76 results about "Fuzzy support vector machine" patented technology

The invention discloses a face gender recognition method based on a fuzzy support vector machine, belonging to intelligent monitoring technology in computer vision, in particular to face recognition technology. In the method, preprocessing is respectively carried out on images in a face training base and images in a face image collecting base in advance; the face characteristic extraction and the characteristic selection are carried out on the obtained two groups of preprocessed images; neural network training is then carried out to generate the fuzzy membership grade; the classifier design of the fuzzy support vector machine is carried out on the obtained fuzzy membership grade; finally, the face gender recognition is carried out and the result is output to be shown. With strong environment adaptability, the method can maintain strong robustness under different light intensities, attitudes and facial expressions. The method can also be applied to various monitoring systems and information collecting systems.

The invention discloses a method for recognizing Chinese speech emotions based on a fuzzy support vector machine. The method for recognizing the Chinese speech emotions based on the fuzzy support vector machine is used for emotion recognition of Chinese speech. The recognition process comprises two stages of rough classification and fine classification, wherein in the rough classification state, the whole situation of a sample to be recognized is extracted, emotional features are counted up, emotions are divided into three rough classifications by means of the rough classification fuzzy support vector machine. In the fine classification state, emotional discrimination in each classification is increased, the inner portion of the rough classification is divided more finely by means of a fine classification fuzzy support vector machine, and therefore every kind of emotions can be recognized. The emotional features have nothing to do with a speaker or the content of a text, training of the support vector machine is guided by fuzzy factors, PCA dimensionality reduction is conducted on fine classification features, and therefore the discrimination is increased. According to the method for recognizing the Chinese speech emotions based on the fuzzy support vector machine, Chinese speech emotion expression which has nothing to do with the speaker and the text content can be achieved by means of overall statistics of voice quality features, and complexity of the algorithm is effectively reduced and real-time performance is improved by means of classification recognition by stages. Due to the fact that the fuzzy support vector machines are applied, better recognition precision can be achieved under the condition of mixed speech emotions.

The invention discloses a behavior identification method based on a fuzzy support vector machine and adopts the fuzzy support vector machine to realize identification for various human behaviors (including normal behaviors such as standing, walking, running, going upstairs/downstairs and abnormal behaviors such as falling down); the behavior identification method is mainly applied to eliminating the influences of isolated points and noise points in the sample points to classification effects and improving behavior identification precision. The main contents for realization of the behavior identification are as follows: firstly, behavior data acquisition is realized by using a tri-axial accelerometer to obtain an X-axis acceleration, a Y-axis acceleration and a Z-axis acceleration; the mean value, the variance and the energy of the resultant acceleration as well as the correlation coefficient between any two dimensions of three-dimensional data are respectively extracted by means of resultant acceleration extracting characteristic values, and a six-dimensional characteristic vector is obtained; secondly, the degree of membership of each sampling point to the affiliated classification is calculated; thirdly, the construction of a classification model is realized by using the fuzzy support vector machine; and fourthly, the identification for human behaviors is realized at the online stage.

A Chinese medicine complexion recognition method based on color modeling comprises the following steps: first collecting multiple pieces of human face images, then dividing each human face image into multiple skin blocks in a m*n size, forming a dataset by the skin blocks (all in four classes), dividing each dataset into a training sample set and a classic sample set, calculating the color feature vector of each skin block in the training sample set and the classic sample set, and respectively calculating the class center and the biggest radius of four classes of samples in each classic sample set; and through carrying out modeling on the modeling feature vectors nu of all classes of classic complexions, calculating the maximum value of the comprehensive deformation degree alpha of all classes of classic sample models and the maximum value of the similarity degree beta of the models, calculating the relative distance between each training sample and each class center in each classic sample set, calculating the class attribution factor lambda_s of each sample in each training sample set, calculating the fuzzy affiliation degree of each sample in each training sample set with the corresponding affiliated class, training a fuzzy support vector machine, and utilizing the well trained fuzzy support vector machine to carry out Chinese medicine complexion recognition.

The invention relates to a network environment-oriented electroencephalogram identification system and a network environment-oriented electroencephalogram identification method. The identification system comprises a plurality of wearable electroencephalogram measuring devices, more than one mobile network terminal and an electroencephalogram data processing device; the electroencephalogram measuring devices are used for processing measured electroencephalograms into network electroencephalogram data and transmitting the network electroencephalogram data to the electroencephalogram data processing device for processing; the electroencephalogram data processing device is used for outputting an identification result and transmitting the identification result to a mobile network terminal for displaying. The identification method comprises the steps of measuring the electroencephalograms of a tested individual by using the electroencephalogram measuring devices and processing the electroencephalograms into the network electroencephalogram data, transmitting the electroencephalogram data to the electroencephalogram data processing device by virtue of a wireless network card of the mobile network terminal, performing segment treatment on the electroencephalograms and extracting electroencephalogram characteristic information by use of the electroencephalogram data processing device, screening electroencephalogram characteristic vector training samples based on reputation computing, and determining the input sample and the degree of membership of a fuzzy support vector machine, thereby completing the identification of the tested individual and outputting the result.

The invention discloses an emotional speaker identification method based on reliability detection of fuzzy support vector machine, which comprises the following steps of: extracting speech component characteristics and combining the speech component characteristics with corresponding weight in a universal broadcast modem (UBM) model to form background model component characteristics; taking the obtained background model component characteristics as a fuzzy membership, and establishing a fuzzy support component machine model in a general model component; carrying out a reliability detection byusing the fuzzy support vector machine model to obtain the reliability characteristics; and calculating the reliability characteristics and identifying the speaker. The method provided by the invention improves the robustness of the speaker identification system and the performance for identifying a speaker.

The invention relates to a speech emotion recognition technology, in particular to a speech emotion recognition method based on a fuzzy support vector machine. The method comprises the steps that input speech signals with emotions are pre-processed, the pre-processing comprises pre-emphasis filtering and windowing framing, the Mel frequency cepstrum coefficient (MFCC) of feature information of the processed speech signals is extracted, dimension reduction processing is carried out on the extracted MFCC by utilizing principal component analysis (KPCA), classification and recognition are carried out according to the MFCC feature information after the dimension reduction is carried out, and recognition results are output. A specific classification and recognition method is carried out by adopting an FSVM algorithm. The speech emotion recognition method based on the fuzzy support vector machine has the advantages that KPCA is adopted for carrying out dimension reduction on MFCC emotion features to reduce redundant information, the recognition effect is better than that with the MFCC features directly used, recognition efficiency of the method is higher, the effect is better, and the recognition speed is higher. The speech emotion recognition method based on the fuzzy support vector machine is especially suitable for intelligent speech emotion recognition.

The invention discloses an emotion recognition method based on an adaptive fuzzy support vector machine. The machine is used for the recognition of three target emotions of frustration, excitement andboredom of patients in the process of robot assisted rehabilitation training. The specific implementation includes the following steps of firstly, obtaining a physiological response signal of targetemotional response of the patients during the rehabilitation training; secondly, using a principal component analysis method to screen out an important characteristic set of the physiological signal indicating target emotional changes; finally, based on the characteristics that different physiological signals are easily disturbed by noise and the physiological signals among different emotions often overlap one another, the emotion recognition method based on the adaptive fuzzy support vector machine is proposed. The method not only improves the representation form of a membership function, butalso defines parameters which controls critical membership degree and an attenuation trend of the membership degree, so that the membership degrees can be adaptively adjusted according to the specific distribution characteristics of different emotional physiological signals.

The invention discloses a straw fermentation fuel ethanol production process key state variable soft measuring method based on a fuzzy support vector machine. The method is characterized in that intelligent calculation is carried out based on a hardware platform, a measuring instrument and computer system software, real-time process data is obtained by the measuring instrument, and online real-time estimation is carried out. The method comprises the steps of: firstly, analyzing the technology mechanism of the straw fermentation fuel ethanol production process, selecting suitable auxiliary variables, establishing a training sample database according to historical batch fermentation data, then mapping the training samples to a high-dimensional core space, and calculating the fuzzy membership corresponding to each sample point in the core space; utilizing the fuzzy support vector machine to carry out training on the training samples after fuzzification, and establishing a soft measuring model; and finally, according to the newest input vectors of the batch fermentation to be predicted, predicting key state variables. According to the invention, the online real-time prediction of the key state variables in the straw fermentation fuel ethanol production process is realized, and the method has great significance on the parameter detection and optimization control of the straw fermentation fuel ethanol production process.

Disclosed are a FSVM-based lysine fermentation process key state variable soft measuring method and a system; the method depends on a hardware platform, a measurement instrument, a computer system and an intelligent controller for software computation, the software obtains real-time process data for soft measurement through the measurement instrument. The method firstly analyze the process mechanism of the lysine fermentation process, selects proper auxiliary variables to create a training sample database according to the historical batch fermentation data, subsequently projects the training samples to a high dimensional nuclear space to calculate the fuzzy degree of membership corresponding to each sample point in the nuclear space; the training samples after fuzzification are trained by FSVM to build a soft measurement module; finally, the predication of key state variables is realized according to the latest input vector of the batch fermentation under predication. The invention realizes the online real-time predication of key state variable of lysine fermentation process, and has great significance on realizing the optimized control and operation of the lysine fermentation process.

The invention discloses an unmanned light boat motion control method based on a fuzzy support vector machine algorithm and belongs to the technical field of motion control. Navigation data and external environment data of an unmanned light boat are collected and then divided into training data and testing data that are subjected to preprocessing operation, and a separation threshold value for preprocessing results is searched. Whether the quantity of the training data is greater than or equal to the threshold value is determined; if yes, a particle swarm algorithm is adopted for the training data; if no, a grid search algorithm is adopted; optimal configuration parameters are obtained for verification and simplification. Preprocessed testing data is input; if the data is linearly separable, the data is input into a fuzzy support vector machine so as to obtain an optimal decision surface; if the data is not linearly separable, the data is mapped to a high dimension space, and training operation is performed to obtain the optimal decision surface. Environment disturbing force data of a boat body is divided into the training data and the testing data, output of the fuzzy support vector machine is obtained, and a motion environment is compensated with the output. Via the method, unmanned light boat motion effects can be improved, and deviation generated due to environment interference can be effectively reduced.

The invention discloses a soft measuring method and system for key variables of a lysine fermentation process based on PSO-FSVM. The method relies on a hardware platform, a measuring instrument and a computer system software used for intelligent computing. The method includes the following steps: analyzing the technological mechanism of the lysine fermentation process, selecting appropriate auxiliary variables, and establishing a training sample database according to historical pot batch data; mapping a training sample into a high dimensional kernel space, and calculating the fuzzy membership degree corresponding to each sample point in the kernel space; then conducting on-line optimization on a kernel function parameter and a penalty coefficient by a particle swarm algorithm, training the fuzzy training sample by a fuzzy support vector machine, and establishing a soft measurement model; finally, achieving prediction of the key state variables according to the latest to-be-predicted pot batch input vector. The soft measuring method and system achieve the on-line real-time prediction of the key state variables in the lysine fermentation process, and are of great significance for the parameter prediction and the optimal control of the lysine fermentation process.

The invention discloses a fault classification method based on an improved fuzzy support vector machine. The method comprises the steps that sample data of a switch cabinet are acquired, an optimization model is established for the sample data by introducing a relaxation factor, and the minimum radius of the hyper-sphere of the optimization module is calculated; the membership degree of the sample data inside and outside the boundary of the hyper-sphere is calculated by using a segmented half-time normal cloud model with the minimum radius of the hyper-sphere acting as the boundary of the hyper-sphere; the sample data of the switch cabinet, fault classification marks and the membership degree form a sample set of the fault types, and the optimal segmented function of the improved FSVM is acquired by using the sample set of the fault types; and the classifier of the improved FSVM is established by using the optimal segmented function of the improved FSVM, and the sample data pass through the classifier of the improved FSVM in turn so that the fault type of each sample data is acquired. The effect of fault classification of the switch cabinet can be enhanced by the method.

The invention discloses an urban lake and reservoir cyanobacterial bloom multivariate predication method based on a fuzzy support vector machine. The urban lake and reservoir cyanobacterial bloom multivariate predication method comprises the steps of: step 1, selecting key influencing factors in modeling of urban lake and reservoir cyanobacterial bloom multivariate predication; step 2, reconstructing a phase space of a urban lake and reservoir cyanobacterial bloom multivariate time sequence; step 3, optimizing and determining nearest neighbor points; step 4, and acquiring an urban lake and reservoir cyanobacterial bloom multivariate fuzzy support vector machine prediction model, and predicting urban lake and reservoir cyanobacterial bloom. The urban lake and reservoir cyanobacterial bloom multivariate predication method proposes the definition of similarity coefficient analysis for selecting the key influencing factors of lake and reservoir cyanobacterial bloom generation, and takes the consistency of time sequence variation trend and the similarity of time domain features into account, so as to determine the degree of similarity between the influencing factors and characterization factors, extract complete strong correlation information, reduce redundant information and improve the robustness and generalization capacity of prediction.

The invention discloses a multi-group image classification method based on characteristic expansion and fuzzy support vector machines, belonging to the field of image processing. The invention aims to solve the problems that the existing multi-group image classification method can not effectively extract the substantive characteristics of an image and the classification precision is relatively low. The method comprises the following steps of: firstly removing the wave bands which can not be used due to the serious pollution from noise and the like, and performing two-dimensional empirical mode decomposition on the remaining wave bands to obtain some two-dimensional intrinsic mode functions; organically combining the two-dimensional intrinsic mode functions, and expanding into the characteristics of multi-group images; and finally, classifying by a fuzzy support vector machine serving as a classifier. The method disclosed by the invention gives full play to the advantage that the two-dimensional empirical mode decomposition can adaptively extract the substantive characteristics of a complex image, and effectively obtains the characteristics of multi-group images; and moreover, by adopting the fuzzy support vector machine as a classifier which integrates the advantages of a support vector machine and a fuzzy function, the classification precision is improved.

The invention discloses a multi-dimensional cloud model-fuzzy support vector machine-based ecological risk evaluation method. The method comprises the following steps of: determining a plurality of evaluation factors and an evaluation standard, considering the plurality of evaluation factors as a plurality of attributes of a support vector machine sample, and dividing ecological risks into a plurality of grades according to the evaluation standard; determining model parameters according to a grading index; considering a super entropy as a product of a parameter k obeying normal distribution and an entropy; generating training samples, with a plurality of attributes, of each grade through a forward multi-dimensional normal cloud generator, and calculating certainty degrees of samples belonging to specific grades so as to generate a training sample set; carrying out model establishment and parameter optimization; selecting a Gaussian function as a kernel function; generating a pluralityof binary classifier by using a one-to-one method so as to form a multiclass classifier; carrying out parameter optimization by utilizing K-fold cross validation and a grid method so as to determine optimum parameters C and sigma; training samples by using the training sample set generated by the cloud generator and then establishing a model; and finally classifying measured data of a region by using the model, and judging an ecological risk grade of the region so as to realize comprehensive ecological risk evaluation under multiple factors.

The invention discloses a multi-classification method based on a fuzzy support vector machine. The basic idea of fuzzy support vector machines (FSVMs) comprises the steps that the fuzzy technology is applied to a support vector machine, different punishment weight coefficients are adopted on different samples, when objective functions are established, different samples have different contributions, smaller weight is given to the samples containing noise or outliers, and therefore the purpose of eliminating the influences of the noise and the outliers is achieved. Inseparable areas exist on the fuzzy support vector machine in multi-class classification problems, namely, regarding to trained classification functions, classification cannot be carried out on data to be classified. In order to prevent the inseparable areas from being generated, a fuzzy membership degree function is introduced, and the problem of the inseparable areas is eliminated to a large extent. In addition, correction is carried out through fuzzy support, and consistency of classification results of all classifiers is further guaranteed.

The invention relates to a bridge structure constant load response time domain fusion analysis method, and belongs to the field of bridge structures. The method comprises the following steps: S1, extracting and analyzing bridge monitoring data at the same temperature, and eliminating the influence of temperature effects; S2, reducing the influence of random interference utilizing a time domain averaging technique; S3, extracting a bridge constant load response characteristic quantity utilizing an autoregressive integrated moving average model; and S4, performing data fusion on the acquired bridge structure constant load response information through a fuzzy support vector machine to obtain a final TDFA(Thulium Doped Fiber Amplifier) analysis result. According to the bridge structure constant load response time domain fusion analysis method, on the premise that the influence of temperature factors is eliminated and the influence of random factors is reduced, the characteristic quantity representing the changing condition of a structure constant load response is directly extracted from complex bridge safety operation monitoring signals, the slow evolution process of variables in a whole monitoring period is deeply analyzed, and scientific reference bases are provided to the technicians for management and maintenance of in-service bridges.

The invention discloses a scene image classification method based on an interval Type-2 fuzzy support vector machine, and relates to the technical field of general image data processing or producing methods. According to the method, a novel fuzzy support vector machine design method based on an interval Type-2 fuzzy set is put forward, and a scene image is effectively classified based on three steps, namely, image classification feature extraction, feature dimension reduction and feature classification. According to the invention, uncertain information is described and processed based on an interval Type-2 fuzzy set, an interval Type-2 fuzzy support vector machine classification model is built, and the accuracy and robustness of scene image classification are improved.

The invention relates to a music genre classification method and system based on a feature weighted fuzzy support vector machine, belonging to the technical field of music content retrieval and pattern recognition. The classifier of the invention selects the fuzzy support vector machine, and can assign the corresponding membership degree according to the difference of the contribution of differentinput samples to the classification, so as to reduce the influence of the noise. The feature weights calculated by reliefF feature selection algorithm are used to determine the membership degree of fuzzy support vector machine, which takes into account the different influence of features with different weights on classification. Aiming at the inseparable point of blind area, the weighted Euclidean distance from the point to the center of each class is used to divide the multi-class probability, which accords with the essence of music diversification.

The invention discloses a robust image watermark detecting method based on fuzzy classification. The robust image watermark detecting method comprises the following steps: in a watermark-containing image correcting process, extracting an eight-circumcircle lower PHT matrix (low 7-order), obtained by circumcircle lower PHT decomposition, of a to-be-detected image, taking the eight-circumcircle lower PHT matrix as a training characteristic to combine with a fuzzy support vector machine theory, training a sample to obtain a training module for performing geometric correction. The fuzzy vector machine FSVM has very good learning capacity and the matrix obtained by PHT decomposition has high precision, so that watermark information under different attacks can be correctly extracted, and the watermark robustness is improved. Moreover, the method has the characteristics of simple calculation, no need of an initial carrier during watermark extraction, and the like, so that the copyright protection practicability of digital image works is strengthened.