A life prediction method for heavy-duty industrial robot reducer based on acceleration test

By constructing stress loading spectrum and accelerated degradation model, considering the effect of periodic shock, the problem of low life prediction accuracy in the existing technology is solved, and more accurate life prediction and operation and maintenance guidance are achieved.

CN118656937BActive Publication Date: 2024-11-26HEBEI UNIV OF TECH +2
View PDF 7 Cites 0 Cited by

Patent Information

Application Number
CN202411110422.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-14
Publication Date
2024-11-26
Estimated Expiration
2044-08-14

AI Technical Summary

Technical Problem

The existing acceleration test methods cannot accurately simulate the degraded environment of heavy-loaded industrial robot reducers under real use conditions, resulting in low life prediction accuracy and the inability to accurately describe the impact of periodic impact loads.

Method used

By obtaining the loaded information set of the reducer to be tested, the stress loading spectrum is constructed, and the degradation data is collected based on the acceleration test, an accelerated degradation model is constructed, the impact factor is introduced, a degradation model is established that takes into account the effect of periodic shocks, and the reliability function is derived. , calculate the predicted lifespan.

Benefits of technology

It improves the accuracy and authenticity of life prediction, can more accurately describe the life of the reducer under different working conditions, and provides effective operation and maintenance guidance.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118656937B_ABST
    Figure CN118656937B_ABST
Patent Text Reader

Abstract

The present application discloses a method for predicting the life of a heavy-duty industrial robot reducer based on an accelerated test, and relates to the technical field of reducer life prediction. The method includes obtaining a set of load information of the reducer to be tested; constructing a stress loading spectrum based on the load information set; conducting an accelerated test based on the stress loading spectrum, collecting degradation data of the reducer to be tested; constructing an accelerated degradation model, and deriving a reliability function; determining unknown parameters in the reliability function based on the degradation data; obtaining stress data of the reducer to be tested under normal working conditions, and calculating the reliability; drawing a reliability curve, and obtaining the median life when the reliability is 50%, as the predicted life of the reducer to be tested. The present application compiles a stress loading spectrum based on real service data to carry out accelerated tests, improves the accuracy of the degradation data obtained, and then constructs an accelerated degradation model, and finally derives a reliability function to complete the life prediction of the reducer.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application generally relates to the technical field of reducer life prediction, and specifically to a method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test. Background Art

[0002] In order to ensure that the reducer has good service performance during actual use, a life test is usually carried out on it after the prototype is developed to simulate the actual use conditions of the reducer, so as to evaluate its durability and reliability and predict its service life.

[0003] At present, in order to save manpower and material resources and improve test efficiency, acceleration tests are generally used to test reducer prototypes, such as constant, cyclic, and step stress test methods. However, these methods cannot simulate the degradation environment of the reducer, which will cause the obtained degradation data to deviate greatly from the actual situation, affecting the subsequent life prediction accuracy. In addition, heavy-duty industrial robots will be subjected to periodic impact loads caused by frequent start-stop and forward and reverse switching during service. The existing accelerated degradation model cannot accurately describe this phenomenon, and the fitting accuracy is low, which will also affect the accuracy of the life prediction results. Therefore, we propose a life prediction method for heavy-duty industrial robot reducers based on acceleration tests to solve the above problems. Summary of the invention

[0004] In view of the above-mentioned defects or deficiencies in the prior art, it is desired to provide a life prediction method for a heavy-duty industrial robot reducer based on an acceleration test, which improves the authenticity and accuracy of life prediction.

[0005] The present application provides a method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test, comprising the following steps:

[0006] Acquire a set of load information of the reducer to be tested; the set of load information includes sampling time and corresponding load torque data;

[0007] Constructing a stress loading spectrum according to the set of loading information; the stress loading spectrum at least includes a plurality of loading data and corresponding loading time and stress level;

[0008] Performing an acceleration test based on the stress loading spectrum, and collecting degradation data of the reducer to be tested; the degradation data at least includes a plurality of operating times and corresponding transmission accuracy values;

[0009] Constructing an accelerated degradation model, and deriving a reliability function based on the accelerated degradation model; the accelerated degradation model is used to describe the degradation process of the reducer;

[0010] determining unknown parameters in the reliability function according to the degradation data;

[0011] Obtaining stress data of the reducer to be tested under normal working conditions, and substituting the stress data into a reliability function to calculate the reliability;

[0012] A reliability curve showing the reliability changing with time is plotted, and the median life when the reliability is 50% in the reliability curve is obtained as the predicted life of the reducer to be tested.

[0013] According to the technical solution provided in the embodiment of the present application, obtaining a set of load information of the reducer to be tested specifically includes the following steps:

[0014] The load torque data of the reducer to be tested during service is collected at a preset sampling frequency and a preset number of sampling points, and a load information set is generated according to the sampling time and the corresponding load torque data.

[0015] According to the technical solution provided in the embodiment of the present application, constructing a stress loading spectrum according to the loaded information set specifically includes the following steps:

[0016] Acquire load data corresponding to at least one load cycle in the load information set, and generate a first load curve according to the load data;

[0017] removing noise data in the first load curve and performing smoothing processing on the first load curve to obtain a second load curve;

[0018] dividing the second load curve into a plurality of stress levels according to a preset data interval;

[0019] Obtaining a load mean value corresponding to each of the stress levels, and expanding each load mean value to obtain an expanded load value; calculating an equivalent torque based on the expanded load value, wherein the equivalent torque is equal to a maximum average load torque allowed by the reducer to be tested;

[0020] Eliminate the enlarged load value that is greater than 2 times the rated torque of the reducer to be tested to obtain the final load data;

[0021] A stress loading spectrum is generated based on the final loading data.

[0022] According to the technical solution provided in the embodiment of the present application, the expanded load value is calculated according to the following formula:

[0023] ;

[0024] in, The maximum allowable average load torque of the reducer obtained from the product manual. is the time corresponding to the first stress level, is the time corresponding to the second stress level, is the time corresponding to the mth stress level, n 1 is the reducer driving speed corresponding to the first stress level, which is the rated speed of the reducer. n 2 is the reducer driving speed corresponding to the second stress level, which is the rated speed of the reducer. n m is the reducer driving speed corresponding to the mth stress level, which is taken as the rated speed of the reducer. T 1 is the load value after the load mean is enlarged corresponding to the first stress level, T 2 is the load value after the load mean is enlarged corresponding to the second stress level. T m is the load value after the load mean corresponding to the mth stress level is expanded, m is the number of stress levels.

[0025] According to the technical solution provided in the embodiment of the present application, an accelerated degradation model is constructed, which specifically includes the following steps:

[0026] Obtain power-law and generalized log-linear models;

[0027] Introducing a shock influence factor into the power law model, and constructing a transfer function based on the generalized log-linear model;

[0028] Based on the power law model with the impact influence factor and the transfer function, an accelerated degradation model is obtained.

[0029] According to the technical solution provided in the embodiment of the present application, an accelerated test is performed based on the stress loading spectrum, and degradation data of the reducer to be tested is collected, which specifically includes the following steps:

[0030] The reducer to be tested is loaded according to the loading data in the stress loading spectrum, and degradation data is collected at preset intervals.

[0031] According to the technical solution provided in the embodiment of the present application, after constructing the stress loading spectrum according to the load information set and before performing the accelerated test based on the stress loading spectrum, the following steps are also included:

[0032] Check whether the reducer to be tested meets the preset conditions; the preset conditions at least include: the appearance is intact and the lubricating material meets the preset standards and there is no leakage;

[0033] When the reducer to be tested meets the preset conditions, an acceleration test is performed based on the stress loading spectrum.

[0034] It can be seen from the above technical solution that the present application has at least the following beneficial effects:

[0035] The present application discloses a life prediction method for a heavy-duty industrial robot reducer based on an acceleration test, the method comprising: obtaining a set of load information of the reducer to be tested, the load information set comprising a sampling time and corresponding load torque data; constructing a stress loading spectrum according to the load information set, the stress loading spectrum comprising at least a plurality of load data and corresponding loading time and stress level; performing an acceleration test based on the stress loading spectrum, and collecting degradation data of the reducer, the degradation data comprising at least a plurality of operating times and corresponding transmission accuracy values; constructing an acceleration degradation model, the acceleration degradation model being used to describe the impact effect on the reducer during service, and deriving a reliability function according to the acceleration degradation model; determining unknown parameters in the reliability function according to the degradation data; obtaining stress data of the reducer to be tested under normal working conditions, substituting the stress data into the reliability function, and calculating the reliability; drawing a reliability curve showing the variation of the reliability with time, and obtaining the median life when the reliability is 50% in the reliability curve as the predicted life of the reducer to be tested.

[0036] This application uses the stress loading spectrum of the reducer to be tested to carry out accelerated tests and obtain degradation data to ensure that the damage accumulation mode of the reducer during the test is consistent with the actual working process. The impact factor is introduced into the existing power law model, and the functional relationship between stress and impact degradation is constructed based on the generalized logarithmic linear model, thereby constructing an accelerated degradation model that takes into account periodic impact effects. The reliability function is then derived based on the accelerated degradation model, and the unknown parameters in the reliability function are estimated based on the degradation data. The life of the reducer is evaluated based on the reliability function. By substituting different torque values, the life of the reducer under different working conditions can be predicted. This method can greatly improve the accuracy of life prediction and provide effective guidance for the operation and maintenance of heavy-duty industrial robots. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Other features, objects and advantages of the present application will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings.

[0038] Figure 1 The figure is a flow chart of the life prediction method of the reducer of a heavy-duty industrial robot based on acceleration test.

[0039] Figure 2 Schematic diagram of the process of constructing stress loading spectrum.

[0040] Figure 3 A graph of the information set being loaded.

[0041] Figure 4 It is a schematic diagram of the first load curve.

[0042] Figure 5 It is a schematic diagram of the second load curve.

[0043] Figure 6 This is a data diagram of the mean load value corresponding to the stress level.

[0044] Figure 7 A schematic diagram of a load information collection of an example harmonic reducer.

[0045] Figure 8 is a graph of degradation data for an example harmonic reducer.

[0046] Fig. 9 It is a schematic diagram of the reliable life curve when the load torque is 87Nm.

[0047] Fig.10 It is a schematic diagram of the reliable life curve when the load torque is 100Nm. DETAILED DESCRIPTION

[0048] The present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the relevant invention, rather than to limit the invention. It is also necessary to explain that, for ease of description, only the parts related to the invention are shown in the accompanying drawings.

[0049] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0050] In order to make the description of the following embodiments clear and concise, a brief introduction to the related technology is first given:

[0051] Industrial robots have the advantages of high working efficiency, high precision and good stability. They play a vital role in traditional manufacturing and are currently widely used in different industries such as mechanical processing, automobile parts manufacturing, food processing, electronics and electrical, logistics, etc. As robot technology continues to mature, the load capacity of the robot end is significantly improved, and heavy-duty industrial robots are emerging in various fields. Therefore, as the core component of heavy-duty industrial robots, the reducer accurately evaluates its service life and plays an important guiding role in the operation and maintenance of heavy-duty industrial robots.

[0052] The type of reducer is, for example, a harmonic reducer or an RV reducer.

[0053] This application uses a harmonic reducer as an example for explanation. By analyzing the failure mechanism of the harmonic reducer, it can be seen that its main failure mechanism is the degradation of accuracy caused by wear; specifically, the wear rate of the harmonic reducer is calculated according to the following formula. From this formula, it can be seen that increasing the load torque has the effect of increasing the wear rate, that is, increasing the wear rate can accelerate the failure of the harmonic reducer. Therefore, the load torque can be used as the acceleration stress to carry out an accelerated test.

[0054] ;

[0055] in, ψ is the wear rate, which represents the speed of wear; K is the operating condition coefficient, which is a fixed constant; T is the load torque; is the flexwheel pitch circle diameter of the reducer, which is a fixed constant; is the tooth angle of the harmonic reducer, which is a fixed constant; n is the driving speed.

[0056] In order to make the heavy-duty industrial robot reducer life prediction method based on acceleration test provided in the embodiment of the present application clearer and easier to understand, the method is introduced below with reference to the accompanying drawings. Figure 1 As shown, the figure is a schematic flow chart of a method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test provided in an embodiment of the present application, the method comprising:

[0057] S100, obtaining a load information set of the reducer to be tested; the load information set includes a sampling time and corresponding load torque data.

[0058] It should be noted that the load information set refers to the load data set that the reducer to be tested is borne during the actual working process.

[0059] The step of obtaining the load information set of the reducer to be tested specifically includes the following steps:

[0060] The load torque data of the reducer to be tested during service is collected at a preset sampling frequency and a preset number of sampling points, and a load information set is generated according to the sampling time and the corresponding load torque data.

[0061] Among them, torque sensors are installed at the joints of heavy-duty industrial robots, and load torque data is collected in real time using the torque sensors. Figure 3 As shown, the figure is a curve chart of the load information set, which represents the measured load data of the third joint of the heavy-load industrial robot during service. The preset sampling period is, for example, 43.75ms, and the sampling point is, for example, 800.

[0062] S200, constructing a stress loading spectrum according to a set of loading information; the stress loading spectrum at least includes a plurality of loading data and corresponding loading time and stress level.

[0063] It should be noted that if Figure 2 As shown in FIG. 1 , a stress loading spectrum is constructed according to a set of loading information, which specifically includes the following steps:

[0064] S201. Obtain load data corresponding to at least one load cycle in a load information set, and generate a first load curve according to the load data.

[0065] In the acquisition process, the heavy-duty industrial robot executes normal working instructions, and the final load information set includes at least one load cycle to reflect the working condition information of the reducer. That is, since the load torque is periodic, only one load cycle is required to be analyzed, and a load cycle period in the load information set obtained in step S100 can be intercepted as the analysis data of this step. For example, Figure 4 As shown, this figure is a schematic diagram of the first load curve.

[0066] S202: remove noise data in the first load curve, and perform smoothing on the first load curve to obtain a second load curve.

[0067] Among them, since the reducer is affected by environmental factors, vibration shock and system errors, some noise data will appear in the measured first load curve. These noise data cannot reflect the actual load and need to be filtered out. This application can use Gaussian filtering to remove noise data. Then, the curve with noise removed is smoothed to obtain the second load curve. For example, Figure 5 As shown, this figure is a schematic diagram of the second load curve.

[0068] S203: Divide the second load curve into a plurality of stress levels according to a preset data interval.

[0069] For example, Figure 5 There are 225 sampling points in the second load curve of , which are divided into 25 groups, that is, divided into 25 stress levels. Here, the preset data spacing is 9 sampling points.

[0070] S204, obtaining the load mean corresponding to each stress level, and expanding each load mean to obtain an expanded load value; calculating the equivalent torque based on the expanded load value, the equivalent torque being equal to the maximum average load torque allowed by the reducer to be tested.

[0071] by Figure 5 Take the second load curve as an example. Each group contains 9 sampling points. The data corresponding to each group of sampling points are averaged to obtain multiple load means, that is, 225 sampling points are converted into 25 sampling points, such as Figure 6 As shown in the figure, this figure is a data schematic diagram of the load mean corresponding to the stress level. There are 25 steps in the figure, and each step represents a stress level.

[0072] Specifically, the expanded load value is calculated according to the following formula:

[0073] ;

[0074] in, is the maximum average load torque that the reducer can withstand according to the product manual. is the time corresponding to the first stress level, is the time corresponding to the second stress level, is the time corresponding to the mth stress level, n 1 is the reducer driving speed corresponding to the first stress level, which is the rated speed of the reducer. n 2 is the reducer driving speed corresponding to the second stress level, which is the rated speed of the reducer. n m For the m The reducer driving speed corresponding to each stress level is taken as the rated speed of the reducer. T 1 is the load value after the load mean value corresponding to the first stress classification is enlarged, T 2 is the load value after the load mean is enlarged corresponding to the second stress level. T m For the m Each stress level corresponds to the load value after the load mean is expanded. m is the number of stress levels.

[0075] The acceleration is achieved by expanding the mean load at each stress level.

[0076] S205, eliminating the expanded load values ​​that are greater than twice the rated torque of the reducer to be tested, and obtaining the final load data.

[0077] The acceleration effect is ensured by eliminating the enlarged load value that is greater than 2 times the rated torque of the reducer to be tested. Here, based on engineering experience, 2 times the rated torque of the reducer to be tested is used as the maximum load torque that does not change the failure mechanism.

[0078] S206. Generate a stress loading spectrum based on the final loading data.

[0079] Furthermore, after constructing the stress loading spectrum according to the load information set and before performing the accelerated test based on the stress loading spectrum, the following steps are also included:

[0080] Check whether the reducer to be tested meets the preset conditions; the preset conditions at least include: good appearance and sufficient lubricating material without leakage; here, good appearance refers to the integrity of the reducer's housing, connecting parts, etc.

[0081] When the reducer to be tested meets the preset conditions, an acceleration test based on the stress loading spectrum is performed.

[0082] By checking whether the reducer to be tested meets the preset conditions, the influence of the reducer itself is eliminated first to ensure the accuracy of life prediction.

[0083] S300, performing an acceleration test based on a stress loading spectrum, and collecting degradation data of the reducer to be tested; the degradation data at least includes multiple operating times and corresponding transmission error values.

[0084] Among them, the acceleration test is carried out based on the stress loading spectrum, and the transmission error degradation data of the reducer to be tested is collected, which specifically includes the following steps:

[0085] The reducer to be tested is loaded according to the loading data in the stress loading spectrum, and degradation data is collected at preset intervals.

[0086] On the premise that the reducer to be tested meets the preset conditions, the reducer prototype to be tested is installed on the accelerated life test bench for simulation. Here, the preset interval length is, for example, 16 hours, the number of sampling times is, for example, 19, and the start-stop process is performed every half an hour to simulate the impact load borne by the reducer.

[0087] In addition, traditional degradation indicators usually include vibration, current, etc., but the thresholds of these degradation indicators are difficult to obtain, and the specific failure criteria cannot be clearly known. This application selects transmission error as a degradation indicator to overcome the problem that traditional degradation indicators are difficult to obtain and the failure threshold is unclear.

[0088] S400, constructing an accelerated degradation model, and deriving a reliability function based on the accelerated degradation model; the accelerated degradation model is used to describe the degradation process of the reducer.

[0089] It should be noted that the construction of the accelerated degradation model specifically includes the following steps:

[0090] Obtain power-law and generalized log-linear models;

[0091] The impact factor is introduced into the power law model, and the transfer function is constructed based on the generalized log-linear model;

[0092] Based on the power law model with the impact factor and the transfer function, an accelerated degradation model is obtained.

[0093] Among them, heavy-loaded industrial robots will be subjected to impact loads during their service, and the impact loads will affect the degradation process of the harmonic reducer. Specifically, every time the impact load occurs, the degradation rate of the harmonic reducer will suddenly change; every time the impact load occurs, the degradation amount of the harmonic reducer will suddenly change.

[0094] In view of this, the impact influence factor is introduced on the basis of the power law model, which characterizes the influence of the impact load on the degradation rate of the harmonic reducer under the accelerated test state. The degradation rate is calculated according to the following formula:

[0095] ;

[0096] in, is the degradation rate, T is the load torque, k is the number of times the impact load occurs, A , e are the parameters to be estimated, μ is the impact factor.

[0097] By substituting different load torques T , the degradation rate of the harmonic reducer under different stress levels can be calculated.

[0098] Then, a transfer function is established based on the generalized log-linear model, which can convert the impact load borne by the harmonic reducer under the accelerated test state into a mutation value of the degradation amount. By substituting different stress change values, the mutation values ​​of the degradation amount under different stress levels can be calculated.

[0099] Specifically, the degradation mutation value is calculated according to the following formula:

[0100] ;

[0101] ;

[0102] ;

[0103] in, h is the damage value caused by a single impact load, , are the unknown parameters in the formula, is the torque mutation value, is the mutation value of degradation caused by a single impact load, is the conversion coefficient.

[0104] Based on the above-mentioned degradation rate calculation formula and degradation mutation value calculation formula, an accelerated degradation model is constructed, which is used to characterize the performance degradation of the harmonic reducer.

[0105] ;

[0106] in, X ( t ) is in t The transmission error degradation value at the moment, σ is the diffusion coefficient, B (t ) is the standard Brownian motion, H is the period of arrival of the periodic shock, which is set according to actual needs, for example, it is set to 0.5 in this solution.

[0107] Furthermore, the reliability function is derived as follows:

[0108] ;

[0109] in, R ( t ) is the reliability function, W is the degradation threshold, is the cumulative distribution function expression of the standard normal distribution, In x is a random variable.

[0110] S500, determining unknown parameters in the reliability function according to the degradation data;

[0111] It should be noted that the unknown parameters refer to the , , , µ , σ、A and e Assuming that there are r products participated in the test, and the total number of measurements for each product was g , every time Collect the transmission error degradation value of the primary harmonic reducer, and It should be selected as an integer multiple of the impact load occurrence cycle. y Product No. j The transmission error increment during the test is z yj .

[0112] The transmission error increment is calculated according to the following formula:

[0113] ;

[0114] in, n yj For the y Products in j The number of times the impact load occurs during each measurement. is the number of times the impact load occurs for the yth product during the j-1th measurement; t yj For the y Products in j The time to collect degradation indicators, t y(j-1) For the yThe time when the degradation index of a product is collected for the j-1th time; For the y Products in j The transmission error degradation value at the time of the first measurement, For the y The transmission error degradation value of a product at the j-1th measurement; For the y Products in j The Brownian motion during the measurement, For the y The Brownian motion of a product at the j-1th measurement.

[0115] The degradation increment follows a multidimensional normal distribution. μ zy is the mean matrix, Z y is the degenerate increment matrix, ∑ y is the covariance matrix, and its likelihood function and log-likelihood function are:

[0116] ;

[0117] ;

[0118] in, θ ={ , , , µ , σ、A , e}, is the likelihood function, is the log-likelihood function.

[0119] Based on the genetic algorithm, the log-likelihood function is maximized to solve the corresponding values ​​of the unknown parameters.

[0120] S600, obtaining stress data of the reducer to be tested under normal working conditions, substituting the stress data into a reliability function, and calculating the reliability.

[0121] S700, draw a reliability curve showing the change of reliability over time, and obtain the median life when the reliability is 50% in the reliability curve as the predicted life of the reducer to be tested.

[0122] This application takes into account the load conditions of the harmonic reducer, uses the stress loading spectrum of the reducer to be tested to carry out accelerated tests, obtains degradation data, and ensures that the damage accumulation mode of the reducer during the test is consistent with the actual working process; this application also proposes an accelerated degradation model that considers the influence of impact loads, introduces an impact influence factor into the existing power-law model, which characterizes the influence of impact loads on the degradation rate, and then establishes a transfer function based on the generalized log-linear model, which characterizes the functional relationship between the transmission error degradation amount and the load torque change value caused by a single impact load, and finally constructs an accelerated degradation model, and derives a reliability function, and estimates the unknown parameters in the reliability function based on the degradation data, and predicts the life of the reducer under different working conditions by substituting different torque values. This method can greatly improve the accuracy of life prediction and provide effective guidance for the operation and maintenance of industrial robots.

[0123] To facilitate understanding, the life prediction method of the heavy-duty industrial robot reducer based on acceleration test in this application is introduced below with reference to specific examples.

[0124] Taking a certain type of harmonic reducer as an example, the load torque is selected as the acceleration stress, and the load information set of the harmonic reducer is collected, such as Figure 7 As shown in Table 1, the sampling period is 43.75 ms and the number of sampling points is 800. The stress loading spectrum is compiled based on the above method, as shown in Table 1.

[0125] Table 1 Stress loading spectrum

[0126]

[0127] The transmission error is selected as the degradation index, and the prototype is selected to carry out the test; the loading test is carried out according to the stress loading spectrum shown in Table 1, and the transmission accuracy degradation data is collected every 16 hours. The timing truncation test is carried out, and the sampling point is 19. All sampling values ​​of the harmonic reducer transmission error are recorded, such as Figure 8 shown.

[0128] Based on the above degradation data, the unknown parameters in the reliability function proposed in this application are estimated. According to the method provided in the previous article, the unknown parameter estimation results are substituted into the above reliability function. At this time, if you want to predict the service life of this harmonic reducer under rated torque, you can substitute the rated load torque 87Nm into the load torque T , bring the threshold 34 into the degradation threshold W , bring the torque mutation value 87Nm into the torque mutation value Δ T , we can conclude Fig. 9 The reliability life curve shown in the figure shows that when the reliability is 50%, the life of the harmonic reducer is about 500h. If you want to predict the reliability of this harmonic reducer under a load torque of 100Nm, you can substitute 100Nm into the load torque.T , bring the threshold 34 into the degradation threshold W , put the torque mutation value 100Nm into the torque mutation value Δ T The reliability function is obtained, and Fig.10 From the reliability life curve shown, we can see that when the reliability is 50%, the reducer life is about 400h. Similarly, this model can be used to calculate the reliability function at any torque size within the allowable torque range and obtain the corresponding life, so as to complete the life prediction of the harmonic reducer.

[0129] The above description is only a preferred embodiment of the present application and an explanation of the technical principles used. Those skilled in the art should understand that the scope of the invention involved in the present application is not limited to the technical solution formed by a specific combination of the above technical features, but should also cover other technical solutions formed by any combination of the above technical features or their equivalent features without departing from the inventive concept. For example, the above features are replaced with (but not limited to) technical features with similar functions disclosed in the present application.

Claims

1. A method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test, characterized in that: The following steps are involved: Acquire a set of load information of the reducer to be tested; the set of load information includes sampling time and corresponding load torque data; Constructing a stress loading spectrum according to the set of loading information; The stress loading spectrum at least includes a plurality of loading data and corresponding loading time and stress level; Performing an acceleration test based on the stress loading spectrum, and collecting degradation data of the reducer to be tested; the degradation data at least includes a plurality of operating times and corresponding transmission error values; Constructing an accelerated degradation model, and deriving a reliability function based on the accelerated degradation model; the accelerated degradation model is used to describe the degradation process of the reducer; determining unknown parameters in the reliability function according to the degradation data; Obtaining stress data of the reducer to be tested under normal working conditions, and substituting the stress data into a reliability function to calculate the reliability; Plotting a reliability curve showing the reliability changing over time, and obtaining a median life when the reliability is 50% in the reliability curve as the predicted life of the reducer to be tested; Constructing an accelerated degradation model includes the following steps: Obtain power-law and generalized log-linear models; Introducing a shock influence factor into the power law model, and constructing a transfer function based on the generalized log-linear model; Based on the power law model with the impact factor and the transfer function, an accelerated degradation model is obtained; The impact factor is introduced on the basis of the power law model, which characterizes the impact of the impact load on the degradation rate of the harmonic reducer under the accelerated test state. The degradation rate is calculated according to the following formula: ; in, is the degradation rate, T is the load torque, k is the number of times the impact load occurs, A , e are the parameters to be estimated, μ is the impact factor; By substituting different load torques T , the degradation rate of the harmonic reducer under different stress levels can be calculated; Then, based on the generalized log-linear model, a transfer function is established, which can convert the impact load borne by the harmonic reducer under the accelerated test state into a degradation mutation value. By substituting different stress change values, the degradation mutation values ​​under different stress levels can be calculated; Specifically, the degradation mutation value is calculated according to the following formula: ; ; ; in, h is the damage value caused by a single impact load, , are the unknown parameters in the formula, is the torque mutation value, is the mutation value of degradation caused by a single impact load, is the conversion coefficient; Based on the above-mentioned degradation rate calculation formula and degradation mutation value calculation formula, an accelerated degradation model is constructed, which is used to characterize the performance degradation of the harmonic reducer; ; in, X ( t ) is in t The transmission error degradation value at the moment, σ is the diffusion coefficient, B ( t ) is the standard Brownian motion, H is the period of arrival of periodic shocks; The reliability function is expressed according to the following formula: ; in, R ( t ) is the reliability function, W is the degradation threshold, is the cumulative distribution function expression of the standard normal distribution, In x is a random variable.

2. The method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test according to claim 1 is characterized in that: Obtaining a set of load information of the reducer to be tested includes the following steps: The load torque data of the reducer to be tested during service is collected at a preset sampling frequency and a preset number of sampling points, and a load information set is generated according to the sampling time and the corresponding load torque data.

3. The method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test according to claim 2 is characterized in that: Constructing a stress loading spectrum according to the loaded information set specifically includes the following steps: Acquire load data corresponding to at least one load cycle in the load information set, and generate a first load curve according to the load data; removing noise data in the first load curve and performing smoothing processing on the first load curve to obtain a second load curve; dividing the second load curve into a plurality of stress levels according to a preset data interval; Obtaining a load mean value corresponding to each of the stress levels, and expanding each load mean value to obtain an expanded load value; calculating an equivalent torque based on the expanded load value, wherein the equivalent torque is equal to a maximum average load torque allowed by the reducer to be tested; Eliminate the enlarged load value that is greater than 2 times the rated torque of the reducer to be tested to obtain the final load data; A stress loading spectrum is generated based on the final loading data.

4. The method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test according to claim 3 is characterized in that: The expanded load value is calculated according to the following formula: ; in, The maximum allowable average load torque of the reducer obtained from the product manual. is the time corresponding to the first stress level, is the time corresponding to the second stress level, is the time corresponding to the mth stress level, n 1 is the reducer driving speed corresponding to the first stress level, which is the rated speed of the reducer. n 2 is the reducer driving speed corresponding to the second stress level, which is the rated speed of the reducer. n m is the reducer driving speed corresponding to the mth stress level, which is taken as the rated speed of the reducer. T 1 is the load value after the load mean is enlarged corresponding to the first stress level, T 2 is the load value after the load mean is enlarged corresponding to the second stress level. T m For the m Each stress level corresponds to the load value after the load mean is expanded. m is the number of stress levels.

5. The method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test according to claim 1 is characterized in that: Performing an accelerated test based on the stress loading spectrum and collecting degradation data of the reducer to be tested specifically includes the following steps: The reducer to be tested is loaded according to the loading data in the stress loading spectrum, and degradation data is collected at preset intervals.

6. The method for predicting the life of a heavy-duty industrial robot reducer based on an acceleration test according to claim 1 is characterized in that: After constructing the stress loading spectrum according to the load information set and before performing the accelerated test based on the stress loading spectrum, the following steps are also included: Check whether the reducer to be tested meets the preset conditions; the preset conditions at least include: the appearance is intact and the lubricating material meets the preset standards and there is no leakage; When the reducer to be tested meets the preset conditions, an acceleration test is performed based on the stress loading spectrum.

Citation Information

Patent Citations

  • Device and method for testing service life of precision cycloidal speed reducer of robot

    CN107991095A

  • Method for predicting service life of flexible gear of harmonic reducer

    CN110889178A

  • Speed reducer fatigue acceleration test and residual life prediction method and storage processing system

    CN113076648A

  • Speed reducer accelerated life test method for industrial robot

    CN113834652A

  • Harmonic reducer accelerated life test method based on small sample data

    CN114757033A