CN113468771B - Vibration estimation method using structure intrinsic parameters - Google Patents
Vibration estimation method using structure intrinsic parameters Download PDFInfo
- Publication number
- CN113468771B CN113468771B CN202111026053.7A CN202111026053A CN113468771B CN 113468771 B CN113468771 B CN 113468771B CN 202111026053 A CN202111026053 A CN 202111026053A CN 113468771 B CN113468771 B CN 113468771B
- Authority
- CN
- China
- Prior art keywords
- vibration
- parameters
- parameter
- estimation
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
Abstract
The embodiment of the invention discloses a vibration estimation method by using structure intrinsic parameters, which comprises the following steps: acquiring target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected; inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters; the vibration parameters of the structure to be measured are estimated according to the output result of the target vibration estimation network model, so that the target vibration parameters of the structure to be measured are obtained, the problem that the estimation result is inaccurate due to the fact that estimation can be carried out only according to historical vibration parameters is solved, and the purpose of estimating the vibration parameters by using the inherent parameters of the structure is achieved. The external environment is simulated through the estimation parameters, the vibration parameters of the structure to be measured are effectively estimated, the accuracy of the vibration parameter estimation is improved, the vibration parameters of the structure in any environment are simulated, the method is easy to realize, and the applicable scene is wide.
Description
Technical Field
The embodiment of the invention relates to the technical field of vibration estimation, in particular to a vibration estimation method utilizing structure intrinsic parameters.
Background
Vibration is a form of behavior of a structure under external excitation. Any excitation causes structural vibration phenomena, which are particularly common in many areas of aerospace, wind energy, road transport, rail transport, and sea transport. When the structure vibrates largely, structural damage may be caused. Therefore, the structural vibration research is an important foundation for guaranteeing the operation safety of various equipment in the fields of aerospace, wind energy, road transportation, railway transportation, marine transportation and the like, and the reliability and the safety of the equipment can be better improved by structural vibration estimation. Through vibration estimation, preventive maintenance can be effectively carried out, and structural vibration of equipment to be operated is obtained in advance, so that unpredictable structural vibration accidents under the traditional technical means are prevented, the working state of the equipment is evaluated, vibration faults are checked, and the method has important significance for guaranteeing safe and stable operation of the equipment.
Currently, vibration estimation is generally performed by collecting historical vibration parameters for analysis, and estimating the vibration parameters according to the analysis result. However, when the vibration estimation is performed in this way, the accuracy of the estimation result is low, and the vibration parameters of the structure cannot be well estimated.
Disclosure of Invention
The invention provides a vibration estimation method by utilizing structure intrinsic parameters, so as to realize accurate estimation of the vibration parameters of a structure.
In a first aspect, an embodiment of the present invention provides a vibration estimation method using structure intrinsic parameters, where the vibration estimation method using structure intrinsic parameters includes:
acquiring target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected;
inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters;
and estimating the vibration parameters of the structure to be detected according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be detected.
In a second aspect, an embodiment of the present invention further provides a vibration estimation apparatus using structure intrinsic parameters, where the vibration estimation apparatus using structure intrinsic parameters includes:
the parameter acquisition module is used for acquiring target frequency domain parameters corresponding to target intrinsic parameters of the structure to be detected;
the input module is used for inputting the target frequency domain parameters into a predetermined target vibration estimation network model as input data, wherein the target vibration estimation network model is obtained by training intrinsic parameters;
and the estimation module is used for estimating the vibration parameters of the structure to be detected according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be detected.
In a third aspect, an embodiment of the present invention further provides a computer device, where the computer device includes:
one or more processors;
a memory for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement a method of vibration estimation using structure-inherent parameters as in any of the embodiments of the present invention.
In a fourth aspect, the embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements a vibration estimation method using structure intrinsic parameters according to any one of the embodiments of the present invention.
The embodiment of the invention provides a vibration estimation method, a device, equipment and a storage medium by utilizing structure intrinsic parameters, which are used for obtaining target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected; inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters; and estimating the vibration parameters of the structure to be measured according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be measured, solving the problem that the estimation result is inaccurate because the estimation can be carried out only according to historical vibration parameters in the existing vibration parameter estimation method, and estimating the vibration parameters by using the inherent parameters of the structure. The method has the advantages that the external environment is simulated through the estimation parameters, the vibration parameter estimation of the structure to be measured can be effectively realized, the accuracy of the vibration parameter estimation is improved, the vibration parameters of the structure under any environment can be simulated, the method is easy to realize, the applicable scene is wide, and the vibration estimation can still be carried out under the condition that the vibration parameters cannot be collected.
Drawings
FIG. 1 is a flow chart of a vibration estimation method using intrinsic parameters of a structure according to a first embodiment of the present invention;
FIG. 2 is a flow chart of a vibration estimation method using intrinsic parameters of a structure according to a second embodiment of the present invention;
FIG. 3 is a diagram of an implementation example of a vibration estimation method using intrinsic parameters of a structure according to a second embodiment of the present invention;
FIG. 4 is a schematic structural diagram of a vibration estimation apparatus using intrinsic parameters of a structure according to a third embodiment of the present invention;
fig. 5 is a schematic structural diagram of a computer device in the fourth embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more clear, embodiments of the present application will be described in further detail below with reference to the accompanying drawings. It should be understood that the embodiments described are only a few embodiments of the present application, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
When the following description refers to the accompanying drawings, like numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the application, as detailed in the appended claims.
In the description of the present application, it is to be understood that the terms "first," "second," "third," and the like are used solely to distinguish one from another and are not necessarily used to describe a particular order or sequence, nor are they to be construed as indicating or implying relative importance. The specific meaning of the above terms in the present application can be understood by those of ordinary skill in the art as appropriate. Further, in the description of the present application, "a plurality" means two or more unless otherwise specified. "and/or" describes the association relationship of the associated objects, meaning that there may be three relationships, e.g., a and/or B, which may mean: a exists alone, A and B exist simultaneously, and B exists alone. The character "/" generally indicates that the former and latter associated objects are in an "or" relationship.
Example one
Fig. 1 is a schematic flowchart of a vibration estimation method using structure intrinsic parameters according to an embodiment of the present application, where the method is applied to estimating vibration parameters of a structure. The method can be performed by a computer device, which can be formed by two or more physical entities or by one physical entity. Generally, the computer device may be a notebook, a desktop computer, a smart tablet, and the like.
As shown in fig. 1, a vibration estimation method using structure intrinsic parameters provided in this embodiment specifically includes the following steps:
s110, acquiring target frequency domain parameters corresponding to the target intrinsic parameters of the structure to be detected.
In this embodiment, the structure to be measured may be specifically understood as a hardware structure having a requirement for estimating vibration parameters, for example, a thin plate; the target intrinsic parameters are specifically understood to be intrinsic parameters associated with the structure to be measured, such as environmental parameters, which may be used to simulate the environment in which the structure to be measured is located, such as height, temperature, speed, angle, and the like, and all of which are intrinsic parameters and may affect vibration parameters of the structure to be measured. Therefore, the environment of the structure to be measured is simulated through the target intrinsic parameters, or the environment of the structure to be measured is reflected through the target intrinsic parameters, so that the vibration estimation is carried out on the structure to be measured. The target frequency domain parameter may be specifically understood as a parameter in a frequency domain form after performing frequency domain transformation.
Specifically, a target intrinsic parameter of the structure to be measured is obtained, the target intrinsic parameter is a time domain parameter, and frequency domain transformation is performed on the target intrinsic parameter to obtain a target frequency domain parameter. When the frequency domain transformation is carried out, one or more target frequency domain parameters can be obtained by changing the phase angle.
It should be noted that, when estimating the vibration by using the structure intrinsic parameters, the structure intrinsic parameters include different types of parameters, the vibration parameters also include different types of parameters, and the intrinsic parameters affecting the vibration parameters may be different for each type of vibration parameters, so that the type of at least one intrinsic parameter related to each type of vibration parameters needs to be predetermined. When the vibration is estimated, the target intrinsic parameters are selected according to the type of the vibration parameter to be estimated and the type of the related intrinsic parameters known in advance, and at this time, more than one target intrinsic parameters may be selected, and the target intrinsic parameters may be composed of different types of intrinsic parameters.
And S120, inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training the inherent parameters.
In this embodiment, the target vibration estimation network model may be specifically understood as a neural network model that can estimate vibration parameters according to intrinsic parameters. The target vibration estimation network model comprises one or more submodels, vibration parameters are estimated according to inherent parameters through the training submodels, then the estimation results of the submodels are processed in a weighting, maximum value, minimum value, median and other calculation modes, the final result is obtained and serves as the estimation result, and when the final estimation result is calculated in a weighting mode, all weight values are determined in a model training mode to obtain weight coefficients.
Specifically, the model is trained in advance, parameters of the model are continuously adjusted according to the loss function in the training process, and finally the target vibration estimation network model meeting the requirements is obtained to complete the training. The trained target vibration estimation network model can directly input data, and an estimation result is obtained according to learning experience. The target vibration estimation network model comprises at least one submodel, each submodel is of different types, vibration parameters are estimated according to intrinsic parameters, and final vibration parameters are determined according to estimation results of the models, so that the estimation results are more accurate.
S130, estimating the vibration parameters of the structure to be measured according to the output result of the target vibration estimation network model, and obtaining the target vibration parameters of the structure to be measured.
In this embodiment, the target vibration parameter may be specifically understood as a vibration parameter corresponding to the structure to be measured in an environment of the target intrinsic parameter, and is a result estimated according to the target intrinsic parameter. And the target vibration estimation network model processes the input target frequency domain parameters according to the learned experience, estimates the vibration parameters of the structure to be detected, outputs the estimation result as the output result of the model to realize the estimation of the vibration parameters, and performs time domain transformation on the output result of the model, namely frequency domain information to obtain the target vibration parameters of the structure to be detected.
The embodiment of the invention provides a vibration estimation method by utilizing structure intrinsic parameters, which comprises the steps of obtaining target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected; inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters; and estimating the vibration parameters of the structure to be measured according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be measured, solving the problem that the estimation result is inaccurate because the estimation can be carried out only according to historical vibration parameters in the existing vibration parameter estimation method, and estimating the vibration parameters by using the inherent parameters of the structure. The method has the advantages that the external environment is simulated through the estimation parameters, the vibration parameter estimation of the structure to be measured can be effectively realized, the accuracy of the vibration parameter estimation is improved, the vibration parameters of the structure under any environment can be simulated, the method is easy to realize, the applicable scene is wide, and the vibration estimation can still be carried out under the condition that the vibration parameters cannot be collected.
Example two
Fig. 2 is a flowchart of a vibration estimation method using structure intrinsic parameters according to a second embodiment of the present invention. The technical scheme of the embodiment is further refined on the basis of the technical scheme, and specifically mainly comprises the following steps:
s210, a standard vibration parameter set is obtained, and at least one training intrinsic parameter set associated with the standard vibration parameter set is determined, wherein the standard vibration parameter set comprises at least one standard vibration parameter, and the training intrinsic parameter set comprises at least one training intrinsic parameter.
In this embodiment, the standard vibration parameter may be specifically understood as a vibration parameter serving as a reference standard, and the vibration parameter in the training process uses a frequency domain signal, so that the standard vibration parameter in this embodiment of the present application takes the frequency domain signal as an example, and obtains a time domain vibration parameter in advance, and performs frequency domain transformation on the time domain vibration parameter to obtain the standard vibration parameter. A standard vibration parameter set may particularly be understood as a data set storing standard vibration parameters. The training intrinsic parameters can be specifically understood as structural intrinsic parameters used in training; a training intrinsic parameter set is understood in particular to mean a data set which stores training intrinsic parameters.
Specifically, a standard vibration parameter set used for training is predetermined, and the type of vibration parameter to be estimated may be selected according to the type of vibration parameter to be estimated, so as to form the standard vibration parameter set. And processing the intrinsic parameters and the vibration parameters of different types in advance, and determining the type of the intrinsic parameter associated with each type of vibration parameter. Further, after obtaining the standard set of vibration parameters, at least one training intrinsic parameter set associated with the standard set of vibration parameters may be determined.
As an optional embodiment of this embodiment, this optional embodiment further optimizes the at least one training intrinsic parameter set determined to be associated with the standard set of vibration parameters to a1-a 5:
a1, obtaining a data file, processing the data file according to the predetermined file attribute characteristics, and determining an original data set, wherein the original data set comprises at least one original sub data set.
In this embodiment, the data file may specifically be a file storing raw data, where the raw data refers to data of a structure in various environments in engineering application, and the raw data may be collected by a sensor. The file attribute characteristics can be specifically understood as format characteristics such as the position, the size and the like of data stored in a file, for example, the data type is stored in the first byte, the specific data content is stored in the second byte and the like. The original data set can be specifically understood as a set formed by original data; an original sub data set may be understood to be a set of original data of the same kind.
It should be noted that the raw data refers to intrinsic parameters.
Specifically, when the original data is stored in the data file, the file attribute characteristics of the file storage are preset, so that the original data is stored according to a certain characteristic rule when being stored. And acquiring a data file, analyzing the data in the data file according to the file attribute characteristics, and acquiring each original data and information and characteristics corresponding to each original data. The original data are classified according to data types, the original data of the same type are stored in the same original sub data set, and each original sub data set forms an original data set. When the original data is read from the data file, if the data volume of the original data is too large, the original data can be processed in an undersampling mode, and one piece of data is obtained at intervals, so that the data volume is reduced.
A2, screening the original data set for at least one alternative set of intrinsic parameters associated with the set of standard vibration parameters.
In the present embodiment, an alternative set of intrinsic parameters may specifically be understood as storing a set of intrinsic parameters related to the standard vibration parameters. And calculating the correlation degree of the standard vibration parameter set and each original sub data set in the original data set, and selecting the original sub data set with higher correlation degree as an alternative inherent parameter set.
As an optional embodiment of this embodiment, this optional embodiment further optimizes at least one alternative set of intrinsic parameters associated with the screening of the standard set of vibration parameters from the raw data set to a21-a 24:
a21, aligning the data in the raw data set and the standard vibration parameter set.
Unifying the data in the original data set and the data in the standard data set on a time unit and a data volume to ensure that the data are in one-to-one correspondence, and aligning the initial first data to realize data alignment.
As an optional embodiment of this embodiment, this optional embodiment further optimizes the aligning of the raw data set and the data in the standard vibration parameter set as follows: unifying time stamp units of the original data set and the standard vibration parameter set data, and performing time stamp matching on the original data set and the standard vibration parameter set data; and adjusting the data volume of each matched original sub data set and standard vibration parameter set to be the same, and aligning the data in sequence.
Specifically, the units of time stamps for all data in the raw data set and the standard vibration parameter set are unified, for example, as numerical values in milliseconds. The data volume in the original sub data set and the standard vibration parameter set may be different for the intrinsic parameters and the vibration parameters due to different acquisition frequencies. Data alignment is divided into two cases, i.e., the same sampling frequency and different sampling frequencies, according to whether the sampling frequencies are the same or not. For the same sampling frequency, when the first time stamps of the two variables are determined to be equal in size, the matching of the time stamps is completed when the matching is successful at the first time stamp, the alignment of the data is completed, and the remaining data are aligned one by one according to the arrangement sequence of the data. For data alignment of different sampling frequencies, one method is to copy and fill the vibration parameters with lower sampling frequency according to the size of the timestamp, for example, calculate the average value, the maximum value, the minimum value, the median, the mode, etc. of the data in a period of time, and take the obtained data as the vibration parameters of the corresponding time point; another method is to perform information compression on the intrinsic parameters with higher sampling frequency, where the information compression is to extract feature values within a period of time, for example, averaging the data in 1 st to 3 rd s to obtain the intrinsic parameters at corresponding time points. After the data volumes of the original sub data set and the standard vibration parameter set are adjusted to be the same through the two modes, the residual data are sequentially aligned one by one according to the arrangement sequence of the data.
Illustratively, the timestamps of the intrinsic parameters in the original data set are, in order: 5/7/10/2021: 00: 00: 00, 5/2021, 7/10: 00: 00: 10, 5/2021, 7/10: 00: 00: 20, …; the timestamps of the standard vibration parameters in the standard vibration parameter set are as follows in sequence: 5/7/10/2021: 00: 0: 10, 5/2021, 7/10: 00: 00: 20, 5/2021, 7/10: 00: 00: 30, …. Timestamp 2021 year 5 month 7 day 10: 00: 00: and 10, completing matching, and carrying out data quantity adjustment and data alignment on subsequent data by taking the timestamp as a start. The data are arranged according to the time sequence, and after the data quantity is adjusted to be the same, each intrinsic parameter has a corresponding standard vibration parameter from the first intrinsic parameter.
It should be noted that the sampling frequency of each original sub data set is generally the same, and therefore, the data amount of the intrinsic parameter in each original sub data set is generally the same, and if different, the data amount can be adjusted to be the same in the above manner. The data alignment has the function of finding the inherent parameters corresponding to the vibration parameters, so that the correlation can be conveniently calculated subsequently.
A22, determining the aligned standard vibration parameter set and the correlation coefficient of each original sub data set.
And calculating the correlation coefficient of the standard vibration parameters in each aligned standard vibration parameter set and the inherent parameters in the original sub-data set, wherein the linear correlation coefficient, such as the Pearson correlation coefficient, is adopted when calculating the correlation coefficient, so that the calculation result is more accurate.
And A23, screening out alternative correlation coefficients meeting preset requirements from the correlation coefficients.
In this embodiment, the alternative correlation coefficient may be specifically understood as a correlation coefficient with the highest correlation degree selected from the correlation coefficients, and may better reflect the correlation with the standard vibration parameter. The preset requirement may be to select a certain number of correlation coefficients, for example, to select 20 correlation coefficients from high to low according to the correlation as candidate correlation coefficients; or selecting a correlation coefficient larger than a certain threshold, for example, selecting a correlation coefficient larger than 0.4 from the correlation coefficients as a candidate correlation coefficient; or considering the two conditions, and comprehensively selecting the alternative correlation coefficients meeting the preset requirements.
And A24, taking each original sub data set corresponding to each alternative correlation coefficient as an alternative inherent parameter set.
And each alternative correlation coefficient corresponds to one original sub data set, the original sub data sets corresponding to the alternative correlation coefficients are respectively determined, and each original sub data set is used as an alternative inherent parameter set.
A3, for each alternative intrinsic parameter set, performing frequency domain transformation on the intrinsic parameters in the alternative intrinsic parameter set to obtain intrinsic frequency domain parameters.
In this embodiment, the specific frequency domain parameter may be specifically understood as a frequency domain signal corresponding to the specific parameter. And performing frequency domain transformation on the intrinsic parameters in each alternative intrinsic parameter set, and transforming the time domain signals into frequency domain signals to obtain the intrinsic frequency domain parameters corresponding to each intrinsic parameter. When frequency domain transformation is carried out, a plurality of corresponding inherent frequency domain parameters can be obtained from one inherent parameter in a mode of changing a phase angle, the contrast quantity between the inherent parameter and the vibration parameter is increased, and the model estimation precision is improved.
A4, obtaining at least one training intrinsic parameter by performing sliding time averaging on the intrinsic frequency domain parameters.
Performing sliding time averaging on each inherent frequency domain parameter, for example, taking a time window with the length of 10, and taking the average value of the 1 st to 10 th inherent frequency domain parameters as a first training inherent parameter; the average value of the 2 nd to 11 th inherent frequency domain parameters is taken as the second training inherent parameter … until the last inherent frequency domain parameter is calculated, resulting in the last training inherent parameter. After the sliding time averaging processing is carried out, the purpose of noise reduction can be achieved, signal noise is reduced, and the accuracy of model estimation is improved.
A5, forming a training intrinsic parameter set corresponding to the alternative intrinsic parameter set based on each training intrinsic parameter.
And taking each training intrinsic parameter as a training intrinsic parameter set corresponding to the alternative intrinsic parameter set, and obtaining the training intrinsic parameter set corresponding to each alternative intrinsic parameter set in the same way.
S220, training the sub-network model to be trained in the given vibration estimation network model to be trained according to the training intrinsic parameters and the corresponding standard vibration parameters, and determining at least one target sub-network model.
In this embodiment, the sub-network model to be trained may be specifically understood as an untrained deep learning based neural network model, and the target vibration estimation network model is an integrated model including at least one sub-network model to be trained. The target subnetwork model can be specifically understood as a network model meeting a convergence condition obtained by training the to-be-trained subnetwork model, the number of the target subnetwork models is the same as that of the to-be-trained subnetwork models, and after one to-be-trained subnetwork model is trained, the target subnetwork model is obtained.
Specifically, each training intrinsic parameter and the corresponding standard vibration parameter are used as a group of input data to be input into the sub-network model to be trained to obtain an estimation result, then a loss function is calculated to perform back propagation, and the next group of data is input until the convergence condition of the model is met to obtain the target sub-network model. The training process of each sub-network model to be trained is the same, and the training intrinsic parameters and the corresponding standard vibration parameters are used as input data for training.
As an optional embodiment of this embodiment, this optional embodiment further trains the sub-network model to be trained in the given vibration estimation network model to be trained according to each of the training intrinsic parameters and the corresponding standard vibration parameters, and determines that at least one target sub-network model is optimized as B1-B3:
and B1, aiming at each sub-network model to be trained, inputting the training intrinsic parameters under the current iteration into the sub-network model to be trained to obtain the estimated vibration parameters.
In this embodiment, the estimated vibration parameters may be specifically understood as the vibration parameters estimated by the sub-network model to be trained in the training process. Firstly, a training intrinsic parameter is selected and input into a to-be-trained sub-network model, and the to-be-trained sub-network model estimates a vibration parameter according to the current parameter of the model to obtain an estimated vibration parameter.
And B2, obtaining a model loss function by adopting a given model loss function expression and combining the estimated vibration parameters and the corresponding standard vibration parameters.
In this embodiment, the model loss function expression may be specifically understood as a loss function expression of the sub-network model to be trained. The model loss function can be specifically understood as the calculation that the estimated vibration parameters and the corresponding standard vibration parameters under the current iteration are brought into the model loss function expression to obtain the model loss function.
B3, performing back propagation on the sub-network model to be trained based on the model loss function to obtain the sub-network model to be trained for the next iteration until the iteration convergence condition is met to obtain the target sub-network model.
And in the training process of the neural network model, continuously updating the parameters of the adjustment model by a back propagation method until the output of the model is consistent with the target, and determining the parameters of the sub-network model to be trained as the parameters of the target sub-network model. After the model loss function is determined, the model loss function is used for carrying out back propagation on the sub-network model to be trained until the sub-network model to be trained meeting the convergence condition is obtained, and the sub-network model to be trained is determined as the target sub-network model. The embodiment of the invention does not limit the specific back propagation process and can be set according to specific conditions.
For example, the sub-network model to be trained in the embodiment of the present application is exemplified by a linear regression model, a support vector regression model, and an artificial neural network model. Training a linear regression model by adopting a linear regression function in a skleran function library in python; training a support vector regression model by adopting an SVR function in a sklern function library in python, and defining initial parameters as a kernel function type of a radial basis function, a penalty coefficient C =100, a tolerance error limit tol =0.1, hyper-parameters gamma =0.1, ϵ =0.5 of the radial basis function and a maximum iteration number of 1000. Training an artificial neural network model by adopting an MLPRegressor function in a sklern function library in python, defining initial parameters as an activation function type as a Sigmoid function, the number of hidden layers of the neural network as a single layer, the number of neurons in the hidden layers of the neural network as 100, a network learning rate eta =100 and the maximum iteration number as 10.
And S230, inputting the training intrinsic parameters under the current iteration into each target sub-network model to obtain corresponding estimated vibration parameters.
In this embodiment, estimating the vibration parameters may be specifically understood as a result of estimating the vibration parameters through the target sub-network model. And after the training of the target sub-network model is finished, inputting the training intrinsic parameters into the target sub-network model for estimation to obtain the estimated vibration parameters correspondingly output by each target sub-network model.
S240, inputting each estimated vibration parameter as input data into an integrated estimation sub-network model in the vibration estimation network model to be trained to obtain a target estimation parameter.
In this embodiment, the integrated estimation sub-network model may be specifically understood as a neural network model for performing comprehensive processing on each estimated vibration parameter, and parameters used in the processing process may be adjusted according to experience; the target estimation parameters can be specifically understood as the final estimation result of the vibration parameters by the vibration estimation network model to be trained.
Specifically, each estimated vibration parameter is input to an integrated estimation sub-network model in the vibration estimation network model to be trained as input data, and the integrated estimation sub-network model performs comprehensive operation, for example, calculation in a weighted summation manner, according to each input estimated vibration parameter to obtain a target estimation parameter.
As an optional embodiment of this embodiment, in this optional embodiment, the estimated vibration parameters are further input as input data to an integrated estimation sub-network model in the to-be-trained vibration estimation network model, and the obtained target estimation parameters are optimized as follows: and performing data processing on each estimated vibration parameter according to the current model parameter in the integrated estimation sub-network model to obtain the target estimation parameter.
In the present embodiment, the current model parameter may be specifically understood as a model parameter of the integrated estimation sub-network model at the current iteration, for example, a weight of each estimated vibration parameter.
And processing each estimated vibration parameter according to the current model parameter in the integrated estimation sub-network model, for example, calculating a weighted value according to the corresponding weighted value, and taking the calculation result as a target estimation parameter.
And S250, obtaining a training loss function by adopting a given training loss function expression and combining the target estimation parameter and the corresponding standard vibration parameter.
In this embodiment, the training loss function expression may be specifically understood as a loss function expression of the integrated estimation subnetwork model. The training loss function may be specifically understood as a loss function calculated according to a target estimation parameter and a standard vibration parameter in a current iteration. And substituting the target estimation parameters and the corresponding standard vibration parameters obtained under the current iteration into a training loss function expression for calculation to obtain a training loss function.
And S260, performing back propagation on the integrated estimation sub-network model based on the training loss function to obtain the integrated estimation sub-network model for the next iteration until an iteration convergence condition is met, and obtaining the target vibration estimation network model.
And in the training process of the neural network model, continuously updating parameters of the adjustment model by a back propagation method until the output of the model is consistent with the target, and determining the parameters of the integrated estimation sub-network model as the parameters of the final integrated estimation sub-network model in the target vibration estimation network model. After the training loss function is determined, the integrated estimation sub-network model is propagated reversely through the training loss function until the integrated estimation sub-network model meeting the convergence condition is obtained. The integrated estimation sub-network model and each target sub-network model at this time constitute a target vibration estimation network model. The embodiment of the invention does not limit the specific back propagation process and can be set according to specific conditions.
After the model is trained by the method of S210-S260 to obtain the target vibration estimation network model, the target vibration estimation network model can estimate vibration parameters in practical application. In practical applications, the target vibration parameters may be estimated according to the structure intrinsic parameters, and the target vibration parameters of the structure to be measured may be estimated by performing the following steps S270-S290. The estimated target vibration parameters can provide technical support for the operation reliability and safety of various devices.
S270, acquiring target frequency domain parameters corresponding to the target intrinsic parameters of the structure to be detected.
And S280, inputting the target frequency domain parameters into a predetermined target vibration estimation network model as input data.
And S290, performing time domain transformation on the output result of the target vibration estimation network model to obtain a target vibration parameter.
The target vibration estimation network model outputs a frequency domain value, so that time domain transformation is carried out on the frequency domain value to obtain a time domain signal serving as a target vibration parameter, and the estimation of the vibration parameter of the structure to be measured is realized.
Fig. 3 is a diagram illustrating an implementation example of a vibration estimation method using structure intrinsic parameters according to an embodiment of the present invention.
And S1, selecting the data files in the folder.
A large number of data files are usually stored in the folder, and the data file to be entered, i.e. the data file containing the required intrinsic parameters, is selected from the folder.
S2, judging whether the variable name is accurate, if so, executing S3; otherwise, execution returns to S1.
And judging whether the variable name in the data file is the type of the required inherent parameter, wherein the step aims to ensure the accuracy of the input data and avoid the data error input. Because the data volume in the data file is usually very large, if the data is recorded incorrectly, the data needs to be recorded again, which wastes time and reduces the working efficiency. Therefore, whether the variable name is accurate or not is judged, if so, subsequent data entry can be carried out, and model training is started; if not, it is necessary to return to S1 to reselect the data file.
And S3, recording variable names.
And S4, matching the time stamps and unifying the units.
The time stamp units of the data are unified, and then the first time stamp is matched and aligned.
S5, judging whether undersampling is needed, if yes, executing S6; otherwise, S7 is executed.
S6, data is undersampled and logged, and S8 is executed.
And S7, data entry.
The data in the data file may have an excessive data volume, and in this case, when the model training is performed, such a large data volume is not required, and therefore, the data is extracted by means of undersampling and then data entry is performed.
And S8, aligning data.
By adjusting the amount of data to be the same, the data is then aligned in order.
And S9, analyzing the characteristic correlation and selecting the characteristics.
And performing characteristic correlation analysis on the intrinsic parameters and the vibration parameters, and selecting the intrinsic parameters with high correlation to form an alternative intrinsic parameter set.
And S10, transforming the time domain signal into a frequency domain signal through fast Fourier transform.
And performing frequency domain transformation on the intrinsic parameters in the alternative intrinsic parameter set through Fast Fourier Transform (FFT) to obtain frequency domain signals, namely the intrinsic frequency domain parameters.
And S11, carrying out moving time average on the frequency domain signals.
And carrying out sliding time averaging to obtain inherent training parameters for subsequent model training.
And S12, training a linear regression model.
And S13, training a support vector regression model.
And S14, training an artificial neural network model.
It should be noted that S12, S13 and S14 are not strictly sequential in execution, and may be performed simultaneously or sequentially. And carrying out model training according to the training intrinsic parameters and the corresponding standard vibration parameters, wherein the obtained model is the model of each target sub-network.
And S15, training an integrated estimation sub-network model.
After the target subnetwork model and the training are finished, the input intrinsic parameters can independently estimate the vibration parameters to obtain respective estimation results. And taking the three different estimation results as the input of the integrated estimation sub-network model, and carrying out weighted summation on the three estimation results by the integrated estimation sub-network model to obtain the final estimation result. The purpose of training the integrated estimation sub-network model is to determine appropriate weight parameters for each target sub-network model and improve the accuracy of the estimation result.
And S16, model estimation, wherein the estimated frequency domain signal is transformed into a time domain signal through inverse FFT.
After model training is completed, the trained models can be used for estimation. The model in this step is a trained target vibration estimation network model, which is composed of each target sub-network model and an integrated estimation sub-network model. And inputting the target frequency domain parameters corresponding to the target intrinsic parameters of the structure to be detected into the target vibration estimation network model to obtain output frequency domain signals. And carrying out inverse FFT (fast Fourier transform) on the signal to obtain a corresponding time domain signal, namely the estimated target vibration parameter.
The embodiment of the invention provides a vibration estimation method by utilizing structure intrinsic parameters, which comprises the steps of obtaining target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected; inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters; and estimating the vibration parameters of the structure to be measured according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be measured, so that the problem that the estimation result is inaccurate because the estimation can be performed only according to the historical vibration parameters in the conventional vibration parameter estimation method is solved. The method analyzes the correlation between the inherent parameters of the structure and the vibration, increases the data contrast by time-frequency domain transformation, further improves the estimation precision, reduces noise of the processed data by using a sliding time averaging method, ensures the validity of the input data during model training, and comprehensively processes the estimation results of each model based on a hybrid estimation method of ensemble learning to obtain the final estimation result. When the estimation results are comprehensively processed, parameters used in the processing process are determined in a model training mode, so that the processing results are more accurate, the structural vibration of the equipment can be effectively estimated, and technical support is provided for improving the operation reliability and safety of various types of equipment.
EXAMPLE III
Fig. 4 is a schematic structural diagram of a vibration estimation apparatus using intrinsic parameters of a structure according to a third embodiment of the present invention, where the apparatus includes: a parameter acquisition module 31, an input module 32 and an estimation module 33.
The parameter acquiring module 31 is configured to acquire a target frequency domain parameter corresponding to a target intrinsic parameter of the structure to be detected; an input module 32, configured to input the target frequency domain parameter as input data into a predetermined target vibration estimation network model, where the target vibration estimation network model is obtained by training intrinsic parameters; and the estimation module 33 is configured to estimate the vibration parameter of the structure to be detected according to an output result of the target vibration estimation network model, so as to obtain the target vibration parameter of the structure to be detected.
The embodiment of the invention provides a vibration estimation device utilizing structure intrinsic parameters, which is characterized in that target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected are obtained; inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters; and estimating the vibration parameters of the structure to be measured according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be measured, solving the problem that the estimation result is inaccurate because the estimation can be carried out only according to historical vibration parameters in the existing vibration parameter estimation method, and estimating the vibration parameters by using the inherent parameters of the structure. The method has the advantages that the external environment is simulated through the estimation parameters, the vibration parameter estimation of the structure to be measured can be effectively realized, the accuracy of the vibration parameter estimation is improved, the vibration parameters of the structure under any environment can be simulated, the method is easy to realize, the applicable scene is wide, and the vibration estimation can still be carried out under the condition that the vibration parameters cannot be collected.
Further, the apparatus further comprises:
the parameter set acquisition module is used for acquiring a standard vibration parameter set and determining at least one training intrinsic parameter set associated with the standard vibration parameter set, wherein the standard vibration parameter set comprises at least one standard vibration parameter, and the training intrinsic parameter set comprises at least one training intrinsic parameter;
the sub-model training module is used for training a sub-network model to be trained in a given vibration estimation network model to be trained according to each training intrinsic parameter and the corresponding standard vibration parameter to determine at least one target sub-network model;
the vibration parameter estimation module is used for inputting the training intrinsic parameters under the current iteration into each target sub-network model to obtain corresponding estimated vibration parameters;
the target parameter estimation module is used for inputting each estimated vibration parameter as input data into an integrated estimation sub-network model in the vibration estimation network model to be trained to obtain a target estimation parameter;
the loss function calculation module is used for acquiring a training loss function by adopting a given training loss function expression and combining the target estimation parameter and the corresponding standard vibration parameter;
and the target model determining module is used for carrying out back propagation on the integrated estimation sub-network model based on the training loss function to obtain the integrated estimation sub-network model for the next iteration until an iteration convergence condition is met to obtain the target vibration estimation network model.
Further, the parameter set obtaining module includes:
the data acquisition unit is used for acquiring a data file, processing the data file according to predetermined file attribute characteristics and determining an original data set, wherein the original data set comprises at least one original sub data set;
a screening unit for screening at least one alternative intrinsic parameter set associated with the standard vibration parameter set from the original data set;
a frequency domain transformation unit, configured to perform frequency domain transformation on the intrinsic parameters in the candidate intrinsic parameter sets to obtain intrinsic frequency domain parameters, for each candidate intrinsic parameter set;
a parameter averaging unit, configured to perform sliding time averaging on each inherent frequency domain parameter to obtain at least one training inherent parameter;
a parameter set forming unit, configured to form, based on each of the training intrinsic parameters, a training intrinsic parameter set corresponding to the candidate intrinsic parameter set.
Further, the screening unit is specifically configured to align the data in the original data set and the standard vibration parameter set; determining the aligned standard vibration parameter set and the correlation coefficient of each original subdata set; screening out alternative correlation coefficients meeting preset requirements from all the correlation coefficients; and taking each original sub data set corresponding to each alternative correlation coefficient as an alternative inherent parameter set.
Further, the implementation of aligning the data in the raw data set and the standard vibration parameter set may be embodied as: unifying the time stamp units of the original data set and the data in the standard vibration parameter set; carrying out timestamp matching on the original data set and the data in the standard vibration parameter set; and adjusting the data volume of each matched original sub data set and standard vibration parameter set to be the same, and aligning the data in sequence.
Further, the sub-model training module comprises:
the estimated parameter determining unit is used for inputting training intrinsic parameters under current iteration into the sub-network model to be trained aiming at each sub-network model to be trained to obtain estimated vibration parameters;
the loss function determining unit is used for obtaining a model loss function by adopting a given model loss function expression and combining the estimated vibration parameters and the corresponding standard vibration parameters;
and the sub-network model determining unit is used for carrying out back propagation on the sub-network model to be trained on the basis of the model loss function to obtain the sub-network model to be trained for the next iteration until an iteration convergence condition is met to obtain the target sub-network model.
Further, the target parameter estimation module is specifically configured to perform data processing on each estimated vibration parameter according to a current model parameter in the integrated estimation sub-network model to obtain the target estimation parameter.
Further, the estimation module 33 is specifically configured to perform time domain transformation on the output result of the target vibration estimation network model to obtain a target vibration parameter.
The vibration estimation device using the structure intrinsic parameters provided by the embodiment of the invention can execute the vibration estimation method using the structure intrinsic parameters provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.
Example four
Fig. 5 is a schematic structural diagram of a computer apparatus according to a fourth embodiment of the present invention, as shown in fig. 5, the apparatus includes a processor 40, a memory 41, an input device 42, and an output device 43; the number of processors 40 in the device may be one or more, and one processor 40 is taken as an example in fig. 5; the processor 40, the memory 41, the input device 42 and the output device 43 in the apparatus may be connected by a bus or other means, which is exemplified in fig. 5.
The memory 41 serves as a computer-readable storage medium for storing software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the vibration estimation method using the structure intrinsic parameters in the embodiment of the present invention (for example, the parameter acquisition module 31, the input module 32, and the estimation module 33 in the vibration estimation apparatus using the structure intrinsic parameters). The processor 40 executes various functional applications of the device and data processing, i.e., implements the above-described vibration estimation method using the structural intrinsic parameters, by executing software programs, instructions, and modules stored in the memory 41.
The memory 41 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 41 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, memory 41 may further include memory located remotely from processor 40, which may be connected to the device over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 42 is operable to receive input numeric or character information and to generate key signal inputs relating to user settings and function controls of the apparatus. The output device 43 may include a display device such as a display screen.
EXAMPLE five
An embodiment of the present invention further provides a storage medium containing computer-executable instructions, which when executed by a computer processor, perform a method for vibration estimation using structure intrinsic parameters, the method including:
acquiring target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected;
inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters;
and estimating the vibration parameters of the structure to be detected according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be detected.
Of course, the storage medium containing the computer-executable instructions provided by the embodiments of the present invention is not limited to the method operations described above, and may also perform related operations in the vibration estimation method using the structure intrinsic parameters provided by any embodiments of the present invention.
From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which can be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.
It should be noted that, in the embodiment of the vibration estimation apparatus using the structural intrinsic parameters, the included units and modules are only divided according to the functional logic, but are not limited to the above division as long as the corresponding functions can be realized; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (9)
1. A vibration estimation method using a structure intrinsic parameter, comprising:
acquiring target frequency domain parameters corresponding to target intrinsic parameters of a structure to be detected;
inputting the target frequency domain parameters serving as input data into a predetermined target vibration estimation network model, wherein the target vibration estimation network model is obtained by training intrinsic parameters;
estimating the vibration parameters of the structure to be detected according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be detected; the training step of the target vibration estimation network model comprises the following steps:
acquiring a standard vibration parameter set, and determining at least one training intrinsic parameter set associated with the standard vibration parameter set, wherein the standard vibration parameter set comprises at least one standard vibration parameter, and the training intrinsic parameter set comprises at least one training intrinsic parameter;
training the sub-network model to be trained in the given vibration estimation network model to be trained according to each training intrinsic parameter and the corresponding standard vibration parameter to determine at least one target sub-network model;
inputting training intrinsic parameters under current iteration into each target sub-network model to obtain corresponding estimated vibration parameters;
inputting each estimated vibration parameter as input data into an integrated estimation sub-network model in the vibration estimation network model to be trained to obtain a target estimation parameter;
obtaining a training loss function by adopting a given training loss function expression and combining the target estimation parameter and the corresponding standard vibration parameter;
carrying out back propagation on the integrated estimation sub-network model based on the training loss function to obtain an integrated estimation sub-network model for the next iteration until an iteration convergence condition is met to obtain a target vibration estimation network model;
the determining at least one training intrinsic parameter set to which the standard set of vibration parameters is associated includes:
acquiring a data file, processing the data file according to predetermined file attribute characteristics, and determining an original data set, wherein the original data set comprises at least one original sub data set;
screening at least one alternative set of intrinsic parameters associated with the set of standard vibration parameters from the raw data set;
for each alternative intrinsic parameter set, performing frequency domain transformation on intrinsic parameters in the alternative intrinsic parameter set to obtain intrinsic frequency domain parameters;
carrying out sliding time averaging on each inherent frequency domain parameter to obtain at least one training inherent parameter;
and forming a training intrinsic parameter set corresponding to the alternative intrinsic parameter set based on each training intrinsic parameter.
2. The method of claim 1, wherein said screening said original data set for an alternative set of intrinsic parameters associated with said standard set of vibration parameters comprises:
aligning the data in the original data set and the standard vibration parameter set;
determining the aligned standard vibration parameter set and the correlation coefficient of each original subdata set;
screening out alternative correlation coefficients meeting preset requirements from all the correlation coefficients;
and taking each original sub data set corresponding to each alternative correlation coefficient as an alternative inherent parameter set.
3. The method of claim 2, wherein said aligning data in said raw data set and a standard vibration parameter set comprises:
unifying the time stamp units of the original data set and the data in the standard vibration parameter set;
carrying out timestamp matching on the original data set and the data in the standard vibration parameter set;
and adjusting the data volume of each matched original sub data set and standard vibration parameter set to be the same, and aligning the data in sequence.
4. The method of claim 1, wherein the training the sub-network model to be trained in the given vibration estimation network model to be trained according to each of the training intrinsic parameters and the corresponding standard vibration parameters to determine at least one target sub-network model comprises:
inputting training intrinsic parameters under current iteration into each sub-network model to be trained to obtain pre-estimated vibration parameters;
obtaining a model loss function by adopting a given model loss function expression and combining the estimated vibration parameters and the corresponding standard vibration parameters;
and performing back propagation on the to-be-trained subnetwork model based on the model loss function to obtain the to-be-trained subnetwork model for the next iteration until an iteration convergence condition is met, and obtaining a target subnetwork model.
5. The method of claim 1, wherein the inputting each estimated vibration parameter as input data into an integrated estimation sub-network model in the vibration estimation network model to be trained to obtain a target estimation parameter comprises:
and performing data processing on each estimated vibration parameter according to the current model parameter in the integrated estimation sub-network model to obtain the target estimation parameter.
6. The method according to claim 1, wherein the estimating the vibration parameter of the structure to be tested according to the output result of the target vibration estimation network model to obtain the target vibration parameter of the structure to be tested comprises:
and performing time domain transformation on the output result of the target vibration estimation network model to obtain target vibration parameters.
7. A vibration estimation apparatus using a structure intrinsic parameter, comprising:
the parameter acquisition module is used for acquiring target frequency domain parameters corresponding to target intrinsic parameters of the structure to be detected;
the input module is used for inputting the target frequency domain parameters into a predetermined target vibration estimation network model as input data, wherein the target vibration estimation network model is obtained by training intrinsic parameters;
the estimation module is used for estimating the vibration parameters of the structure to be detected according to the output result of the target vibration estimation network model to obtain the target vibration parameters of the structure to be detected;
the parameter set acquisition module is used for acquiring a standard vibration parameter set and determining at least one training intrinsic parameter set associated with the standard vibration parameter set, wherein the standard vibration parameter set comprises at least one standard vibration parameter, and the training intrinsic parameter set comprises at least one training intrinsic parameter;
the sub-model training module is used for training a sub-network model to be trained in a given vibration estimation network model to be trained according to each training intrinsic parameter and the corresponding standard vibration parameter to determine at least one target sub-network model;
the vibration parameter estimation module is used for inputting the training intrinsic parameters under the current iteration into each target sub-network model to obtain corresponding estimated vibration parameters;
the target parameter estimation module is used for inputting each estimated vibration parameter as input data into an integrated estimation sub-network model in the vibration estimation network model to be trained to obtain a target estimation parameter;
the loss function calculation module is used for acquiring a training loss function by adopting a given training loss function expression and combining the target estimation parameter and the corresponding standard vibration parameter;
the target model determining module is used for carrying out back propagation on the integrated estimation sub-network model based on the training loss function to obtain an integrated estimation sub-network model for the next iteration until an iteration convergence condition is met to obtain a target vibration estimation network model;
the parameter set obtaining module comprises:
the data acquisition unit is used for acquiring a data file, processing the data file according to predetermined file attribute characteristics and determining an original data set, wherein the original data set comprises at least one original sub data set;
a screening unit for screening at least one alternative intrinsic parameter set associated with the standard vibration parameter set from the original data set;
a frequency domain transformation unit, configured to perform frequency domain transformation on the intrinsic parameters in the candidate intrinsic parameter sets to obtain intrinsic frequency domain parameters, for each candidate intrinsic parameter set;
a parameter averaging unit, configured to perform sliding time averaging on each inherent frequency domain parameter to obtain at least one training inherent parameter;
a parameter set forming unit, configured to form, based on each of the training intrinsic parameters, a training intrinsic parameter set corresponding to the candidate intrinsic parameter set.
8. A computer device, the device comprising:
one or more processors;
a memory for storing one or more programs,
when executed by the one or more processors, cause the one or more processors to implement the vibration estimation method using the structure-inherent parameter as recited in any one of claims 1 to 6.
9. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out a vibration estimation method using structure-inherent parameters according to any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111026053.7A CN113468771B (en) | 2021-09-02 | 2021-09-02 | Vibration estimation method using structure intrinsic parameters |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111026053.7A CN113468771B (en) | 2021-09-02 | 2021-09-02 | Vibration estimation method using structure intrinsic parameters |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113468771A CN113468771A (en) | 2021-10-01 |
| CN113468771B true CN113468771B (en) | 2021-12-14 |
Family
ID=77867437
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111026053.7A Active CN113468771B (en) | 2021-09-02 | 2021-09-02 | Vibration estimation method using structure intrinsic parameters |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113468771B (en) |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP4277915B2 (en) * | 2007-04-03 | 2009-06-10 | 株式会社デンソー | Vehicle control device |
| EP2610605B1 (en) * | 2010-08-26 | 2023-04-12 | Nissan Motor Co., Ltd. | Device for estimating vehicle body vibration and controller for suppressing vehicle body vibration using same |
| US10239635B2 (en) * | 2017-06-08 | 2019-03-26 | The Boeing Company | Methods for balancing aircraft engines based on flight data |
-
2021
- 2021-09-02 CN CN202111026053.7A patent/CN113468771B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN113468771A (en) | 2021-10-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110675623B (en) | Short-term traffic flow prediction method, system and device based on hybrid deep learning | |
| CN108091166B (en) | Method, device, equipment and storage medium for predicting number change of available parking spaces | |
| CN114297036B (en) | Data processing method, device, electronic equipment and readable storage medium | |
| CN116431975A (en) | Environment monitoring method and system for data center | |
| CN108879732A (en) | Transient stability evaluation in power system method and device | |
| CN113465734B (en) | Real-time estimation method for structural vibration | |
| CN111797861A (en) | Information processing method, information processing apparatus, storage medium, and electronic device | |
| CN110634060A (en) | User credit risk assessment method, system, device and storage medium | |
| CN108053424A (en) | Method for tracking target, device, electronic equipment and storage medium | |
| CN114290960A (en) | Method and device for acquiring battery health degree of power battery and vehicle | |
| CN112712214A (en) | Method, system, device and storage medium for predicting track of maritime search and rescue object | |
| CN111738319A (en) | Clustering result evaluation method and device based on large-scale samples | |
| CN112711032B (en) | Radar target detection method and system based on graph data and GCN | |
| CN113468771B (en) | Vibration estimation method using structure intrinsic parameters | |
| CN116647405B (en) | Multipath QUIC abnormal flow detection method, system and equipment | |
| CN111325255B (en) | Specific crowd delineating method and device, electronic equipment and storage medium | |
| CN116885711A (en) | Wind power prediction method, device, equipment and readable storage medium | |
| CN115170304B (en) | Method and device for extracting risk feature description | |
| CN116451081A (en) | Data drift detection method, device, terminal and storage medium | |
| Wang et al. | Temperature forecast based on SVM optimized by PSO algorithm | |
| WO2023050649A1 (en) | Esg index determination method based on data complementing, and related product | |
| CN112463964B (en) | Text classification and model training method, device, equipment and storage medium | |
| CN115222018A (en) | Non-invasive load decomposition method and device | |
| CN114282657A (en) | Market data long-term prediction model training method, device, equipment and storage medium | |
| CN114662832A (en) | Population mobility analysis method and device and electronic equipment |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20230626 Address after: Room 409, 4th Floor, Building C, No. 2 Cuizhu Second Street, Qianshan, Xiangzhou District, Zhuhai City, Guangdong Province, 519070 Patentee after: Zhiwei Technology (Zhuhai) Co.,Ltd. Address before: 30 Shuangqing Road, Haidian District, Beijing 100091 Patentee before: TSINGHUA University |