least absolute shrinkage and selection operator (LASSO), which is named LPRM-GR and facilitates the subsequent parameter estimation of replenished items and helps maintaining the effectiveness of item replenishment in MCAT. In conjunction with the proposed approach, the regular BIC and weighted BIC are applied, respectively, to select the optimal item-trait patterns. Simulation for evaluating the LPRM-GR in pattern recognition accuracy of replenished items and the corresponding item estimation accuracy is conducted under multiple conditions across different numbers with respect to dimensionality, response-category numbers, latent trait correlation, stopping rules, and item selection criteria. Results show that the proposed method with the two types of BIC both have good performance in pattern recognition for item replenishment in the two- to four-dimensional MCAT with the MGRM, for which the weighted BIC is generally superior to the regular BIC. The proposed method has relatively high accuracy and efficiency in identifying the patterns of graded-response items, and has the advantages of easy implementation and practical feasibility.

Desorption Electrospray Ionization Mass Spectrometry Imaging Illustrates the Quality Characters of Isatidis Radix

For a long history, herbal medicines have made significant contributions to human health all around the world. However, the exploration of an effective approach to illustrate their inner quality remains a challenge. So, it is imperative to develop new methods and technologies to characterize and identify quality markers of herbal medicines. Taking Isatidis Radix, the dried root of Isatis indigotica as an example, desorption electrospray ionization (DESI), in combination with quadrupole-time-of-flight mass spectrometry (Q-TOF/MS), was applied in this work for the first time to reveal the comprehensive spatial distribution of metabolites and, further, to illustrate quality characters of this herbal medicine. After simple pretreatment, 102 metabolites including alkaloids, sulfur-containing compounds, phenylpropanoids, nucleosides, amino acids, organic acids, flavonoids, phenols, terpenes, saccharides, peptides, and sphingolipids were characterized, some of which were successfully localized and visualized in the transverse section of the root. Based on the ion images, samples with different quality characters were distinguished unambiguously by the pattern recognition method of orthogonal partial least squares discrimination analysis (OPLS-DA). Simultaneously, 11 major influencing components exerting higher ion intensities in superior samples were identified as the potential quality markers of Isatidis Radix. Desorption electrospray ionization (DESI) mass spectrometry imaging (MSI), together with chemometric analysis could not only improve the understanding of the plant biology of herbal medicines but also be beneficial in the identification of quality markers, so as to carry out better quality control of herbal medicines.

Establishment of the thin-layer chromatography-surface-enhanced Raman spectroscopy and chemometrics method for simultaneous identification of eleven illegal drugs in anti-rheumatic health food

The risk of the illegal addition of anti-inflammatory and analgesic chemical drugs in anti-rheumatic health foods should not be ignored. Market supervision and rapid on-site detection technology need to be strengthened. Thin-layer chromatography-surface-enhanced Raman spectroscopy (TLC-SERS), which has the advantages of simple operation, fast separation, and qualitative and quantitative detection, was used in this study. And these eleven chemical drugs (acetaminophen, acetylsalicylic acid, aminophenazone, dexamethasone, diclofenac sodium, hydrocortisone, indometacin, naproxen, phenylbutazone, piroxicam, prednisone 21-acetate) that may be added to anti-rheumatic health foods have been simultaneously identified by TLC-SERS combined with chemometrics method. The characteristic signals of the separated drug spots were collected by SERS, which were optimized by the gold colloidal nanoparticles' volume and integration time. Then SERS was subjected to principal component analysis (PCA) to reduce dimensionality and combined with the pattern recognition methods in chemometrics, such as PCA-Linear Discriminant Analysis (LDA), PCA-K Nearest Neighbor and PCA-Support Vector Machine, and eleven drug components were judged and identified. Moreover, the predictive performances of different models were also analyzed and compared. The results showed that the TLC plate and four organic solvents of petroleum ether, chloroform, ethyl acetate and acetic acid were selected as the developing agent. The dropping amount of gold colloidal nanoparticles and the integration time were set and optimized. The limit of detection of the simultaneous detection method of SERS was 10–100 mg/L. Furthermore, SERS was preprocessed by Gap-Segment 2nd Derivative, and then the PCA-LDA model was established, and the model's prediction accuracy can reach 100%. The method is simple, rapid, sensitive and accurate, and has low experimental instruments and equipment requirements. It is suitable for the on-site simultaneous detection of various anti-inflammatory and analgesic chemical drugs in health foods. It can also provide guarantee and support for the establishment of appropriate rapid detection methods and the development of supervision technology.

A novel way to formalize stable graph cores by using matching-graphs

The increasing amount of data available and the rate at which it is collected leads to rapid developments of systems for intelligent information processing and pattern recognition. Often the underlying data is inherently complex, making it difficult to represent it by linear, vectorial data structures. This is where graphs offer a versatile alternative for formal data representation. Actually, quite an amount of graph-based methods for pattern recognition has been proposed. A considerable part of these methods rely on graph matching. In the present paper, we propose a novel encoding of specific graph matching information. The basic idea is to formalize the stable cores of individual classes of graphs – discovered during intra-class matchings – by means of so called matching-graphs. We evaluate the benefit of these matching-graphs by researching two classification approaches that rely on this novel data structure. The first approach is a distance based classifier focusing on the matching-graphs during dissimilarity computation. For the second approach, we propose to use sets of matching-graphs to embed input graphs into a vector space. The basic idea is to produce hundreds of matching-graphs first, and then represent each graph g as a vector that shows the occurrence of, or the distance to, each matching-graph. In a thorough experimental evaluation on seven real world data sets we empirically confirm that our novel approaches are able to improve the classification accuracy of systems that rely on comparable information as well as state-of-the-art methods.

Fault Diagnosis of a Solenoid Valve Based on Multi-Feature Fusion

This paper presents a non-intrusive solenoid valve fault diagnosis method, which adopts the eigenvalue extraction scheme and combines the time-frequency analysis and time-domain analysis of the current signal through the solenoid valve fault pattern recognition method. Specifically, the research group selected the solenoid valve driving end current value and its second-order change rate curve as the characteristic curve, and then it extracted the time-domain parameters and the corresponding frequency band energy of the second-order change rate as the characteristic value. Then, the research group conducted time-frequency analysis on the basis of the current signal, created a multi-feature fusion feature vector and then input it into a multi-class support vector machine (SVM) for solenoid valve fault diagnosis. To evaluate the method, the research group developed a working solenoid valve model that integrated a magnetic circuit, circuit and spool motion, and the research group examined four typical solenoid valve states (normal, spring broken, spool stuck and spool slightly stuck), for which the research group collected the current signals during solenoid valve energization and closing. The trials involved analyzing the current characteristics of the drive end, both in simulation and experimental setups, and the research group demonstrated that the proposed drive end current detection diagnosis method can improve the accuracy of recognition by 15.22% compared with traditional neural network recognition methods. The SVM strategy exploiting a multi-feature fusion strategy can improve recognition accuracy by 8.7% and verification accuracy by 42.11% compared with SVM on the basis of the energy feature values.

Analysis of Gut Microbiome Structure Based on GMPR+Spectrum

The gut microbiome is related to many major human diseases, and it is of great significance to study the structure of the gut microbiome under different conditions. Multivariate statistics or pattern recognition methods were often used to identify different structural patterns in gut microbiome data. However, these methods have some limitations. Minimal hepatic encephalopathy (MHE) datasets were taken as an example. Due to the physical lack or insufficient sampling of the gut microbiome in the sequencing process, the microbiome data contains many zeros. Therefore, the geometric mean of pairwise ratios (GMPR) was used to normalize gut microbiome data, then Spectrum was used to analyze the structure of the gut microbiome, and lastly, the structure of core microflora was compared with Network analysis. GMPR calculates the Intraclass correlation coefficient (ICC), whose reproducibility was significantly better than other normalization methods. In addition, running-time, Normalized Mutual Information (NMI), Davies-Boulding Index (DBI), and Calinski-Harabasz index (CH) of GMPR+Spectrum were far superior to other clustering algorithms such as M3C, iClusterPlus. GMPR+Spectrum can not only perform better but also effectively identify the structural differences of intestinal microbiota in different patients and excavate the unique critical bacteria such as Akkermansia, and Lactobacillus in MHE patients, which may provide a new reference for the study of the gut microbiome in disease.

Fast Detection of Diarrhetic Shellfish Poisoning Toxins in Mussels Using NIR Spectroscopy and Improved Twin Support Vector Machines

Diarrhetic shellfish poisoning (DSP) toxins are potent marine biotoxins. It can cause a severe gastrointestinal illness by the consumption of mussels contaminated by DSP toxins. New methods for effectively and rapidly detecting DSP toxins-contaminated mussels are required. In this study, we used near-infrared (NIR) reflection spectroscopy combined with pattern recognition methods to detect DSP toxins. In the range of 950-1700 nm, the spectral data of healthy mussels and DSP toxins-contaminated mussels were acquired. To select optimal waveband subsets, a waveband selection algorithm with a Gaussian membership function based on fuzzy rough set theory was applied. Considering that detecting DSP toxins-contaminated mussels from healthy mussels was an imbalanced classification problem, an improved approach of twin support vector machines (TWSVM) was explored, which is based on a centered kernel alignment. The influences of parameters of the waveband selection algorithm and regularization hyperparameters of the improved TWSVM (ITWSVM) on the performance of models were analyzed. Compared to conventional SVM, TWSVM, and other state-of-the-art algorithms (such as multi-layer perceptron, extreme gradient boosting and adaptive boosting), our proposed model exhibited better performance in detecting DSP toxins and was little affected by the imbalance ratio. For the proposed model, the F-measure reached 0.9886, and detection accuracy reached 98.83%. We explored the physical basis for the detection model by analyzing the relationship between the occurrence of overtone and combination bands and selected wavebands. This study supports NIR spectroscopy as an innovative, rapid, and convenient analytical method to detect DSP toxins in mussels.

Machine-learning-based hybrid recognition approach for longitudinal driving behavior in noisy environment

Driving behavior recognition has attracted wide attention as it can act as an important reference input of many vehicle intelligent control systems. In this paper, a real-time recognition of driver’s longitudinal driving behavior is investigated by proposing a hybrid adaptive pattern recognition method. Firstly, a framework of integrated behavior recognition model is established consisting of two sub models to cluster and label the sample data, respectively. Secondly, a new fast high-stability clustering method is proposed to solve the problem of clustering samples with non-negligible noise. And the test results show the clustering process can be completed within a short time lower than 0.1 s with different number of cluster centers. Then, support vector machine and artificial neural network are employed and trained to construct a heuristic self-labeling approach to label the clustered sample automatically in real time with high accuracy (92.9%) under an experimental driving cycle generated from our real-vehicle test bench. Subsequently, the two established modules are integrated and offline trained by 96 thousand historical data, and employed to a 400-second-long online application under a smoothly-varying driving condition, and three further applications under extreme driving cycles with about 1000 s, respectively. Simulation results show that the proposed integrated model is capable of eliminating the interference of noise, and has a relatively stable performance of recognition for different driving cycles with different driving behaviors (92.7% of overall performance for commonly cycles, and greater than 91.2% for extreme cycles), enhancing the capacity for online application.