28 results about "Relevance vector machine" patented technology

The invention discloses a sliding bearing friction wear prediction method based on fluid mechanics, and relates to the technical field of mechanical state monitoring, and the method comprises the steps: collecting single-point temperature, local pressure, vibration time domain signals and a bearing pedestal inclination angle through a sensor, and generating sensor data; performing field reconstruction based on a fluid mechanics conservation equation on the sensor data to obtain multi-field data; performing spatial alignment on the multi-field data, inputting the multi-field data into a long short-term memory (LSTM) network, and generating wear state characteristics; training a sparse correlation vector machine regression model RVM by using the historical wear data, establishing a nonlinear mapping relationship between the wear state characteristics and the wear depth, taking the wear state characteristics as the input of the vector machine regression model RVM, and outputting the predicted wear depth; sensor data and the predicted wear depth are fused in real time through Kalman filtering, and when prediction deviation exceeds a covariance threshold value, vector machine regression model RVM parameters are updated.

The invention discloses a power distribution network intelligent early warning and section positioning method based on transient recording data, and the method comprises the steps: carrying out the fault type recognition through the comprehensive analysis of the transient recording data, combining with an advanced algorithm, such as a relevance vector machine (RVM), carrying out the line abnormality recognition through a deep neural network, and accurately extracting the feature information of fault data, the limitation of an existing power distribution automation master station fault studying and judging method is effectively overcome, the problem of inaccurate fault studying and judging caused by the data problem of the intelligent terminal is avoided, the accuracy of fault studying and judging of the power distribution automation master station is integrally improved to a relatively high level, a reliable basis can be provided for fault processing, and the working efficiency is improved. Multi-source data such as a wave recording file, a topological structure and remote signaling and telemetering are fully fused, the operation state of the power distribution network is analyzed from multiple angles, compared with an analysis method of a single data source, the actual situation of the power distribution network can be reflected more comprehensively and accurately, and the reliability and accuracy of fault diagnosis and positioning are improved.

The invention provides an urban lifeline pipeline corrosion risk early warning method based on digital twinning, and the method comprises the following steps: S1, obtaining pipeline internal operation parameters, pipeline external environment parameters and external corrosion rate data according to a monitoring control system and detection equipment; s2, constructing an internal corrosion rate prediction model based on digital object fusion and a physical guidance neural network; s3, constructing an external corrosion rate prediction model based on a particle swarm optimization algorithm and a relevance vector machine; s4, performing completion and time registration on the external corrosion rate data of the pipeline according to a spline interpolation method; s5, constructing a pipeline corrosion risk grading early warning method; and S6, according to the steps S1 to S5, establishing an underground pipeline full-life service cycle corrosion risk early warning system based on digital twinning. According to the method, real-time corrosion risk grading early warning and residual life prediction are achieved for the multiphase flow underground pipeline in the full-life service cycle, and detection, maintenance and maintenance work of the multiphase flow underground pipeline under the corrosion risk is guided.

The invention discloses a wind power prediction method and a related device, and relates to the technical field of data prediction, and the wind power prediction method comprises the steps: obtaining historical wind power data; performing empirical mode decomposition on the historical wind power data to obtain a plurality of target components, the plurality of target components including a plurality of intrinsic mode function (IMF) components and a residual component; obtaining target component features corresponding to the plurality of target components; using a power prediction model based on a relevance vector machine obtained by pre-training, and taking the target component characteristics corresponding to the plurality of target components as a basis to obtain prediction values corresponding to the plurality of target components at the prediction time; and adding the predicted values respectively corresponding to the plurality of target components at the same moment to obtain a wind power data predicted value of the predicted time. The wind power prediction method disclosed by the invention has relatively high prediction precision and calculation efficiency, and also has relatively high stability and generalization ability.

The invention discloses a low-voltage user electricity larceny identification system, which comprises the following steps: S1, collecting electricity utilization data of a low-voltage user, and extracting a feature vector from the electricity utilization data; s2, using a relevance vector machine (RVM) to carry out feature selection on the power utilization features of the power consumers to obtain an optimal feature set; s3, performing clustering analysis on the optimal feature set by using an improved K-means clustering algorithm based on information gain, automatically identifying a clustering number and an initial clustering center, and performing clustering iteration to realize abnormal power consumption detection; and S4, identifying abnormal electricity consumption behaviors according to a clustering result, and generating a report containing abnormal electricity consumption behavior information and an electricity stealing suspected user list. The invention further discloses a corresponding system. According to the invention, clustering analysis is carried out on the optimal feature set through the improved K-means clustering algorithm based on information gain, abnormal electricity consumption detection is realized, clustering performance can be improved, and electricity stealing identification efficiency and accuracy are improved.

The invention discloses an online ship motion non-parametric modeling method suitable for multiple ships and multiple working conditions. The method comprises the steps that S1, a four-degree-of-freedom ship motion data set is acquired; s2, establishing a four-degree-of-freedom ship motion model, and establishing a non-parameter identification model based on the four-degree-of-freedom ship motion model; s3, establishing a hybrid kernel correlation vector machine; and S4, carrying out adaptive training on the nonparametric identification model to obtain a trained nonparametric identification model so as to complete an online ship motion nonparametric modeling process. The invention provides a self-adaptive nonparametric identification model establishment method suitable for various types of ships, the limitation that modeling is only carried out for a single ship type in previous research is made up, and the effectiveness and wide application prospect of the model under different ships and different working conditions are shown. And in ship motion modeling, consideration on the influence of rolling is added, and the defects of a traditional planar motion model are overcome, so that key factors of ship motion are reflected more comprehensively.

The application relates to a rotating machine residual life prediction method and system, and belongs to the field of mechanical life prediction. The method comprises the following steps: collecting signals capable of reflecting mechanical life of a rotating machine to be measured at least at two positions, extracting one-dimensional time characteristic data from the signals capable of reflecting mechanical life by using principal component analysis pooling, training a related vector machine by using the one-dimensional time characteristic data, so that the full life cycle of the rotating machine is divided into a period without obvious failure trend and a failure tendency period; training a deep separable convolution gate recurrent unit network by using data of the failure tendency period, and predicting the residual life of the rotating machine in the failure tendency period. The application can improve the prediction accuracy when the residual service life of the rotating machine is predicted.

The invention discloses a bearing residual service life prediction method based on a relevance vector machine and an NNHTS model, and the method comprises the steps: firstly extracting the time domain features of a bearing vibration signal, and constructing a comprehensive health index through filtering noise reduction and principal component analysis; secondly, optimizing variational mode decomposition parameters by using a genetic algorithm, and decomposing the health index sequence into high-frequency and low-frequency components; then, training and predicting the high-frequency component and the low-frequency component by adopting an NNHTS model and a correlation vector machine respectively; and finally, superposing the predicted values of the components, and reconstructing a health index decline curve to predict the residual life. According to the method, through decomposition, hierarchical modeling and integration strategies, the fitting ability of deep learning to complex fluctuation and the sparse probability modeling advantage of a relevance vector machine to trend are combined, and the prediction precision and working condition adaptability are remarkably improved.

The invention discloses a rolling bearing reliability evaluation and prediction method based on hybrid noise reduction and MKRVM, and the method comprises the steps: obtaining the vibration signal data of the whole life of a rolling bearing; carrying out mixed noise reduction on the vibration signal by using maximum kurtosis deconvolution improved by an intrinsic time scale decomposition algorithm ITD and an arithmetic optimization algorithm to obtain a vibration signal subjected to secondary noise reduction; respectively carrying out feature extraction on the vibration signal subjected to secondary noise reduction by using three feature extraction methods, namely a time domain feature extraction method, a frequency domain feature extraction method and a time-frequency domain feature extraction method; forming a feature vector set by the respectively extracted features; a stacked auto-encoder dimension reduction method and a logistic regression reliability evaluation model are adopted to obtain a reliability curve for evaluating the current bearing; inputting the feature vector set into a bearing reliability prediction model for constructing a hybrid kernel relevance vector machine, and predicting bearing degradation features to obtain predicted degradation features; and the predicted degradation characteristics are embedded into the logistic regression reliability evaluation model to obtain a bearing reliability prediction curve with higher precision.

The present invention provides a DVL velocity measurement model training method based on artificial bee colony optimization relevance vector machine, comprising: building an AUV test platform, obtaining three-axis gyroscope and three-axis accelerometer data of the inertial navigation system, three-axis DVL velocity information, latitude and longitude output by GNSS, and depth training samples provided by a depth meter; using the heading, pitch, roll, heading change rate and three-axis DVL velocity information output by the inertial navigation system as an RVM model input set, and using the three-axis inertial navigation system velocity information as an RVM model output set; initializing RVM kernel function parameters, mapping standardized data to feature space; calculating the mean and variance of the posterior distribution; updating and calculating hyperparameters and noise variance to obtain a sparse model; using an artificial bee colony algorithm to optimize the hybrid kernel function parameters in each RVM, wherein the GNSS and depth meter are used to constrain the objective function in the artificial bee colony algorithm. This method reduces the positioning error of SINS / DVL integrated navigation from 3‰ of the range to 1.5‰ of the range.

The present application provides an array layout method and device based on the joint sparse recovery technology. The method includes: determining the array aperture size, the number of dictionary grid points, the total number of reference direction diagrams, and the number of Hermitian reference direction diagrams according to the reference multi-direction diagram for array optimization; solving the joint sparse problem to obtain the measurement vector and the observation matrix of the sparse matching task; estimating the sparse weight vector by using the relevance vector machine algorithm; and assigning weights to individual patterns to obtain the excitation vector of the direction diagram. The advantages of the present application are as follows: By using the method of the present application, only by changing the element excitation, it is possible to reconstruct the same multi-direction diagram as the uniform array with fewer elements, that is, the same element positions are shared among different direction diagrams; The array layout method based on the joint sparse recovery technology not only meets the constraints on the expected main lobe width and sidelobe level, but also matches the entire reference direction diagram with precise details.

This invention discloses an online prediction method for battery SOC based on an incremental correlation vector machine (RVM) strategy using a multi-kernel ensemble approach. The method includes the following steps: Step 1, data preprocessing; Step 2, training set sampling; Step 3, kernel function selection; Step 4, model training; Step 5, model validation; Step 6, adaptive kernel parameters; Step 7, RVM model ensemble; Step 8, model prediction; Step 9, incremental learning strategy; and Step 10, online incremental prediction. From a practical perspective, this invention addresses the complexities and diverse needs of various applications. Drawing on the ideas of incremental learning and ensemble learning, it generates highly differentiated RVM individual learning models containing multiple kernel functions through dual perturbation of training samples and kernel functions. Combined with a novel incremental ensemble strategy, it avoids the problem of model overlearning, improves the model's generalization ability and robustness, and expands its application scope.

The invention discloses an electrocardiosignal anomaly detection method adopting a relevance vector machine. The electrocardiosignal anomaly detection method comprises the steps that a relevance vector machine model is built; the method comprises the following steps of: acquiring individual clinical electrocardiosignal data, and obtaining an implicit function between an electrocardiosignal vector and a state target value by utilizing an association vector machine model: after determining a weight coefficient vector, acquiring individual daily electrocardiosignal data, constructing an electrocardiosignal vector initial value, substituting the electrocardiosignal vector initial value into the implicit function, and solving a corresponding state target value initial value and a probability thereof; and judging whether the initial electrocardio state of the relative individual is abnormal or not and the credibility. Therefore, according to the electrocardiosignal anomaly detection method, in the Bayesian framework, through maximum likelihood estimation, subtle anomalies of the electrocardiosignals after the individual suffers from the heart disease are identified, the generalization ability is quite high, the method is not sensitive to noise, a probabilistic evaluation result can be given, credibility information is contained, and the method is easy to accept.

The application provides a city lifeline pipeline corrosion risk early warning method based on digital twinning, which comprises the following steps: S1, obtaining pipeline internal operation parameters, pipeline external environmental parameters and external corrosion rate data according to a monitoring control system and detection equipment; S2, constructing an internal corrosion rate prediction model based on digital-physical fusion and a physically guided neural network; S3, constructing an external corrosion rate prediction model based on a particle swarm optimization algorithm and a relevance vector machine; S4, complementing and time registering the pipeline external corrosion rate data according to a spline interpolation method; S5, constructing a pipeline corrosion risk grading early warning method; and S6, establishing a digital-twinning-based underground pipeline full-life service cycle corrosion risk early warning system according to steps S1 to S5. The application realizes real-time corrosion risk grading early warning and residual life prediction of multiphase flow underground pipelines in the full-life service cycle, and guides the detection, repair and maintenance work of the multiphase flow underground pipelines under the corrosion risk.

The invention belongs to the field of power system relay protection, and particularly relates to a real-time fault detection positioning method based on double-end synchronous high-frequency sampling and multi-dimensional feature fusion, which extracts transient traveling wave features through dynamic window function time-frequency analysis, intelligently identifies fault types in combination with a relevance vector machine, and improves fault detection accuracy. A three-dimensional space parameter correction factor is used for dynamically compensating an environment error, and finally millisecond-level accurate positioning of a complex fault is realized through a hybrid positioning equation and a three-level verification system, so that the safety protection capability of a power grid is remarkably improved.

The invention provides a titanium alloy milling surface roughness prediction method based on Fe3O4 nano-particle reinforced cutting fluid. The titanium alloy milling surface roughness prediction method comprises the following steps: preparing the cutting fluid containing Fe3O4 nano-particles; establishing a milling force, milling noise and surface roughness acquisition system, and carrying out TC4 titanium alloy wet milling tests under different process parameters; drawing a change curve of milling force, noise and surface roughness test data, and comparing the performance difference before and after addition of the Fe3O4 nanoparticles; and constructing a prediction model based on a convolutional neural network and a relevance vector machine to realize accurate prediction of the surface roughness. The cutting fluid has the advantages that the Fe3O4 nanoparticles are added into the cutting fluid, so that the milling force and noise are remarkably reduced, the surface roughness is improved, and intelligent prediction of the surface quality can be effectively realized by combining experimental testing with a machine learning model.

The invention provides an electric vehicle ownership prediction method and system, and belongs to the technical field of electric vehicle ownership prediction. The method comprises the following steps: acquiring historical time sequence data of a multi-source input variable related to the inventory of the electric vehicle, and preprocessing and standardizing the historical time sequence data; decomposing the data into a trend term, a season term and a residual term by adopting a time sequence decomposition algorithm; screening out an optimal feature subset for each component by using an information measurement algorithm; constructing a combined kernel function formed by linearly combining a plurality of basic kernel functions; optimizing and solving an optimal kernel weight combination by adopting an improved particle swarm optimization algorithm; and training a relevance vector machine prediction model by using the optimal kernel weight combination and the optimal feature subset, and integrating prediction results of all components to obtain a final electric vehicle inventory prediction value. According to the method, multiple influence factors are comprehensively considered, feature selection and kernel function combination are optimized, and the accuracy and stability of electric vehicle inventory prediction are effectively improved.

The invention relates to a QL-RRVM-based supercapacitor remaining service life prediction method and an auxiliary module, and the method comprises the steps: constructing an improved RRVM, integrating a reinforcement learning-based feature extraction module in front of the improved RRVM, and obtaining a QL-RRVM model for predicting a degradation trend according to a time sequence; original data of the supercapacitor are collected, the state of the supercapacitor at the current moment is obtained based on any action, after a model is input, based on a predicted target value and an expected return guide decision, the remaining service life of the supercapacitor is dynamically estimated through the degradation trend obtained through prediction; and the auxiliary module cooperates with the supercapacitor management terminal to execute the method. The model training and prediction process is efficient, simple, convenient and easy to implement, is suitable for various degradation systems, has wide application prospects, improves the interpretability of prediction results, and greatly improves the real-time performance and efficiency of the system on the premise of ensuring the prediction precision; effective information loss is suppressed, and more efficient prediction is realized.

The invention provides an intelligent detection method for a direct-current series arc fault in a more-electric aircraft scene, and relates to the field of electrical parameter detection, and the method comprises the following steps: collecting direct-current series current data in a working process of a direct-current power supply system; inputting the direct-current series current data into an arc fault detection model to obtain a direct-current series arc fault probability and a direct-current series arc fault type; the arc fault detection model comprises four layers of time convolution networks TCN and one layer of relevance vector machine RVM which are connected in sequence; in the arc fault detection model before training, the number of channels of four layers of TCN is doubled layer by layer; in the trained arc fault detection model, pruning processing is carried out on all TCN channels, so that the parameter quantity of the TCN is reduced to 30% of the parameter quantity before training; and comparing the DC series arc fault probability with an alarm threshold, and when the DC series arc fault probability is greater than the alarm threshold, determining that the DC power supply system will have a corresponding type of DC series arc fault.

The invention provides a nonlinear dynamic data anomaly detection method for multiple process stages of silicon single crystal growth, and belongs to the technical field of integrated circuit silicon single crystal preparation. Comprising the following steps: acquiring a historical data set and a monitoring data set in a silicon single crystal growth process, and constructing a reconstruction training set and a reconstruction test set; performing iterative training on a relevance vector machine model by using the reconstructed training set to obtain a relevance vector machine training model; performing single-step prediction on the monitoring data set to obtain a prediction sequence, and obtaining a residual sequence by using the prediction sequence; dividing the residual error sequence into a plurality of segmented windows, and sequentially traversing each segmented window by using all sliding windows to obtain a plurality of sliding window data sets; and a dynamic threshold strategy is introduced, anomaly detection is performed on the corresponding sliding windows according to the candidate threshold set generated by each sliding window data set, and all anomaly detection results are fused into a final anomaly detection result. According to the invention, the accuracy and reliability of abnormal detection of the nonlinear data can be effectively improved.

The invention provides an unmanned aerial vehicle inspection terminal data anomaly detection method and system based on behavior coding cooperating with Transform-RVM, and relates to the technical field of artificial intelligence, the method comprises the following steps: preprocessing sensor original data of an unmanned aerial vehicle inspection terminal, and generating a time sequence sample set; the time sequence samples are converted into behavior coding vectors, and space-time fusion features are generated by fusing equipment space topology information through a graph convolutional network; inputting the space-time fusion features into a Transform editor, and extracting deep features through a multi-head self-attention mechanism, residual connection, layer normalization and a feedforward network; constructing a probability classification model by adopting a relevance vector machine RVM, modeling a parameter optimization process by utilizing a neural differential equation, and performing efficient training in combination with a dynamic sparsity control and adjoint sensitivity method to obtain a highly sparse RVM model; and performing anomaly detection on the unmanned aerial vehicle inspection terminal data acquired in real time by using the highly sparse RVM model. According to the scheme, the anomaly detection precision of the unmanned aerial vehicle inspection terminal data can be improved.

The application relates to a rolling bearing fault diagnosis method and system. Vibration signals of a rolling bearing to be detected are acquired, a plurality of decomposition algorithms are used for signal decomposition, a plurality of feature vectors under the plurality of decomposition methods are respectively obtained, the plurality of feature vectors are respectively input into a trained correlation vector machine model for fault diagnosis, a plurality of fault diagnosis outputs are obtained, the plurality of fault diagnosis outputs are fused, and a final diagnosis result is determined. The application adopts a plurality of decomposition methods to decompose the vibration signals, and fuses a plurality of fault diagnosis probability outputs obtained, so that fault features can be more effectively extracted, and the accuracy of fault diagnosis is improved.

The present application proposes a method for detecting anomalies of nonlinear dynamic data in multiple process stages of silicon single crystal growth, which belongs to the technical field of integrated circuit silicon single crystal preparation. The method includes: obtaining historical data sets and monitoring data sets during the growth of silicon single crystals, and constructing a reconstruction training set and a reconstruction test set; iteratively training the correlation vector machine model using the reconstruction training set to obtain a correlation vector machine training model; performing single-step prediction on the monitoring data set to obtain a prediction sequence, and using the prediction sequence to obtain a residual sequence; dividing the residual sequence into multiple segmented windows, and using all sliding windows to traverse each segmented window in turn to obtain multiple sliding window data sets; introducing a dynamic threshold strategy, performing anomaly detection on the corresponding sliding window according to the candidate threshold set generated for each sliding window data set, and fusing all anomaly detection results into the final anomaly detection result. The present application can effectively improve the accuracy and reliability of nonlinear data anomaly detection.

The present invention discloses a method and system for predicting icing on transmission lines based on multi-source satellite remote sensing, belonging to the technical field of predicting icing on transmission lines. The method includes: obtaining a historical data set; normalizing the indicators of remote sensing data; obtaining the misclassification rate of the indicators based on the random forest method; screening the main indicators according to the misclassification rate; screening the historical data set according to the main indicators to obtain a training set; training using the training set based on the multi-core relevance vector machine method to obtain an icing prediction model for transmission lines; predicting the remote sensing data according to the icing prediction model to obtain the icing thickness. Based on the random forest-multi-core relevance vector machine method, a multi-source satellite remote sensing data and actual icing data are used for training to construct an icing prediction model, and the icing prediction model is used to predict the icing thickness according to the remote sensing data; multiple kernel functions are combined to construct the icing prediction model to enhance the prediction effect of the model.

The present application discloses a water quality grade evaluation method, device and medium, which relate to the field of water quality grade evaluation. A water quality sample set is obtained; the average weight of each water quality evaluation factor is determined based on the sample set, and the target water quality evaluation factor is screened out based on the average weight to determine the target water quality evaluation factor combination; the target water quality parameters corresponding to the target water quality evaluation factor in each sample are screened based on the target water quality evaluation factor combination to obtain a target sample set; the mapping relationship between the target water quality parameter and the water quality grade of each target sample in the target sample set is determined, and a related vector machine is configured based on the mapping relationship to construct a water quality grade evaluation model, so as to determine the water quality grade of the water according to the water quality grade evaluation model. By screening out the water quality evaluation factors that contribute greatly to the water quality grade evaluation and constructing a water quality grade evaluation model, the complexity of the water quality evaluation model is reduced and the efficiency of the water quality grade evaluation is improved.

The invention relates to a power distribution network power supply reliability evaluation method and system, and the method comprises the steps: building a fault power failure influence factor evaluation system based on the analysis of power distribution network fault power failure influence factors of an analytic hierarchy process; building an improved firefly algorithm and relevance vector machine combined FA-RVM prediction model, and predicting a power distribution network fault power failure index; based on a simulated annealing-particle swarm optimization algorithm, constructing an SA-PSO-RVM prediction model; and designing a double-model complementary power distribution network fault power failure prediction platform, and predicting power distribution network fault power failure indexes, the double models being an IFA-RVM prediction model and an SA-PSO-RVM prediction model. The advanced algorithm model is utilized to predict the distribution network fault power failure, the requirements of safe, reliable and economical operation of the power system are met, the reliability of the power grid can be improved, accidents of the power system are reduced, and the economic benefits of power grid operation are improved.

The present invention discloses a method and device for fall recognition based on machine learning, and the method comprises: obtaining a feature set to be recognized in real time, inputting samples in the feature set to be recognized into a pre-built fall recognition model for recognition and obtaining recognition results; wherein, the construction process of the fall recognition model is: collecting human body motion data, and annotating the motion data with fall motions and normal human body activity motions to generate an annotation set; pre-processing the motion data in the annotation set to generate a data set; extracting features from the data set based on preset effective features to generate a sample set, and randomly dividing the sample set into a training set and a test set according to a preset ratio; inputting the training set and the test set into a pre-built relevance vector machine model and training and optimizing model parameters based on the crow algorithm; and using the trained and optimized relevance vector machine model as a fall recognition model. The present invention can accurately and efficiently recognize falls.

The invention discloses an intracranial aneurysm rupture state evaluation method based on haemodynamic mode characteristics of an incident flow and an impact domain, and the method comprises the steps: firstly constructing an intracranial aneurysm three-dimensional geometric model through the image segmentation and three-dimensional reconstruction method based on clinical 3DRA image data; and a computational fluid mechanics method is applied to obtain a hemodynamic cloud picture and corresponding hemodynamic parameters. The method comprises the following steps: determining an impact domain and an incident flow by applying hemodynamic flow field mode feature distribution, further exporting hemodynamic cloud charts in the two areas, constructing an 8-dimensional hemodynamic cloud depth information feature calculation model based on a Pointnet model, and calculating to obtain 8-dimensional depth information features of hemodynamic clouds of the impact domain and the incident flow. Morphological and hemodynamic parameters of the intracranial aneurysm and 8-dimensional depth information characteristics of hemodynamic clouds of an impact domain and an incident flow are integrated, and the rupture state of the intracranial aneurysm is classified and evaluated by adopting a relevance vector machine.