Data-driven operation condition transmission path analysis method

Abstract

The invention provides a data-driven operation condition transmission path analysis method, which comprises the following steps of: obtaining vibration signals of an indication point and a target point of a target mechanical system, and obtaining a predicted optimal transmissibility through calculation; constructing a target function, substituting the predicted optimal transmissibility, and obtaining a final data-driven transmissibility through iterative optimization; and on the basis of the data-driven transmissibility, predicting an actual target point signal, and evaluating the vibration contribution of each operation condition transmission path to the actual target point signal. According to the method, the defect that the analysis precision is reduced due to blind selection of target test data is overcome, and relatively high analysis precision can be achieved.

Description

Technical Field

[0001] The invention belongs to the field of vibration and noise analysis, and in particular relates to a data-driven operating condition transfer path analysis method. Background Art

[0002] Transfer path analysis (TPA) is a key technology for evaluating the vibration and noise transfer characteristics of mechanical systems through experiments. Operational transfer path analysis (OTPA) does not require disassembly of the system, so it is more time-saving, labor-saving, and easy to use than traditional TPA technology. In engineering practice, the overall test data of OTPA is usually generated after the test data has undergone time-frequency conversion and crosstalk reduction processing, so it inevitably contains errors. Technicians usually select target test data based on personal experience and subjective judgment and directly perform OTPA without considering the impact of the target test data itself on the accuracy of OTPA. In other words, the accuracy of OTPA is different for different selected target test data, and blind selection often leads to low accuracy of OTPA. Therefore, there is an urgent need for a new OTPA technology that can consider the impact of target test data on OTPA accuracy, and then find the relatively optimal target test data for OTPA to achieve higher analysis accuracy.

[0003] The above information disclosed in this Background section is only for enhancement of understanding of the background of the invention and therefore it may contain information that does not form the prior art that is already known to a person of ordinary skill in the art. Summary of the invention

[0004] The purpose of the present invention is to provide a data-driven operating condition transfer path analysis method, which overcomes the defect of blindly selecting target test data leading to reduced analysis accuracy, considers the statistical distribution characteristics of the error of the target test data, gradually selects the relatively optimal target test data through iterative optimization, and finally achieves higher analysis accuracy. In order to achieve the above purpose, the present invention provides the following technical solutions:

[0005] A data-driven operating condition transfer path analysis method, specifically comprising:

[0006] In the first step, the vibration signals of the indicator point and the target point of the target mechanical system are obtained, and the predicted optimal transmission rate is obtained by calculation;

[0007] The second step is to construct an objective function, substitute the predicted optimal delivery rate, and obtain the final data-driven delivery rate through iterative optimization;

[0008] In the third step, based on the data-driven transfer rate obtained in the second step, the actual target point signal can be predicted, and the vibration contribution of each operating condition transfer path to the actual target point signal can be evaluated.

[0009] Preferably, in the first step, the optimal transmission rate is predicted by calculation, and the specific method includes:

[0010] dividing the vibration signal into actual data and overall test data;

[0011] Select target test data from the overall test data and obtain the transmission rate through calculation;

[0012] The optimal transmission rate is obtained by minimizing the prediction error of the actual target point signal.

[0013] Preferably, the vibration signal is divided into actual data and overall test data, and the specific method is as follows:

[0014] For the target mechanical system, the vibration signals of n indication points and 1 target point are obtained. After time-frequency conversion and crosstalk reduction processing, a total of M groups of data are obtained. [ X ˜ , Y ˜ ] ( Represents the indicator point data matrix, with dimensions , Represents the target point data matrix, with dimension , both only represent data at a certain frequency point), through The norm of is maximized, which can be obtained from [ X ˜ , Y ˜ ] Select 1 set of actual data [ X ˜ P , Y ˜ P ] :

[0015] [ X ˜ P , Y ˜ P ] = a r g m a x X ˜ P ( ∑ i = 1 N f ‖ X ˜ P ‖ i ) ,

[0016] in represents the Frobenius norm, is the number of frequency points, and the remaining data is Overall test data of the group [ X ˜ E T , Y ˜ E T ] .

[0017] Preferably, the transfer rate is obtained by calculation, and the specific method is as follows:

[0018] from Overall test data of the group [ X ˜ E T , Y ˜ E T ] Select m ( ) group, as the target test data [ X ˜ E , Y ˜ E ] , the number of all combinations is as follows:

[0019] ,

[0020] The transmissibility was calculated by total least squares method. ( The dimension is ):

[0021] .

[0022] Preferably, the method for obtaining the predicted optimal transmission rate is as follows:

[0023] By minimizing the prediction error of the actual target point signal, the optimal transmission rate is obtained :

[0024] .

[0025] Preferably, in the second step, constructing the objective function includes: constructing an objective function about the target test data and the transmission rate:

[0026]

[0027] The gamma function ( ), condition number is a matrix The ratio of the largest singular value to the smallest singular value of .

[0028] Preferably, in the second step, the iterative optimization is performed to obtain the final data-driven transfer rate, and the specific method is as follows:

[0029] Substitute the predicted optimal transmission rate obtained in the first step into the objective function, and obtain the first target test data by minimizing the objective function:

[0030] [ X ˜ E 1 , Y ˜ E 1 ] =  a r g m i n X ˜ E , Y ˜ E Φ ( X ˜ E , Y ˜ E , m , T ˜ P ) ,

[0031] And the first transmission rate is calculated by the total least squares method:

[0032] ;

[0033] Then substitute the first transmission rate into the objective function, and obtain the selected second target test data by minimizing the objective function:

[0034] [ X ˜ E 2 , Y ˜ E 2 ] = a r g m i n X ˜ E , Y ˜ E Φ ( X ˜ E , Y ˜ E , m , T ˜ 1 T L S ) ,

[0035] And the second transmission rate is calculated by the total least squares method:

[0036] ;

[0037] By analogy, repeat the above steps and calculate the k-1th transmission rate by the total least squares method:

[0038] ,

[0039] Select the kth target test data:

[0040] [ X ˜ E k , Y ˜ E k ] = a r g m i n X ˜ E , Y ˜ E Φ ( X ˜ E , Y ˜ E , m , T ˜ ( k − 1 ) T L S ) ,

[0041] The iteration stops until the target test data selected for the Kth time is the same as the previous one:

[0042] [ X ˜ E K , Y ˜ E K ] = [ X ˜ E ( K − 1 ) , Y ˜ E ( K − 1 ) ] ,

[0043] The final data-driven transfer rate is calculated by the total least squares method:

[0044] .

[0045] Preferably, in the third step, the predicted actual target point signal is determined according to the following formula:

[0046] Prediction value of actual target point signal .

[0047] Preferably, in the third step, the vibration contribution of each operating condition transmission path to the actual target point signal is determined according to the following formula:

[0048] ,

[0049] in, is the vibration contribution of the ith path, ; is a matrix The i-th element of ; is a matrix The ith element of .

[0050] In the above technical solution, the data-driven operating condition transfer path analysis method provided by the present invention has the following beneficial effects: the selection of target test data in the traditional operating condition transfer path analysis method depends on the personal experience and subjective judgment of the technicians, which is usually blind, so this selection is often inappropriate, resulting in lower accuracy. The data-driven operating condition transfer path analysis method proposed by the present invention emphasizes "learning from data", can consider the statistical distribution characteristics of the errors of the target test data, gradually find the relatively optimal target test data, and finally calculate a more accurate transfer rate, which achieves a significant improvement in analysis accuracy compared to the traditional method. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can also be obtained based on these drawings.

[0052] Figure 1 An algorithm flow chart of an embodiment of a data-driven operating condition transfer path analysis method of the present invention;

[0053] Figure 2 A schematic diagram of a multi-path simulation model of an embodiment of a data-driven operating condition transfer path analysis method of the present invention;

[0054] Figure 3 A diagram showing an error comparison between the transmissibility obtained by an embodiment of the data-driven operating condition transfer path analysis method of the present invention and the results of other methods;

[0055] Figure 4 A curve diagram of actual target point signal prediction values ​​of an embodiment of a data-driven operating condition transfer path analysis method of the present invention;

[0056] Figure 5 This is a graph of the vibration contribution of the sixth transfer path to the actual target point signal in an embodiment of the data-driven operating condition transfer path analysis method of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] Therefore, the following detailed description of the embodiments of the present invention provided in the accompanying drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] It should be noted that similar reference numerals and letters denote similar items in the following drawings, and therefore, once an item is defined in one drawing, further definition and explanation thereof is not required in subsequent drawings.

[0060] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the referred device or element must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0061] In addition, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0062] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be fixed connection, detachable connection, or integrated; it can be directly connected, or indirectly connected through an intermediate medium, it can be the internal communication of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances. In the present invention, unless otherwise clearly specified and limited, the first feature "above" or "below" the second feature can include the first and second features directly contacting, and can also include the first and second features not directly contacting but contacting through another feature between them. Moreover, the first feature "above", "above" and "above" the second feature include the first feature directly above and obliquely above the second feature, or simply means that the first feature is higher in level than the second feature. The first feature "below", "below" and "below" the second feature include the first feature directly below and obliquely below the second feature, or simply means that the first feature is lower in level than the second feature.

[0063] In order to enable those skilled in the art to better understand the technical solution of the present invention, Figures 1 to 5 The present invention is further described in detail.

[0064] The present invention proposes a data-driven operating condition transfer path analysis method, the steps comprising:

[0065] In the first step, the vibration signals of the indicator point and the target point of the target mechanical system are obtained, and the predicted optimal transmission rate is obtained by calculation;

[0066] The second step is to construct an objective function, substitute the predicted optimal delivery rate, and obtain the final data-driven delivery rate through iterative optimization;

[0067] In the third step, based on the data-driven transfer rate obtained in the second step, the actual target point signal can be predicted, and the vibration contribution of each operating condition transfer path to the actual target point signal can be evaluated.

[0068] In one embodiment, the first step further includes: obtaining vibration signals of the indicator point and the target point of the target mechanical system, and dividing them into actual data and overall test data. Selecting the target test data from the overall test data, calculating the transmissibility by the overall least squares method, and then obtaining the predicted optimal transmissibility by minimizing the prediction error of the actual target point signal.

[0069] The second step also includes: constructing an objective function about the target test data and the transmission rate, substituting the predicted optimal transmission rate obtained in the first step, obtaining the selected first target test data by minimizing the objective function, and calculating the first transmission rate by the total least squares method. Substituting the first transmission rate into the objective function, obtaining the selected second target test data by minimizing the objective function, and calculating the second transmission rate by the total least squares method. And so on, repeating the above steps until the selected target test data is the same as the previous one, the iteration stops, and the final data-driven transmission rate is calculated by the total least squares method.

[0070] The third step also includes: predicting the actual target point signal based on the data-driven transmissibility obtained in the second step, and evaluating the vibration contribution of each operating condition transmission path to it.

[0071] The algorithm flow of the above three steps is as follows Figure 1 As shown, the first step is on the left side of the figure, the second step is in the middle of the figure, and the third step is on the right side of the figure.

[0072] In another embodiment, in the first step, for the target mechanical system, the vibration signals of n indicating points and 1 target point are obtained, and after time-frequency conversion and crosstalk reduction processing, a total of M groups of data are obtained. [ X ˜ , Y ˜ ] ( Represents the indicator point data matrix, with dimensions , Represents the target point data matrix, with dimension , which only represent the data at a certain frequency point). The norm of is maximized, which can be obtained from [ X ˜ , Y ˜ ] Select 1 set of actual data [ X ˜ P , Y ˜ P ] :

[0073] [ X ˜ P , Y ˜ P ] = a r g m a x X ˜ P ( ∑ i = 1 N f ‖ X ˜ P ‖ i ) ,

[0074] in represents the Frobenius norm, is the number of frequency points. The remaining data is Overall test data of the group [ X ˜ E T , Y ˜ E T ] .

[0075] From the overall test data [ X ˜ E T , Y ˜ E T ] Select m ( ) group, as the target test data [ X ˜ E , Y ˜ E ] , the number of all combinations is as follows:

[0076] ,

[0077] The transmissibility was calculated by the total least squares method ( The dimension is ):

[0078] ,

[0079] By minimizing the prediction error of the actual target point signal, the optimal predicted transmission rate is obtained:

[0080] .

[0081] In another embodiment, in the second step, an objective function related to the target test data and the transmissibility is constructed:

[0082] ,

[0083] The gamma function ( ), condition number is a matrix Substitute the predicted optimal transmission rate obtained in the first step into the ratio of the maximum singular value to the minimum singular value of , and obtain the first target test data selected by minimizing the objective function:

[0084] [ X ˜ E 1 , Y ˜ E 1 ] =  a r g m i n X ˜ E , Y ˜ E Φ ( X ˜ E , Y ˜ E , m , T ˜ P ) ,

[0085] And the first transmission rate is calculated by the total least squares method:

[0086] .

[0087] In another embodiment, in the second step, the first transmission rate is substituted into the objective function, and the selected second target test data is obtained by minimizing the objective function:

[0088] [ X ˜ E 2 , Y ˜ E 2 ] = a r g m i n X ˜ E , Y ˜ E Φ ( X ˜ E , Y ˜ E , m , T ˜ 1 T L S ) ,

[0089] And the second transmission rate is calculated by the total least squares method:

[0090] .

[0091] By analogy, repeat the above steps and calculate the k-1th transmission rate by the total least squares method:

[0092] ,

[0093] Select the kth target test data:

[0094] [ X ˜ E k , Y ˜ E k ] = a r g m i n X ˜ E , Y ˜ E Φ ( X ˜ E , Y ˜ E , m , T ˜ ( k − 1 ) T L S ) ,

[0095] The iteration stops until the target test data selected for the Kth time is the same as the previous one:

[0096] [ X ˜ E K , Y ˜ E K ] = [ X ˜ E ( K − 1 ) , Y ˜ E ( K − 1 ) ] ,

[0097] The final data-driven transfer rate is calculated by the total least squares method:

[0098] .

[0099] In another embodiment, in the third step, based on the data-driven transfer rate obtained in the second step, the predicted value of the actual target point signal can be calculated:

[0100] ,

[0101] And evaluate the vibration contribution of each operating condition transfer path to the actual target point signal:

[0102] ,

[0103] in, is the vibration contribution of the ith path, , is a matrix The i-th element of is a matrix The ith element of .

[0104] In another embodiment, the vibration data used is generated by a multi-path simulation model. The model is a lumped parameter model composed of The mass blocks are connected to each other through stiffening and damping elements, such as Figure 2 shown. Figure 2 The mass of the mass block kg, of which , the stiffness of the connection , the damping of the connection ,in .make By applying excitation on the active end, the vibration frequency domain data of 6 indicator points and 1 target point on the passive end can be obtained, distributed in After crosstalk reduction and adding normal distribution error, we can get Group data.

[0105] The 13 groups of data are divided into 12 groups of overall test data and 1 group of actual data, and then 6 to 12 groups are selected from the overall test data as the target test data. Note that the target test data of the present invention is selected for each frequency point separately, so there are 2510 combinations at each frequency point ( ). In the first step of the present invention, the predicted optimal transmission rate can be obtained; in the second step of the present invention, the data-driven transmission rate can be obtained. For convenience, technicians usually select target test data for all frequency points at the same time. Similarly, there are 2510 combinations in total, and 2510 transmission rates and their errors can be obtained. The average value of these errors can be regarded as the expectation of the error level caused by blindly selecting target test data. For the variance of the error of target test data of different sizes, the comparison of the above three transmission rate errors is as follows Figure 3 As shown. It can be seen that compared with the average error of the transfer rates of all combinations and the error of the predicted optimal transfer rate, the total error of the data-driven transfer rate in the present invention is reduced by 37.6% and 15.2%, respectively. This shows the superiority of the data-driven operating condition transfer path analysis method in the present invention, that is, it can achieve higher accuracy.

[0106] In the third step of the present invention, the predicted value of the vibration signal of the actual target point can be obtained, such as Figure 4 This prediction is shown in Hz, the amplitude distribution at a total of 200 frequency points; based on Article 6 ( Figure 2 Taking the rightmost path in the figure as an example, the vibration contribution of the path can be obtained, as follows: Figure 5 , which shows that this contribution is Amplitude distribution at a total of 200 frequency points of Hz. Figure 4 and Figure 5 The content reveals the vibration characteristics of the mechanical system and can provide useful information for controlling the vibration of the mechanical system.

[0107] The above is only a description of the data-driven operating condition transfer path analysis method for a multi-path simulation model, but the present invention is not limited to this application scope.

[0108] Finally, it should be noted that the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without making any creative work are within the scope of protection of the present application.

[0109] The above description is only by way of illustration of certain exemplary embodiments of the present invention. It is undoubted that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the above drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A data-driven operating condition transfer path analysis method, characterized in that: It includes the following steps: In the first step, the vibration signals of the indicator point and the target point of the target mechanical system are obtained, and the predicted optimal transmission rate is obtained by calculation; The second step is to construct an objective function, substitute the predicted optimal delivery rate, and obtain the final data-driven delivery rate through iterative optimization; The third step is to predict the actual target point signal based on the data-driven transferability and evaluate the vibration contribution of each operating condition transfer path to the actual target point signal.

2. A data-driven operating condition transfer path analysis method according to claim 1, characterized in that: Preferably, in the first step, the step of obtaining the predicted optimal transmission rate by calculation specifically includes: dividing the vibration signal into actual data and overall test data; Select target test data from the overall test data and obtain the transmission rate through calculation; The optimal transmission rate is obtained by minimizing the prediction error of the actual target point signal.

3. A data-driven operating condition transfer path analysis method according to claim 2, characterized in that: The vibration signal is divided into actual data and overall test data, specifically including: For the target mechanical system, the vibration signals of n indication points and 1 target point are obtained. After time-frequency conversion and crosstalk reduction processing, a total of M groups of data are obtained. , Represents the indicator point data matrix, with dimensions , Represents the target point data matrix, with dimension , both only represent the data at a certain frequency point. The norm of is maximized, which can be obtained from Select 1 set of actual data : ， in represents the Frobenius norm, is the number of frequency points, and the remaining data is Overall test data of the group .

4. A data-driven operating condition transfer path analysis method according to claim 3, characterized in that: The transmission rate is obtained by calculation, specifically including: From the overall test data Select m ( ) group, as the target test data , the number of all combinations is as follows: ， The transmissibility was calculated by total least squares method. ,in The dimension is : 。 5. A data-driven operating condition transfer path analysis method according to claim 4, characterized in that: The step of obtaining the predicted optimal transmission rate specifically includes: By minimizing the prediction error of the actual target point signal, the optimal transmission rate is obtained : 。 6. A data-driven operating condition transfer path analysis method according to claim 5, characterized in that: In the second step, constructing the objective function includes: constructing an objective function related to the target test data and the transmission rate: ， Among them, the gamma function ( ), condition number is a matrix The ratio of the largest singular value to the smallest singular value of .

7. A data-driven operating condition transfer path analysis method according to claim 6, characterized in that: In the second step, the iterative optimization is performed to obtain the final data-driven delivery rate, specifically including: Substitute the predicted optimal transmission rate obtained in the first step into the objective function, and obtain the first target test data by minimizing the objective function: ， And the first transmission rate is calculated by the total least squares method: ； Then substitute the first transmission rate into the objective function, and obtain the selected second target test data by minimizing the objective function: ， And the second transmission rate is calculated by the total least squares method: ； By analogy, repeat the above steps and calculate the k-1th transmission rate by the total least squares method: ， Select the kth target test data: ， The iteration stops until the target test data selected for the Kth time is the same as the previous one: ， The final data-driven transfer rate is calculated by the total least squares method: 。 8. A data-driven operating condition transfer path analysis method according to claim 7, characterized in that: In the third step, the predicted actual target point signal is determined according to the following formula: Prediction value of actual target point signal .

9. A data-driven operating condition transfer path analysis method according to claim 8, characterized in that: In the third step, the vibration contribution of each transmission path to the actual target point signal is determined according to the following formula: ， in, is the vibration contribution of the ith path, ; is a matrix The i-th element of ; is a matrix The ith element of .

Images