Vibration or shock time domain waveform simulation method and system

Abstract

The present application relates to a kind of vibration or impact environment time domain waveform simulation method and system, belong to environmental engineering field, method includes: S1, collection N group vibration or impact time domain data;S2, from N group time domain data respectively determine corresponding N data points;S3, based on N data points, the amplitude x N (t m ) of N data points is solved Constant C (t m ) contained;S4, respectively calculate the residual VAR N (t m ) of the amplitude x m (t N ) of N data points and constant C (t m );S5, determine the distribution type that random variable obeys;S6, by regression calculation, solve all L undetermined parameters;S7, utilize distribution f to generate infinite random variables;S8, obtain T length vibration or impact environment time domain test waveform.The method of the present application combines the advantages of traditional frequency spectrum control method and time domain waveform method, the vibration or impact environment characteristics experienced by equipment in time domain is explicit and parameterized, can be extended for similar vibration or impact environment test examination, with good universality, high engineering application value.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of environmental engineering, and particularly relates to a vibration or impact environment time-domain waveform simulation method and system. BACKGROUND

[0002] In recent years, with the continuous improvement of the combat performance of weapon equipment, the induced vibration or impact environment is also increasingly complex, for example, the vibration or impact environment induced by airplane machine gun shooting is characterized by a large number of repeated impacts, and the vibration or impact environment induced by vehicle driving on a bumpy road is characterized by non-Gaussian distribution of vibration, and the like. The influence of these complex vibration or impact environments on nearby installed equipment cannot be ignored. Therefore, in order to ensure that the equipment is normally used in the above vibration or impact environment, it is necessary to carry out vibration or impact environment test in the laboratory in advance to evaluate the equipment.

[0003] In the traditional vibration or impact environment test, a frequency spectrum control method is generally used. For example, a power spectral density is used to generate a time-domain waveform in the random vibration test, with the premise that the random vibration environment experienced by the equipment is ergodic and obeys Gaussian distribution. The impact response spectrum is used to generate a time-domain waveform in the impact test, but the actual impact environment with complex time-domain characteristics cannot be accurately simulated. Therefore, for non-stationary vibration or non-Gaussian distributed vibration or impact environment, the frequency spectrum control method has certain limitations.

[0004] In order to make up for the shortcomings of the frequency spectrum control method, an engineering field has developed a test technology for directly simulating the actual waveform of the vibration or impact environment in the time domain. The principle is to connect a group or multiple groups of vibration or impact environment time-domain signals measured and collected at the beginning and the end, and combine them into a group of test waveforms as test condition input to the test bench controller to evaluate the equipment. However, this method requires a sufficient number of time-domain signal samples to fully represent the typical key events such as transient vibration, high-level non-stationary vibration or complex impact in the vibration or impact environment. However, the number of samples obtained in actual work is limited, and the method does not further analyze the samples, so it cannot guarantee that the limited samples can effectively represent all the vibration or impact characteristics of the equipment, and therefore there is a risk of insufficient test evaluation. SUMMARY

[0005] In view of the shortcomings of the prior art, the present application provides a vibration or impact environment time-domain waveform simulation method and system, which combines the advantages of the traditional frequency spectrum control method and the time-domain waveform method, explicitly and parameterizes the vibration or impact environment characteristics experienced by the equipment in the time domain, can be extended for test evaluation of similar vibration or impact environment, has good universality, and has high engineering application value.

[0006] To achieve the above object, the present application discloses the following technical scheme:

[0007] Specifically, the present application provides a vibration or impact environment time domain waveform simulation method, which comprises the following steps:

[0008] S1, for the equipment performing a certain task, arranging a sensor to collect vibration or impact environment data, repeatedly collecting N groups (N≥2) of time domain data, the sampling frequency and the start and end time of each group of time domain data being the same;

[0009] S2, for the N groups of time domain data with a time length of T, determining N data points corresponding to any time t m in the time length T from the N groups of time domain data, the amplitude of the data point being x N (t m ), wherein 0≤t m ≤T;

[0010] S3, based on the N data points, solving the constant C(t N ) contained in the amplitude x m (t m ) of the N data points;

[0011] S4, respectively calculating the residual VAR N (t m ) of the amplitude x N (t m ) of the N data points and the constant C(t m ), i.e.:

[0012] VAR N (t m )=x N (t m )-C(t m )

[0013] defining the residual VAR N (t m ) as a random variable obeying a certain distribution;

[0014] S5, performing non-parametric estimation on the N random variables VAR N (t m ) to determine the distribution type f obeyed by the N random variables VAR N (t m );

[0015] S6, assuming that the distribution f contains L undetermined parameters (p1,…p L ), performing regression calculation on the distribution f (p1,…p N ) by using the N random variables VAR m (t L ) to solve all the L undetermined parameters p1,…p L , wherein N≥L;

[0016] S7, generating infinite virtual random variables U using the distribution f under the specified probability C P (t m ), the joint constant C(t m ), infinite data points x'(t m ) corresponding to the time t m ) can be constructed, and the calculation formula of each data point x'(t m ) is as follows:

[0017] x'(t m ) = C(t m ) + UP(t m )

[0018] Wherein, x'(t m ) is the amplitude of the simulated generated time domain waveform at time t m ; C(t m ) is a constant, U P (t m ) is a certain virtual random variable;

[0019] S8, repeat steps S1 to S7 to construct the time series signal {x'(0), x'(t1), x'(t2), … x'(T)} within T time length, that is, the vibration or shock environment time domain waveform within T time length can be obtained.

[0020] The installation position of the sensor in step S1 can represent the typical vibration or shock environment characteristics of the equipment, which refers to the non-Gaussian non-stationary characteristic signal in the actual vibration or shock environment.

[0021] Preferably, the non-parametric estimation in step S5 refers to analyzing and calculating the distribution parameters using the information of the sample itself when the distribution of the sample is unknown.

[0022] Preferably, the distribution test in step S5 refers to judging whether a group of samples conforms to a certain specified distribution.

[0023] Preferably, the regression calculation in step S6 includes linear regression calculation and nonlinear regression calculation.

[0024] Preferably, the distribution type in step S5 includes uniform distribution, exponential distribution, normal distribution and chi-square distribution.

[0025] Preferably, the specific solving method of the constant C(t m ) in step S3 is: constructing N data points at time t m with N as the independent variable and x N (t mThe scatter plot of the dependent variable is used to perform regression calculations on the scatter plots to obtain the regression function. The intersection of the regression function and the dependent variable axis is considered as a constant C(t). m ).

[0026] Another aspect of the present invention provides a time-domain waveform simulation system for vibration or shock environments, which includes an acquisition unit, an amplitude determination unit, an amplitude decomposition unit, a random variable distribution determination unit, a distribution parameter solution unit, a data point construction unit, and a vibration or shock environment time-domain waveform simulation unit.

[0027] The acquisition unit repeatedly acquires N sets (N≥2) of data, with the same sampling frequency and start and end time for each set of data;

[0028] The amplitude determination unit determines the amplitude at any time t within the duration T for N sets of time-domain data. m From N sets of time-domain data, N corresponding data points with an amplitude of x can be determined. N (t m ), where 0≤t m ≤T;

[0029] The amplitude decomposition unit will t m The magnitudes of N data points at time t are decomposed into a constant C(m) and a random variable VAR(t). m ) N ,Right now:

[0030] x N (t m )=C(m)+VAR(t m ) N ,

[0031] Where, x N (t m ) is the amplitude, C(m) is a constant, and VAR(t) is the amplitude. m ) N It is a random variable;

[0032] The random variable follows a distribution that determines the random variable VAR(t) for N data points. m ) N Perform non-parametric estimation to determine VAR(t) m ) N The distribution f it follows;

[0033] The distribution parameter solving unit uses N random variables VAR(t) m ) N For distribution f(p1,…p) L Perform regression calculations to solve for all L undetermined parameters p1,…p L Where N≥L;

[0034] The data point construction unit generates an infinite number of random variables U using the distribution f at a specified probability C P (t m ), a joint constant C(m), to construct a plurality of data points x'(t m ) corresponding to the t m time

[0035] x'(t m ) = C(m) + U P (t m )

[0036] Wherein, x'(t m ) is the amplitude of the simulated generated time domain waveform at t m time;

[0037] The vibration or shock environment time domain waveform simulation unit can traverse all time points within the T duration to construct a time series signal {x'(0), x'(t1), x'(t2), … x'(T)} within the T duration, that is, the vibration or shock environment time domain waveform within the T duration can be obtained.

[0038] Compared with the prior art, the present application has the following beneficial effects:

[0039] (1) The vibration or shock environment time domain waveform simulation method provided by the present application considers the time-varying characteristics of the original sample and decomposes the amplitude of the original sample into a constant and a random variable; by statistically analyzing the random variable, a consistent rule is extracted to reconstruct the time domain waveform. Compared with the traditional multi-group signal concatenation method, the present application can more accurately and effectively reflect the random characteristics of the vibration or shock environment, and can effectively reduce the risk of insufficient test evaluation when used for environmental test evaluation.

[0040] (2) The vibration or shock environment time domain waveform simulation method provided by the present application combines the advantages of the traditional frequency spectrum control method and the time domain waveform method, explicitly and parameterizes the vibration or shock environment characteristics experienced by the device in the time domain, and can be extended for test evaluation of similar vibration or shock environments, has good universality, and has high engineering application value. BRIEF DESCRIPTION OF DRAWINGS

[0041] Figure 1 is a flowchart of the vibration or shock environment time domain waveform simulation method of the present application;

[0042] Figure 2 is a time domain waveform schematic diagram in an embodiment of the present application;

[0043] Figure 3 is a measured time domain waveform schematic diagram in an embodiment of the present application;

[0044] Figure 4 Constant diagram for 3 groups of measured time domain waveform sets in the embodiment of the present application;

[0045] Figure 5 Constant diagram for 3 groups of residual waveforms with a time length of 0.5S in the embodiment of the present application;

[0046] Figure 6 Constant diagram for the residual obeying normal distribution with parameters N(μ,σ) in the embodiment of the present application;

[0047] Figure 7 Constant diagram for the final reference time domain waveform in the embodiment of the present application. DETAILED DESCRIPTION

[0048] The exemplary embodiments, features and aspects of the present application will be described in detail below with reference to the accompanying drawings. The same reference numerals in the drawings denote the same or similar elements. Although various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless specifically indicated.

[0049] The present application provides a vibration or impact environment time domain waveform simulation method, as shown in Figure 1 and Figure 2 , which specifically comprises the following steps:

[0050] S1, for the equipment performing a certain task, a sensor is arranged to collect vibration or impact environment data. N groups (N≥2) of time domain data under vibration or impact environment are repeatedly collected by the arranged sensor, and each group of time domain data has the same sampling frequency and start and end time. The time domain data under vibration or impact environment refers to the vibration or impact response of the equipment under excitation, i.e. vibration or impact time domain data. When the sensor is installed to collect data, the installation position of the sensor should be able to represent the typical vibration or impact environment characteristics of the equipment, so as to ensure that the collected time domain data is the time domain data under vibration or impact environment. The typical vibration or impact environment characteristics refer to the non-Gaussian non-stationary characteristic signal in the actual vibration or impact environment.

[0051] S2, for the N groups of time domain data with a time length of T, at any time t m , N corresponding data points can be determined from the N groups of time domain data, and the amplitude of the data point is denoted as x N (t m ), wherein 0≤t m ≤T.

[0052] S3, at any time t m , the amplitude x N (t m) are different, so it can be considered that the data of N time domain data at the same time has time-varying characteristics. Therefore, the "addition model" based on vibration or impact environment data can decompose the amplitude x m (t N ) of N data points at time t m ) into a constant C(m) and a random variable VAR(t m ) N That is:

[0053] x N (t m )=C(m)+VAR(t m ) N

[0054] Among the amplitudes x N (t m ) of N data points, the constant C(m) contained is the same, while the random variable VAR(t m ) N is different, and the difference of the random variable also leads to the difference of the amplitude of each data point. Therefore, based on the amplitudes of multiple data points, the value of C(m) can be calculated first to determine the constant part of the amplitudes of N data points corresponding to time t m .

[0055] In the specific application process, the constant C(m) is obtained by calculating the arithmetic mean of multiple data points at time t m . In one application example, the specific calculation process is as follows: N data points at time t m , a scatter plot is constructed with N as the independent variable and x N (t m ) as the dependent variable, the scatter plot is regressed to obtain a regression function, and the intersection of the regression function and the dependent variable axis is regarded as the constant C(t m ).

[0056] S4, calculate the residual VARN(tm) of the amplitude xN(tm) of N data points and the constant C(tm) respectively, that is:

[0057] VARN(tm)=xN(tm)-C(tm)

[0058] Define the residual VARN(tm) as a random variable subject to a certain distribution.

[0059] Use the constant C(tm) of N data points calculated in step S3 to calculate the residual VARN(tm) of N data points, that is, the random variable of multiple data points at this time.

[0060] S5, calculate the N random variables VAR N (t mPerform non-parametric estimation to determine N random variables VAR. N (t m The distribution type f that it follows.

[0061] S6. Assume that distribution f contains L undetermined parameters (p1,…p). L Using N random variables VAR N (t m For the distribution f(p1,…p) L Perform regression calculations to solve for all L undetermined parameters p1,…p L , where N≥L.

[0062] Nonparametric estimation refers to statistical testing and analysis of a sample using information from the sample itself when the sample distribution is unknown. Distribution testing determines whether a sample represents a background population that differs significantly from a specified distribution. Distribution types include uniform, exponential, normal, and chi-square distributions. Regression calculations include linear and nonlinear regression calculations.

[0063] S7. Given a specified probability C, generate an infinite number of virtual random variables U using distribution f. P (t m ), joint constant C(t) m ), capable of constructing the corresponding t m Infinitely many data points x'(t) at time t m ), for each data point x'(t m The calculation formula for ) is as follows:

[0064] x'(t m )=C(t m )+U P (t m )

[0065] Where x'(t) m To simulate and generate time-domain waveforms at t m Amplitude at time; C(t) m ) is a constant, U P (t m ) is a certain dummy random variable.

[0066] Under a specified probability C, infinitely many virtual random variables U are generated using distribution f through simulation software. P (t m ).

[0067] S8, repeat steps S1 to S7 to calculate the infinite simulation amplitudes of multiple time points, and construct the time series signal corresponding to all time points in T duration based on the infinite time domain waveform amplitudes of all time points, that is, {x'(0), x'(t1), x'(t2), … x'(T)}, and the vibration or impact environment time domain waveform of T duration can be obtained by reconstructing the time domain waveforms of all time points. Embodiments

[0069] The content of the application will be described in detail below through embodiments.

[0070] The vibration and impact environment data of a certain type of vehicle during driving on a gravel road were measured and collected, including the vibration or impact response of the driving direction, vertical and lateral directions, and the sampling frequency was 10240Hz. The process of the specific time domain waveform simulation method is as follows:

[0071] S1, the certain type of vehicle repeatedly drives three times under the same working condition, and the time domain waveform of the vertical passage of the mounting surface of the certain type of equipment on the vehicle is taken as the analysis object, and 3 groups of measured time domain waveforms are collected by using the installed sensors, and the time duration is 0.5S, as shown in Figure 1 Figure 3

[0072] S2, for the collected time domain data with a time duration of T, for any time t m (0≤t m ≤T) in T duration, the corresponding N data points can be determined from N groups of time domain data, and the amplitude is x N (t m ).

[0073] S3, for the 3 groups of measured time domain waveforms, the arithmetic mean value of each time point in 0.5S duration is calculated. The constant of the 3 groups of measured time domain waveform sets is obtained, that is, the constant C(m), as shown in Figure 4

[0074] S4, subtract the constant C(m) from the 3 groups of measured time domain waveforms at the same time, to obtain 3 groups of residual waveforms with a time duration of 0.5S, that is, random variables, as shown in Figure 5

[0075] S5, non-parametric estimation and distribution test are performed on the 3 groups of random variables to determine that the random variables follow the normal distribution with parameters N(μ,σ). The regression calculation of N(μ,σ) is performed by using the 3 groups of random variables, as shown in Figure 6

[0076] S6, for a given probability C=95%, the corresponding deviation value U P (t m ) is obtained by substituting into the distribution N(μ,σ).​​​​​

[0077] S7. In summary, a t can be constructed. m The expression for the data point at time:

[0078] x'(t m )=C(t m )+U p (t m )

[0079] Where x'(t) m To simulate and generate time-domain waveforms at t m Amplitude at time; C(t) m ) is a constant, UP(t) m ) is a certain dummy random variable.

[0080] S8. Based on the calculated residual distribution N(μ,σ), repeat steps S1-S6 to generate a series of random components with a given probability C = 95%, and add them to the constant to obtain the final reference time-domain waveform, as shown in the appendix. Figure 7 As shown.

[0081] Another aspect of the present invention provides a time-domain waveform simulation system for vibration or shock environments, which includes an acquisition unit, an amplitude determination unit, an amplitude decomposition unit, a random variable distribution determination unit, a distribution parameter solution unit, a data point construction unit, and a vibration or shock environment time-domain waveform simulation unit.

[0082] The acquisition unit repeatedly acquires N sets of data (N≥2), with the same sampling frequency and start and end time for each set of data.

[0083] For N sets of time-domain data, each with a duration of T, the amplitude determination unit determines the amplitude at any time t within the duration of T. m From N sets of time-domain data, N corresponding data points with an amplitude of x can be determined. N (t m ), where 0≤t m ≤T.

[0084] The amplitude decomposition unit will t m The magnitudes of N data points at time t are decomposed into a constant C(m) and a random variable VAR(t). m ) N ,Right now:

[0085] x N (t m )=C(m)+VAR(t m ) N ,

[0086] Where, x N (t m) is an amplitude, C(m) is a constant, VAR(t m ) N is a random variable.

[0087] The distribution determining unit determines the distribution of the random variable VAR(t m ) N The non-parametric estimation is performed to determine the distribution f of the random variable VAR(t m ) N VAR(t m ).

[0088] The distribution parameter solving unit solves all L undetermined parameters p1,…p N ) by performing regression calculation on the distribution f(p1,…p L ). L Wherein, N≥L.

[0089] The data point constructing unit generates infinite random variables U P (t m ) under the specified probability C, and constructs multiple data points x'(t m ) corresponding to the time t m ) by using the distribution f and combining the constant C(m), that is

[0090] x'(t m ) = C(m) + U P (t m )

[0091] Wherein, x'(t m ) is an amplitude of the simulation generated time domain waveform at the time t m .

[0092] The vibration or impact environment time domain waveform simulation unit can traverse all time points within the time T, and construct a time sequence signal {x'(0), x'(t1), x'(t2),…x'(T)} within the time T, that is, the vibration or impact environment time domain waveform within the time T can be obtained.

[0093] The time domain test waveform simulation method provided by the application combines the advantages of the traditional frequency spectrum control method and the time domain waveform method, explicitly and parameterizes the vibration or impact environment characteristics experienced by the device in the time domain, can be extended for test examination of similar dynamic and impact environments, has good universality, and has high engineering application value.

[0094] The above-described embodiments only describe the preferred embodiments of the application, and do not limit the scope of the application. Without departing from the design spirit of the application, various modifications and improvements of the technical solutions of the application made by those skilled in the art shall fall within the protection scope of the claims of the application.

Claims

1. A method for simulating time-domain waveforms in vibration or shock environments, characterized in that: It includes the following steps: S1. For equipment performing a certain task, deploy sensors to collect vibration or shock environment data, and repeatedly collect N sets (N≥2) of time domain data. The sampling frequency and start and end time of each set of time domain data are the same. S2. For N sets of time-domain data, each with a duration of T, at any time t within the duration T... m From N sets of time-domain data, N corresponding data points can be determined, and the amplitude of each data point is x. N (t m ), where 0≤t m ≤T; S3. Based on N data points, solve for the amplitude x of the N data points. N (t m The constant C(t) included m ); constant C(t) m The specific solution method for t is as follows: m Given N data points at time points, construct a system with N as the independent variable and x as the inverse variable. N (t m The scatter plot of the dependent variable is used to perform regression calculations on the scatter plots to obtain the regression function. The intersection of the regression function and the dependent variable axis is considered as a constant C(t). m ); S4. Calculate the amplitude x for each of the N data points. N (t m ) and constant C(t) m VAR of residuals N (t m ),Right now: VAR N (t m )=x N (t m )-C(t m ) ; Define residual VAR N (t m Let be a random variable that follows a certain distribution; S5. VAR of N random variables N (t m Perform non-parametric estimation to determine N random variables VAR. N (t m The distribution type f it follows; S6. Assume that distribution f contains L undetermined parameters (p1,…p). L Using N random variables VAR N (t m For the distribution f(p1,…p) L Perform regression calculations to solve for all L undetermined parameters p1,…p L Where N≥L; S7. Given a specified probability C, generate an infinite number of virtual random variables U using distribution f. P (t m ), joint constant C(t) m ), capable of constructing the corresponding t m Infinitely many data points at time Each data point The calculation formula is as follows: ； in, To simulate the generation of time-domain waveforms at t m Amplitude at any given moment; It is a constant. Let it be a certain dummy random variable; S8. Traverse all time points within time period T, repeating steps S1 to S7 to construct a time series signal within time period T. That is, it is possible to obtain the time-domain waveform of a vibration or shock environment with a duration of T.

2. The method for simulating time-domain waveforms in vibration or shock environments according to claim 1, characterized in that: The installation location of the sensor in step S1 can represent the typical vibration or shock environment characteristics of the equipment. The typical vibration or shock environment characteristics refer to the non-Gaussian non-stationary characteristic signals in the actual vibration or shock environment.

3. The method for simulating time-domain waveforms in vibration or shock environments according to claim 1, characterized in that: The non-parametric estimation mentioned in step S5 refers to analyzing and calculating the distribution parameters using information from the sample itself when the distribution of the sample is unknown.

4. The method for simulating time-domain waveforms in vibration or shock environments according to claim 1, characterized in that: In step S5, the distribution test refers to determining whether a set of samples follows a certain specified distribution.

5. The method for simulating time-domain waveforms in vibration or shock environments according to claim 4, characterized in that: The regression calculation in step S6 includes linear regression calculation and nonlinear regression calculation.

6. The method for simulating time-domain waveforms in vibration or shock environments according to claim 5, characterized in that: The distribution types mentioned in step S5 include uniform distribution, exponential distribution, normal distribution, and chi-square distribution.

7. A vibration or shock environment time-domain waveform simulation system for the vibration or shock environment time-domain waveform simulation method according to claim 1, characterized in that: It includes an acquisition unit, an amplitude determination unit, an amplitude decomposition unit, a random variable distribution determination unit, a distribution parameter solution unit, a data point construction unit, and a vibration or shock environment time-domain waveform simulation unit; The acquisition unit repeatedly acquires N sets (N≥2) of data, with the same sampling frequency and start and end time for each set of data; The amplitude determination unit determines the amplitude at any time t within the duration T for N sets of time-domain data. m From N sets of time-domain data, N corresponding data points with an amplitude of x can be determined. N (t m ), where 0≤t m ≤T; The amplitude decomposition unit will t m The magnitudes of N data points at time t are decomposed into a constant C(m) and a random variable VAR(t). m ) N ,Right now: x N (t m )=C(m)+VAR(t m ) N ; Where, x N (t m ) is the amplitude, C(m) is a constant, and VAR(t) is the amplitude. m ) N It is a random variable; The random variable follows a distribution that determines the random variable VAR(t) for N data points. m ) N Perform non-parametric estimation to determine VAR(t) m ) N The distribution f it follows; The distribution parameter solving unit uses N random variables VAR(t) m ) N For distribution f(p1,…p) L Perform regression calculations to solve for all L undetermined parameters p1,…p L Where N≥L; The data point construction unit generates an infinite number of random variables U using distribution f under a specified probability C. P (t m ), joint constant C(m), construct the corresponding t m Multiple data points at time ,Right now ； in, To simulate the generation of time-domain waveforms at t m Amplitude at any given moment; The environmental time-domain waveform simulation unit can traverse all time points within a time period T to construct a time-series signal within a time period T. That is, it is possible to obtain the time-domain waveform of a vibration or shock environment with a duration of T.

Images