39 results about "Modal matrix" patented technology

The invention discloses a rolling bearing fault diagnosis method based on improved variational model decomposition and an extreme learning machine. The method comprises: vibration signals of a rollingbearing under different types of faults are collected, the vibration signals are filtered by means of maximum correlation kurtosis deconvolution, parameter optimization is carried out on the maximumcorrelation kurtosis deconvolution method by using a particle swarm algorithm, and an enveloped energy entropy after signal deconvolution is used as a fitness function; the mode number of variationalmodel decomposition is improved by an energy threshold and improved variational model decomposition of the filtered vibration signals is realized to obtain mode matrixes of the corresponding vibrationsignals; singular value decomposition is carried out on the mode matrixes to obtain a singular value vector and a rolling bearing fault feature set is constructed; and the fault feature set is trained by using an extreme learning machine and a rolling bearing fault diagnosis model is established. Therefore, stable feature extraction of the complex vibration signal of the rolling bearing is realized, so that the diagnostic accuracy is improved.

The invention provides a design method for resolving a modal space controller of a hydraulic drive six-degree-of-freedom parallel mechanism. A modal matrix is directly calculated according to system structure parameters; a six-degree-of-freedom strong coupling multi-input multi-output system is converted into a six-decoupling single-input single-output system in a modal space by conversion of the modal matrix; and parameters of a dynamic pressure feedback controller in the modal space are designed according to a hydraulic dynamic pressure feedback technology. The problem about coupling of a system is solved, and each decoupled modal space channel has the optimal control performance. The design method is only related to the system structure parameters, so that the application range of a modal decoupling controller is greatly expanded.

The invention relates to a CAE-based vibration fatigue simulation test method of a new energy vehicle battery pack.The CAE-based vibration fatigue simulation test method comprises the steps that virtual connection parts among a virtual battery pack, a virtual vibration test bench and a virtual battery pack and a virtual vibration test bench areconstructed to obtain a vibration fatigue virtual testdevice; modal calculation is performed on the virtual battery pack within the test vibration frequency range, and the natural frequency and modal matrix of the virtual battery pack are obtained; andactual vibration excitation load is applied to the vibration fatigue virtual test device under the natural frequency and modal matrix of the virtual battery pack; and the vibration fatigue parametersof the virtual battery pack under the actual vibration excitation load are calculated, whether the vibration fatigue parameters meet the vibration fatigue parameter threshold range conditions or not is determined, and if so, the initial parameters of the battery pack are stored to be taken as the design parameters of the battery pack for designing the battery pack. The CAE-based vibration fatiguesimulation test method evaluates whether the virtual battery pack is subjected to vibration fatigue damage or not during the test, and reduces the test cost of the battery pack design.

The invention discloses a method for realizing a satellite solar wing sensor layout on the basis of an EI (Effective Independence) method. The method comprises the following steps of: through the structural layout characteristics of a satellite solar wing, selecting a plurality of initial measurement points to construct a modal matrix [Phi]; taking the minimum covariance of an estimation error as optimal estimation to construct an independent modal matrix A; carrying out feature value decomposition on the A to extract an order with a high contribution degree as an optimized measurement point; and combining the requirements of the weight and the reliability of a sensor on the solar wing to determine an optimal layout method for a satellite solar wing modal measurement point. By use of the method, the optimal sensor arrangement of solar wing modal in-orbit measurement can be realized, and the implementation of in-orbit modal measurement is guaranteed.

The invention relates to a natural image classification method and device on basis of multi-modal matrix filling. The method comprises the steps of carrying out feature extraction on natural image data with labels, natural image data without labels and natural image data for testing, and obtaining different feature representations; adopting a matrix filling algorithm to generate estimation labels of all features of the data with the labels; carrying out linear combination on all the estimation labels to be approximate to the corresponding known true labels corresponding to the estimation labels to obtain a combination coefficient; for all features, utilizing the natural image data with the labels and adopting the matrix filling algorithm to predict the labels of the natural image data without the labels and the labels of the natural image data for testing; adopting the combination coefficient to combine the predicted labels of all the features to obtain labels combining multiple features; classifying the natural image data on the basis of the labels combining the multiple features. The natural image classification method and device on the basis of the multi-modal matrix filling are easy to achieve, high classification accuracy can be obtained, meanwhile, the advantages of image classification on the basis of matrix filling are inherited, and the natural image classification method and device on the basis of the multi-modal matrix filling are suitable for the fields of network picture summarizing and classifying, image retrieval and the like.

The invention discloses a grid structure optimization method for realizing reliable splitting based on node relevancy. According to the method, aiming at the problem that a splitting strategy is not converged due to the fact that a grouping mode is not remarkable and a splitting range is widely distributed in the current splitting process, the association degree between system nodes is changed byadopting a method of adjusting the number of lines, so that the purpose of reducing a splitting decision space is achieved. According to the invention, generalized characteristic analysis is carried out based on a slow homology theory, and a dominant mode is acquired to construct a modal matrix; a slow homology matrix is acquired through Gauss elimination of principal elements; a node classification criterion is improved, and a node relevancy model is constructed; system stability constraint conditions are comprehensively considered, and a DPSO intelligent optimization algorithm is utilized toobtain an optimization scheme of homology group strong coupling and non-homology group weak connection; the splitting decision space is highly centralized, which ensures that splitting sections underdifferent disturbances can fall into a weak connection area with high probability; internal oscillation of a homology group is avoided; the splitting effectiveness is improved; and the method has important significance in maintaining the safety and stability of a system.

The invention relates to an engineering testing method of the vibration field of a non-uniform arrangement structure. Based on the condition number of a structural vibration modal matrix, vibration measuring points causing an increase of the condition number of the matrix are removed in the loop iteration process, so that reduction of the number of the measuring points and optimization of the positions of the measuring points are achieved. Then based on a mode superposition method, only measuring data of the optimized measuring points with the number reduced is adopted, and vibration field reconstruction of the whole large structure is achieved. According to the method, a few sensors are optimized for position arrangement, and the vibration response field of the whole structure is reconstructed by means of the measuring data of the sensors, so that space and weight are saved, the testing workload is significantly relieved, and a monitoring position optimization method is provided for avibration monitoring system.

The invention relates to a distributed optical fiber computing method for wing strain field reconstruction based on a modal superposition principle, and belongs to the field of structural health monitoring. The method comprises the following steps that 1, based on measurement of fiber grating sensor wing structure strain, calculating the strain of the surface of a wing structure through fiber grating center wavelength offset; and 2, calculating a strain mode of the wing structure, and optimizing the fiber grating sensor network layout of the wing structure through the strain mode; 3, deducinga strain matrix corresponding to the position of the measurement point of the wing structure sensor and a conversion formula between strain matrixes of all points of the wing structure; and 4, for theproblem that the strain mode matrix is uncertain, adopting a regularization method to optimize the solution of the strain mode coordinates, and reducing the strain field reconstruction error. According to the method, response signals of fiber gratings with a small number of discrete points are collected, and a wing structure strain field is deduced and calculated through a formula.

The invention relates to a wind turbine tower stress state calculation method based on a modal superposition method, Firstly, the finite element model of the tower is established, then the truncated modal matrix of the tower is calculated by the finite element method, and the principal vibration coordinates of the tower are calculated according to the real-time vibration values at the monitoring points. Finally, the real-time stress state of the tower is calculated by the modal superposition method. Using this method, the stress states of the whole tower and flange bolts can be calculated quickly according to the real-time vibration values of the finite vibration measuring points of the tower. The method has the characteristics of high calculation accuracy and fast calculation speed, and can meet the requirements of real-time online calculation.

The invention discloses an electric performance-oriented space mesh antenna key dynamical modal selection method. The method comprises the following steps of: inputting geometric parameters, material parameters, electric parameters of modal guillotine factors; calculating an ideal antenna far electric field; establishing an antenna structure finite element model; extracting node, unit and shape function information; calculating a unit electric performance first-order coefficient matrix; grouping an overall electric performance first-order coefficient matrix; carrying out antenna structure modal analysis; outputting a modal matrix; calculating a first-order sensitivity, for the modal, of the electric performance; outputting a modal shape equivalent mass fraction; calculating significance factors, for the modal, of the electric performance; judging whether the significance factors are greater than the modal guillotine factors or not; if the judging result is positive, selecting the significance factors; and otherwise, abandoning the significance factors. According to the method disclosed by the invention, the deficiencies of the traditional dynamical modal selection method are overcome, the influences, on the electric performance of antennas, of modals in different orders are integrally considered in the modal selection process, and the problem of electric performance-oriented space mesh antenna key dynamical modal selection is solved.

The invention discloses a bearing composite fault diagnosis method based on minimum correlation local non-negative matrix factorization. The method comprises the following steps: firstly, introducing a grasshopper optimization algorithm to optimize variational modal decomposition parameters, adaptively determining an optimal modal number and a penalty factor, inputting the optimal parameters into variational modal decomposition to decompose a vibration signal to obtain a plurality of modal components, and reconstructing the modal components to obtain an input matrix; secondly, introducing the minimum correlation constraint into a local non-negative matrix to form a minimum correlation local non-negative matrix factorization algorithm, reconstructing a plurality of modal components and an original signal into a modal matrix, and calculating the optimal dimension of the minimum correlation local non-negative matrix factorization; then performing optimal dimensionality decomposition on the input matrix by using a minimum correlation local non-negative matrix decomposition algorithm to obtain a basis matrix W and a coefficient matrix H, and finally performing envelope spectrum analysis on the basis matrix W to separate coupled bearing composite fault signals. The effectiveness of the method is verified through simulation analysis of the composite fault signal. The experimental data analysis result also shows that the method can effectively separate and diagnose the bearing composite fault.

The invention discloses a patent text retrieval method and device based on multi-modal matrix vector representation, a word vector set of all words is obtained through training according to an existing patent data set, so that word vectors can contain information of all keywords, an image vectorization representation model is obtained through training of the existing patent data set, and the retrieval efficiency is improved. According to the method, the attached drawings in the patent are extracted, the graph vectors corresponding to the attached drawings are obtained, the graph vectors and the word vectors are combined, when the patent is retrieved, a large amount of useful information contained in the attached drawings in the patent is fully utilized, and meanwhile, some retrieval requirements of searching for a text through a graph, searching for a graph through a text and searching for a graph through a graph in the current market are met; and the patent retrieval result is more accurate.

The invention provides a method for detecting singular and movement bifurcation of an articulated bar system mechanism by utilizing a mechanism displacement modal submatrix method. The method comprises the following steps of (1) determining a driving node and a driven node in free nodes of the articulated bar system mechanism; (2) according to configuration generated in the moving process of the articulated bar system mechanism, establishing a balance matrix of the articulated bar system mechanism, and obtaining a whole mechanism displacement modal matrix through singular value decomposition; (3) decomposing the whole mechanism displacement modal matrix into a mechanism displacement modal matrix Ua (2l*m) corresponding to the driving node and a mechanism displacement modal matrix Up (2k*m) corresponding to the driven node; and (4) computing a singular value of a mechanism displacement modal submatrix, and judging that the singular and movement bifurcation happen to the articulated bar system mechanism if the minimum singular value tends to zero. The method is strong in operability, provides structural design and construction design basis for a novel space structure-mechanism, and has wide application prospect.

The invention discloses a rotary machinery shafting fatigue strength detection method which comprises the following steps of S1, establishing a concentrated mass torsional vibration model of a targetshafting to obtain a motion equation of the target shafting under the action of a dynamic torque T; s2, establishing a modal matrix Phi and calculating to obtain a regularization factor alpha so as toobtain a regularization modal matrix Phin; s3, converting the motion equation into a regular modal equation; s4, solving the responses and torque distribution of all shaft sections of the target shaft system; and S5, detecting the fatigue strength of the target shaft system. According to the present invention, the torque distribution of the whole shaft system can be calculated only by arranging one measuring point, so that the field test process is simpler and more efficient, and the test efficiency is improved. According to the invention, the position of the measuring point can be selected at any shaft section of a shaft system, the method has very strong application capability in specific engineering practice, and is beneficial to adapting to the complex conditions of an engineering site.

The invention relates to a geological disaster multi-modal monitoring data fusion imaging method based on a resistivity method. The method is characterized in that resistivity monitoring data is usedas a complete modal subspace, a spatial and temporal discontinuous data mode including water level, deep displacement and soil humidity is used as an incomplete modal data set, and a geometric structure of the data is coded through a graph regularization factor to ensure the local similarity of each modal data; deep learning and incomplete multi-modal analysis are fused to construct a deep semantic matching model fusing a modal deep neural network and incomplete multi-modal matrix decomposition; the deep semantic matching characteristics of the multi-modal data are obtained by combining deep matrix decomposition, optimization and layer-by-layer modal semantic matching and updating, and the incomplete multi-modal data in the shared space at least comprises classification and clustering analysis. The effectiveness of a dynamic change inversion section of a landslide seepage field, a geologic body structure field and a deep displacement field generated by coupling is improved.

The invention provides a response solving method and device for a rotor system. The method comprises the following steps: establishing a system dynamic model for the rotor system according to a finite element method; adding an equivalent virtual support stiffness matrix to the system dynamic model to obtain a system equivalent dynamic model of the rotor system; performing modal analysis on the system equivalent dynamic model to obtain a modal matrix; performing order reduction on the modal matrix, and obtaining an order-reduced system equivalent dynamic model based on the modal matrix after order reduction; and performing response solving on the reduced-order system equivalent dynamic model according to a state space method to obtain the dynamic response of the rotor system. According to the method, the equivalent virtual support is added in the system dynamic model, so that stable modal separation of the algorithm is ensured, and the order reduction is performed on the modal matrix, so that the order-reduced system equivalent dynamic model is obtained, and the calculation efficiency and the calculation precision of rotor system response are improved. The device is used for implementing the method.

The invention relates to a structure strain sensor measuring point optimization method. The invention particularly relates to a sensor measuring point optimization method for reducing a strain field reconstruction error. The layout advantages of fiber grating sensors are utilized to overcome the defects in the prior art, strain field reconstruction of a structure can be realized based on a strainfield reconstruction method of a modal superposition idea. The method comprises the following steps: step (1), firstly, establishing a finite element model of an actual structure, and performing modelcorrection on the established finite element model to ensure that the established finite element model has precision meeting requirements; (2) selecting a modal order required by strain field reconstruction, and extracting a strain modal matrix of the finite element model of the structure for the corrected finite element model; and (3) constructing a function relationship between the local strainmeasurement information and the overall strain field as shown in the specification: {epsilon all}=[T]{epsilon loc}, wherein [T] is a conversion matrix, {epsilon all} is a strain field vector, and {epsilon loc} is a local measurement point matrix.

The invention discloses a method for condensing FETI engineering numerical values based on regional decomposition. The method comprises the steps of solving a generalized flexibility matrix and a boundary flexibility matrix through constructing a rigid matrix only related to the boundary freedom degree of a floating sub-domain and a rigid body modal matrix of the rigid matrix. The formed new 'hat' matrix can fully utilize and maintain the sparsity of the floating sub-domain rigid matrix, and the order of the new 'hat' matrix is equal to the number of floating sub-domain boundary degrees of freedom and is far smaller than the scale of the original 'hat' matrix, so that the storage space is greatly reduced, and the triangular decomposition solving efficiency of the generalized stiffness matrix is improved; and the calculation of the local flexibility matrix becomes economical and feasible. Meanwhile, an intermediate result of the new 'hat' matrix calculation just provides a Dirichlet preprocessor with a better convergence characteristic for PCPG iterative solution of a frame equation set.

The invention discloses a rolling bearing fault diagnosis method based on improved variational mode decomposition and extreme learning machine, which is characterized in that: the vibration signals of rolling bearings under different types of faults are collected, and the maximum correlation kurtosis deconvolution is used to filter the vibration signals, Using the particle swarm algorithm to optimize the parameters of the maximum correlation kurtosis deconvolution method, the envelope energy entropy after signal deconvolution is proposed as the fitness function; the energy threshold is proposed to improve the number of modes in the variational mode decomposition , realize the improved variational mode decomposition of the filtered vibration signal, and obtain the modal matrix of the corresponding vibration signal; perform singular value decomposition on the modal matrix, obtain a singular value vector and construct a rolling bearing fault feature set; use extreme learning The computer trains the fault feature set to establish a rolling bearing fault diagnosis model. The invention realizes the stable feature extraction of the complex vibration signal of the rolling bearing, thereby improving the diagnostic accuracy.

The invention discloses a calculation method for the vibration state of a wind turbine generator system tower based on a modal superposition method. The calculation method comprises the steps of firstly, establishing a finite element model of a tower; then, calculating with a finite element method to obtain a truncation modal matrix of the tower, and calculating according to real-time vibration values of monitoring points to obtain a vibration principal coordinate vector of the tower; and finally, calculating with a modal superposition method to obtain the real-time vibration state of the whole tower. According to the method, the vibration state of the whole tower can be obtained through quick calculation according to the real-time vibration values of finite vibration monitoring points ofthe tower; and the calculation method has the characteristics of high calculation accuracy and speed, and the requirements of real-time online calculation can be met.

The invention provides a small current grounding fault line selection method. The method comprises the following steps: 1, collecting a bus three-phase voltage value and a bus zero sequence voltage; 2, judging whether a single-phase earth fault occurs or not; 3, if the single-phase earth fault does not occur, returning to the step 1, and if so, starting a wave recording device; 4, starting the wave recording device to record fault data of one cycle after an event node; 5, carrying out singularity detection on the zero sequence voltage value 3U0 of the bus by adopting a Hilbert-Huang transform technology, and confirming the fault moment of the power distribution network; 6, performing SVD filtering on the transient zero-sequence current of each outgoing line; 7, performing morphological filtering on the transient zero-sequence current of each outgoing line; 8, performing empirical mode decomposition on the transient zero-sequence current of each outgoing line and constructing a mode matrix; 9, processing the modal matrix by using a singular value decomposition technology and constructing a fault eigenvalue matrix; and 10, extracting characteristic values by adopting a PSO-FCM clustering technology to judge a fault line.

The embodiment of the invention discloses a sensor measuring point position determination method and device. The method comprises the steps of acquiring a strain modal matrix of a multi-order modal; determining a Fischer information matrix according to the strain modal matrix; calculating a value of a determinant of the Fischer information matrix; and when the value of the determinant of the Fischer information matrix is determined to be the maximum value, indicating that the position of the target node identified in each order of modality in the multi-order modalities is the measuring point position of the sensor, wherein the target node is the node with the maximum modal strain energy in the plurality of nodes included in the order of modalities. Thus, when the value of the determinant of the Fischer information matrix is determined to be the maximum value, the position of the target node identified in each order of modality in the multi-order modalities is the measuring point position of the sensor, and redundancy of sensor configuration is avoided.

The invention relates to a low-voltage direct-current power distribution network node inertia evaluation method based on contribution factors, which comprises the following steps: introducing a modal matrix to decompose the dynamic process of each node state variable into basic mode components, calculating the contribution factors of the basic mode components to each state variable, calculating a value time interval for least square index fitting, and calculating an equivalent time constant of least square exponential fitting, and calculating a node inertia index. The method has the advantages that the modal matrix is introduced, the dynamic process of each node state variable is decomposed into superposition of a series of basic modes, and each basic mode has the determined attenuation characteristic; contribution factors are defined, a node inertia evaluation method based on the contribution factors is provided, and it is pointed that the contribution factors of all modes are related to a system structure, system parameters and input parameters; a value time interval selection method and an equivalent time constant calculation method based on least square exponential fitting are provided, the curve fitting precision is guaranteed, and the evaluation accuracy is high.

The invention provides a spatial six-degree-of-freedom motion platform modal control method. In the invention, a spatial six-degree-of-freedom motion platform modal matrix is used to convert a strong-coupling physical space system into a decoupled modal space system, a modal control concept is introduced on the basis of the conventional spatial six-degree-of-freedom hydraulic motion platform control, the modal conversion matrix is used to decouple a strong dynamic coupling six-degree-of-freedom hydraulic motion platform, an expected physical input signal and a motion platform actual output signal are converted into modal signals to perform independent modal control regulation, the drive and control of the spatial six-degree-of-freedom hydraulic motion platform are realized, the coupling influences among the actuators and degrees of freedom in a spatial six-degree-of-freedom hydraulic motion are weakened effectively, indexes such as the single degree of freedom motion and multiple degree of freedom motion reproduction of the six-degree-of-freedom hydraulic motion platform are improved, and the bandwidths of degrees of freedoms besides the close degree of freedom of a first order modal are improved.

The invention provides a modal modeling method of a kinematic system with spatial six degrees of freedom, which comprises the following steps: scattering a parallel kinematic system with complex spatial six degrees of freedom into a plurality of substructures; computing the modalities of the previous orders of the substructures by adopting a substructure transfer matrix method to acquire low-order modality characteristics of the substructures; constructing a modal matrix; converting physical coordinates into modal coordinates by modal coordinate conversion; establishing a modal kinematic second order differential equation for each substructure; removing the redundant degree of freedom; carrying out modal synthesis on each substructure, thereby establishing an integral modal kinematic equation of the parallel kinematic system with spatial six degrees of freedom to acquire a modal model of the parallel kinematic system with spatial butted six degrees of freedom; and finally, converting the modal model into physical coordinates through coordinate conversion to acquire a physical spatial dynamic equation. The modal modeling method of the kinematic system, which is provided by the invention, has overlarge working space, overlong kinematic routes of supporting legs and high bandwidth and accuracy requirements of the system, and overcomes the limitation of multi-rigid body modeling of a spatial butted overlarge kinematic system.