Data set construction method for gear life evaluation, gear life evaluation method and gear life evaluation system

Abstract

The invention discloses a data set construction method for gear service life evaluation and a gear service life evaluation method and system.The data set construction method comprises the steps that multiple gears in a gear box are combined in pairs to form multiple gear sets, and the gear sets with the two gears high in operation similarity are selected through first-time screening; each gear set with high similarity of two gears in use is screened out through the second screening, and each gear set with high similarity of life-related characteristics of two gears is screened out through the third screening. Taking the service life data and the service life related characteristic data of each gear set screened out for the third time as a data set for evaluating the service life of the gear; in the gear life evaluation method, a neural network is trained by using a data set for gear life evaluation; the system is used for realizing a data set construction method for gear life evaluation and a gear life evaluation method. The gear service life evaluation efficiency and accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to the field of gear life assessment, and specifically to a method for constructing a data set for gear life assessment, a gear life assessment method, and a system. Background Art

[0002] In reality, the key role played by gears in mechanical transmission systems and their various advantages have promoted the development of many technical industries. However, there are still many deficiencies in gear life assessment.

[0003] In the prior art, when assessing the life of gears, the most considered is to achieve gear life assessment through relevant data such as gear wear. One problem with this assessment method is that the accuracy of gear life assessment is not high. There is no in-depth analysis of life-related characteristics during gear life assessment, resulting in an inaccurate and incomplete assessment model or method. More importantly, in the application of gears, gears need to perform multiple work combinations. However, this assessment method does not consider the number of gears during gear life assessment, resulting in poor accuracy.

[0004] Therefore, due to the lack of classified combination assessment of gears in the prior art, the efficiency and accuracy of gear life assessment cannot be effectively improved. Summary of the Invention

[0005] The present invention provides a method for constructing a data set for gear life assessment, a gear life assessment method, and a system to solve the problems of low efficiency and poor accuracy existing in the prior art gear life assessment method.

[0006] To achieve the above object, the technical solutions adopted by the present invention are as follows:

[0007] The method for constructing a data set for gear life assessment is as follows:

[0008] Any one of the multiple gears in the gearbox is combined with each of the other gears in pairs to form multiple gear groups; obtain multiple types of operation data sets of each of the two gears in each gear group during the historical operation period, and perform similarity analysis based on the multiple types of operation data sets of the two gears in each gear group during the historical operation period to screen out each gear group with high running similarity between the two gears included, thereby completing the first round of screening;

[0009] Obtain the usage duration data of each time of the two gears in each gear group screened in the first round during the historical operation period, and perform usage similarity analysis based on the usage duration data of each time of the two gears in each gear group screened in the first round during the historical operation period to screen out each gear group with high usage similarity between the two gears included, thereby completing the second round of screening;

[0010] Obtain the life data of each of the two gears in each gear set selected in the second round during their respective historical operating cycles, as well as multiple characteristic data of each of the two gears in each gear set selected in the second round during their respective historical operating cycles. Conduct a correlation analysis based on the life data and multiple characteristic data of each gear in each gear set selected in the second round during their respective historical operating cycles. Screen out several life-related characteristic data from the multiple characteristic data of each gear in each gear set according to the results of the correlation analysis. Then, conduct a life-related characteristic similarity analysis based on the life-related characteristic data of the two gears in each gear set to screen out each gear set with a high similarity of life-related characteristics of the two gears included, thereby completing the third round of screening;

[0011] Finally, use the life data and several life-related characteristic data of each gear in each gear set obtained from the third round of screening as the gear life evaluation dataset.

[0012] Furthermore, during the first round of screening, compare the operating data in multiple types of operating data groups of the two gears in each gear set. Mark the corresponding operating data groups with equal operating data as the same operating group, and mark the corresponding operating data groups with unequal operating data as the operating deviation group;

[0013] Count the number of same operating groups, and perform a ratio process on the number of same operating groups and the number of operating data groups to obtain the same operating value of each gear set;

[0014] After standardizing the operating data in each operating deviation group, perform a difference process on the operating data in the standardized operating deviation group and take the absolute value to obtain the operating data difference corresponding to each operating deviation group. Then, sum and average the operating data differences of all operating deviation groups to obtain the deviation operating value of each gear set;

[0015] Based on the same operating value and deviation operating value of each gear set, calculate the operating similarity value of each gear set;

[0016] Finally, according to the operating similarity values of each gear set, screen out each gear set with a high operating similarity of the two gears included.

[0017] Furthermore, compare the operating similarity value of each gear set with a preset operating similarity threshold, and select the gear sets with an operating similarity value greater than or equal to the operating similarity threshold as the gear sets with a high operating similarity of the two gears included.

[0018] Further, during the second-round screening, sum up the usage durations of each pair of gears in each gear set over the historical operation cycle to obtain the total usage duration of the two gears in the historical operation cycle; take the absolute value after performing a difference operation on the total usage durations of the two gears in the historical operation cycle to obtain the total usage duration deviation of the two gears; perform a ratio operation on the total usage duration deviation of the two gears and the duration corresponding to the historical operation cycle to obtain the total usage deviation value of the corresponding gear set.

[0019] Perform a ratio operation on the total usage duration of the two gears in each gear set over the historical operation cycle and the total number of times the two gears are used in the historical operation cycle to obtain the average usage duration of the two gears in the historical operation cycle; take the absolute value after performing a difference operation on the average usage durations of the two gears in the historical operation cycle to obtain the average usage duration difference of the two gears; perform a ratio operation on the average usage duration difference of the two gears and the duration corresponding to the historical operation cycle to obtain the average usage deviation value of the corresponding gear set.

[0020] Perform a difference operation on the start time point of each use of the two gears in each gear set and the end time point of the previous use of the two gears to obtain the single-use interval duration of the two gears; sum up and average the single-use interval durations of the two gears to obtain the average interval duration of the two gears; perform a difference operation on the average interval durations of the two gears and take the absolute value of the difference to obtain the average interval duration difference of the two gears; perform a ratio operation on the average interval duration difference of the two gears and the duration corresponding to the historical operation cycle of the two gears to obtain the usage interval deviation value of the corresponding gear set.

[0021] Calculate the usage similarity value of each gear set based on the total usage deviation value, average usage deviation value, and usage interval deviation value of each gear set.

[0022] Finally, based on the usage similarity values of each gear set, screen out each gear set with a high similarity in the life-related characteristics of the two gears it contains.

[0023] Further, compare the usage similarity value of each gear set with a preset usage similarity threshold respectively. If the usage similarity value of the gear set is less than the usage similarity threshold, then the gear set is regarded as a gear set with a high usage similarity of the two gears it contains.

[0024] Further, during the third-round screening, compare the life-related characteristic data of the two gears in each gear set, and mark the same life-related characteristic data as the same life-related characteristic.

[0025] Count the number of the same life-related characteristics obtained by comparing each gear set, and perform a ratio operation with the total number of life-related characteristics of the gear set to obtain the life characteristic similarity value of each gear set.

[0026] Finally, according to the same value of the life characteristics of each gear set, each gear set with a high similarity of the two gear life-related characteristics included is screened out.

[0027] Furthermore, the same value of the life characteristics of each gear set is compared with a preset same threshold of life characteristics, and the gear set with the same value of life characteristics greater than or equal to the same threshold of life characteristics is taken as the gear set with a high similarity of the two gear life-related characteristics included.

[0028] Furthermore, in the third round of screening, the Pearson correlation coefficient method is used to perform correlation analysis and calculation on the life data and each characteristic data of each gear in each gear set during the historical operation period, and the characteristic correlation value of each characteristic data of the gear is obtained. After taking the absolute value of the characteristic correlation value of each characteristic data, it is only compared with the preset threshold, and the characteristic data corresponding to the characteristic correlation value greater than the preset threshold is taken as the life-related characteristic data of the gear.

[0029] A gear life evaluation method constructs a data set according to the above-mentioned method for constructing a data set for gear life evaluation; then trains a neural network model with the data set to obtain a life evaluation prediction model; finally, inputs the life-related characteristic data of the target gear into the life evaluation prediction model, and the life evaluation prediction model predicts and evaluates the life of the target gear.

[0030] A gear life evaluation system includes a combined primary screening module, a secondary screening module, a final screening module, and a life evaluation module, where:

[0031] The combined primary screening module performs the first screening in the above-mentioned method for constructing a data set for gear life evaluation;

[0032] The secondary screening module performs the second screening in the above-mentioned method for constructing a data set for gear life evaluation;

[0033] The final screening module performs the third screening in the above-mentioned method for constructing a data set for gear life evaluation and obtains a data set;

[0034] The life evaluation module generates a neural network model, trains the neural network model with the data set to obtain a life evaluation prediction model, and the life evaluation module obtains the life-related characteristic data of the target gear and inputs it into the life evaluation prediction model, thereby predicting and evaluating the life of the target gear.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] 1. After classifying the gears into groups, based on the similarity analysis of operating data such as the operating temperature and load of the gears, a first-round screening of the gear groups is carried out in terms of the operating similarity of the gears. After the first-round screening, through the deviation analysis of the service life data of the gears in the historical operating cycle, a second-round screening of the gear groups is carried out in terms of the long service life of the gears. By screening the gear groups, the present invention is conducive to finding gears with high similarity, which is conducive to subsequent unified life assessment and thus improves the efficiency of gear life assessment.

[0037] 2. According to the real-time operating data of the gears in the historical operating cycle and the actual life records of the gears, the life-related characteristics of the gears are extracted, and through the comparison and analysis of the life-related characteristics of the gears in the gear groups, the gear groups are further screened. A life assessment model is constructed based on the life-related characteristics of the gears in the gear groups after screening. While improving the efficiency of gear life assessment, the present invention improves the accuracy of gear combination assessment through the comparison of life-related characteristics. Brief Description of the Drawings

[0038] Figure 1 is a flowchart of the steps of the method for constructing a dataset for gear life assessment according to an embodiment of the present invention.

[0039] Figure 2 is a block diagram of a gear life assessment system according to an embodiment of the present invention. Detailed Embodiments

[0040] In order to enable those skilled in the art to better understand the solution of the present invention, the following will describe in detail the embodiments of the present invention in conjunction with the drawings and embodiments, so as to fully understand the implementation process of how the present invention uses technical means to solve technical problems and achieve corresponding technical effects and be implemented accordingly. The embodiments of the present invention and each feature in the embodiments can be combined with each other without conflict, and the formed technical solutions are all within the protection scope of the present invention.

[0041] Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0042] It should be noted that the terms "including" and "having" in the specification and claims of the present invention and any variations thereof are intended to cover non-exclusive inclusion.

[0043] Embodiment 1

[0044] As Figure 1 shown, this embodiment discloses a method for constructing a dataset for gear life assessment, including the following steps:

[0045] Step 1. In this embodiment, for a gearbox with multiple gears, each gear in the gearbox has historically worked in pairs with every other gear. Each gear is combined with every other gear in pairs to form multiple gear groups. Suppose there are gears numbered 1, 2, 3, 4... n in the gearbox, where n represents the gear number. For the gear numbered 1, the numbered combinations of the gear groups of this gear are (1-2), (1-3), (1-4)... (1-n). Moreover, if two gears have already formed a gear group, they will not be grouped repeatedly. For example, if the gear numbered 1 forms the gear group (1-2), then when the gear numbered 2 is combined, the gear group (2-1) will not be formed again.

[0046] Obtain various operation data groups of each of the two gears in each gear group during the historical operation period. In this embodiment, the various operation data groups of each gear include, but are not limited to, the average temperature operation data group, the average humidity operation data group, the average rotation speed operation data group, and the average load (torque) operation data group. Specifically, sensors with corresponding functions can be installed in the gearbox to monitor and obtain multiple operation data of each gear under various conditions. The sensors include, but are not limited to, temperature sensors, humidity sensors, rotation speed sensors, and torque sensors.

[0047] Perform similarity analysis based on the various operation data groups of the two gears in each gear group during the historical operation period to screen out each gear group with a high running similarity between the two gears included, thereby completing the first round of screening. The specific process is as follows:

[0048] First, compare the operation data in the various operation data groups of the two gears in each gear group in the same category, that is, compare the operation data in the average temperature operation data group of the two gears with each other, compare the operation data in the average humidity operation data group of the two gears with each other, compare the operation data in the average rotation speed operation data group of the two gears with each other, and compare the operation data in the average load operation data group of the two gears with each other. Mark the corresponding operation data groups with equal operation data as the operation same group, and mark the corresponding operation data groups with unequal operation data as the operation deviation group. For example, if the operation data in the average temperature operation data group of the two gears are equal, then mark the average temperature data groups of the two gears as the operation same group; if the operation data in the average rotation speed operation data group of the two gears are not equal, then mark the average rotation speed data groups of the two gears as the operation deviation group.

[0049] Count the number A of the operation same groups, and perform a ratio process on the number A of the operation same groups and the number B of the operation data groups to obtain the same operation value XT = A / B of each gear group.

[0050] Standardize the operation data within each operation deviation group to obtain the standardized operation data Xi within each operation deviation group. The standardization formula is as follows:

[0051]

[0052] Where: x represents the initial operation data within the operation deviation group, μ represents the mean of the operation data within the operation deviation group, and δ represents the standard deviation of the operation deviation group.

[0053] Then, perform a difference operation on the operation data within the standardized operation deviation group and take the absolute value to obtain the operation data difference corresponding to each operation deviation group. Then sum and average the operation data differences of all operation deviation groups to obtain the deviation operation value PC for each gear group. For example, assume that all existing operation deviation groups are (X1, X2), (Y1, Y2)......(Z1, Z2), where 1 and 2 are the numbers of the two gears included in each gear group. Then the calculation formula for the deviation operation value PC is as follows:

[0054]

[0055] Where, n is the gear number; |X1 - X2| represents the operation data difference of the operation deviation group (X1, X2), |Y1 - Y2| represents the operation data difference of the operation deviation group (Y1, Y2), and |Z1 - Z2| represents the operation data difference of the operation deviation group (Z1, Z2).

[0056] Then, based on the same operation value XT and the deviation operation value PC of each gear group, calculate the operation similarity value YX through summation;

[0057] Finally, according to the operation similarity values YX of each gear group, compare the operation similarity value YX of each gear group with a preset operation similarity threshold. If the comparison result shows that the operation similarity value of a certain gear group is greater than or equal to the operation similarity threshold, then regard this gear group as the gear group with high running similarity of the two gears it contains and retain this gear group. If the comparison result shows that the operation similarity value of a certain gear group is less than the operation similarity threshold, then discard this gear group. Thus, through the first round of screening, each gear group with high running similarity of the two gears it contains is screened out.

[0058] In this embodiment, the meaning and significance of obtaining the operation similarity value YX are as follows: The operation similarity value YX is calculated from the same operation value XT and the deviation operation value PC. The same operation value XT reflects the degree of similarity of the operation data of the two gears included in the gear set, while the deviation operation value PC reflects the overall deviation degree between the operation data of the two gears. The larger the same operation value XT and the smaller the deviation operation value PC, the higher the similarity of the two gears in terms of operation. The purpose of obtaining the operation similarity value YX is to perform the first-round screening of the gear set based on the similarity of the two gears in terms of operation, facilitating the subsequent unified life assessment of the gears and improving the efficiency of gear life assessment.

[0059] Step 2: After the first-round screening of the gear set, obtain the data of the usage duration of each time during the historical operation cycle of the two gears in each gear set screened in the first round. Based on the data of the usage duration of each time during the historical operation cycle of the two gears in each gear set screened in the first round, perform usage similarity analysis to screen out each gear set with high usage similarity of the two gears included, thereby completing the second-round screening. The specific process is as follows:

[0060] First, through the historical usage record reports of the two gears in each gear set obtained after the first-round screening, obtain the start time point and end time point of each use during the historical operation cycle of the two gears. Perform a difference operation based on the start time point and end time point of each use of the two gears to obtain the usage duration of each time of the two gears in each gear set.

[0061] Sum up the usage durations of each time of the two gears in each gear set during the historical operation cycle to obtain the total usage duration of the two gears during the historical operation cycle; perform a difference operation on the total usage durations of the two gears during the historical operation cycle and then take the absolute value to obtain the deviation of the total usage durations of the two gears; perform a ratio operation on the deviation of the total usage durations of the two gears and the duration corresponding to the historical operation cycle to obtain the usage total deviation value of the corresponding gear set.

[0062] Perform a ratio operation on the total usage duration of the two gears in each gear set during the historical operation cycle and the total number of times of use of the two gears during the historical operation cycle to obtain the average usage duration of the two gears during the historical operation cycle. Among them, the total number of times of use of the two gears during the historical operation cycle is obtained by counting the number of start time points of each use of the two gears, and the total number of times of use of the two gears during the historical operation cycle is equal to the number of start time points of each use of the two gears.

[0063] Take the absolute value after taking the difference between the average usage durations of the two gears in each gear set during the historical operation cycle to obtain the average usage duration difference between the two gears; take the ratio of the average usage duration difference between the two gears to the duration corresponding to the historical operation cycle to obtain the average usage deviation value of the corresponding gear set.

[0064] Take the difference between the start time point of each use of the two gears in each gear set and the end time point of the previous use of the two gears to obtain the single-use interval duration between the two gears; sum up the single-use interval durations of the two gears and take the average to obtain the average interval duration between the two gears; take the difference between the average interval durations of the two gears and take the absolute value of the difference to obtain the average interval duration difference between the two gears; take the ratio of the average interval duration difference between the two gears to the duration corresponding to the historical operation cycle of the two gears to obtain the use interval deviation value of the corresponding gear set.

[0065] Multiply the total use deviation value, average use deviation value, and use interval deviation value of each gear set to calculate the use similarity value of each gear set.

[0066] Finally, according to the use similarity values of each gear set, compare the use similarity values of each gear set with a preset use similarity threshold respectively. When the use similarity value of a certain gear set is less than the use similarity threshold, then this gear set is used as a gear set with high similarity in the use of the two gears it contains, and this gear set is retained. When the use similarity value of a certain gear set is greater than or equal to the use similarity threshold, then this gear set is discarded. Thus, through the second-round screening, each gear set with high similarity in the life-related characteristics of the two gears it contains is screened out.

[0067] In this embodiment, the use similarity value reflects the similarity of the gears in the gear set in terms of total usage duration life, average usage duration life, and usage interval; the smaller the use similarity value, the higher the similarity. The purpose of understanding the similarity of the gears in the gear set in terms of total usage duration life, average usage duration life, and usage interval is to judge the similarity degree of the gears in the gear set in terms of operation and usage while being highly similar, which is beneficial to judging whether they can be combined and evaluated for life simultaneously, thereby improving the efficiency of gear life evaluation.

[0068] Step 3: Obtain the respective life data of each of the two gears in each gear set selected in the second round during their historical operation cycles, as well as multiple characteristic data of each of the two gears in each gear set selected in the second round during their historical operation cycles. Conduct a correlation analysis based on the life data and multiple characteristic data of each gear in each gear set selected in the second round during their historical operation cycles. According to the results of the correlation analysis, screen out several life-related characteristic data from the multiple characteristic data of each gear in each gear set. Then, conduct a life-related characteristic similarity analysis based on the life-related characteristic data of the two gears in each gear set to screen out each gear set with a high similarity in life-related characteristics of the two gears it contains, thereby completing the third round of screening.

[0069] In this embodiment, the historical operation cycle is divided into N cycle nodes, and the characteristic data values of each gear in each gear set at each cycle node are obtained and integrated into a characteristic data group (TZ1, TZ2, TZ3......TZN) for each gear, where TZN represents the characteristic data value of each gear at the Nth cycle node.

[0070] Specifically, in this embodiment, vibration signal data, acoustic emission signal data of each gear in each gear set during their historical operation cycles, and oil fluid chemical composition data in the gearbox are obtained through sensors. Characteristic data is extracted based on these data, and the obtained characteristic data includes, but is not limited to, vibration signal characteristic data, acoustic wave signal characteristic data, and oil fluid chemical composition characteristic data in the gearbox.

[0071] Among them, in this embodiment, by analyzing the vibration signal, the time-domain characteristics and frequency-domain characteristics of the vibration signal are extracted. The time-domain characteristics include statistical quantities such as the mean value, variance, peak value, root mean square value (RMS), margin, and skewness of the vibration signal. The frequency-domain characteristics are extracted by performing a Fourier transform on the vibration signal to extract frequency-domain characteristics such as the spectral centroid, spectral variance, gear meshing frequency and its harmonics, and sidebands.

[0072] In this embodiment, by capturing the tiny acoustic waves generated during the operation of the gear, the frequency, amplitude and other acoustic wave signal characteristics of the acoustic emission signal are analyzed.

[0073] In this embodiment, by detecting the metal particles and chemical composition changes in the gearbox oil fluid, the size, shape, quantity of wear particles in the oil fluid and the chemical composition of the oil fluid are analyzed as the oil fluid chemical composition characteristics.

[0074] In this embodiment, the life data values of each gear at each cycle node are obtained through actual life records and integrated into a life data group (SM1, SM2, SM3......SMN) for each gear, where SMN represents the life data value of each gear at the Nth cycle node.

[0075] In this embodiment, the Pearson correlation coefficient method is used to perform a correlation analysis and calculation on the life data and each characteristic data of each gear in each gear group during the historical operation period, and a characteristic correlation value R is obtained. The calculation formula is as follows:

[0076]

[0077] Where: N represents the number of characteristic data values (i.e., the total number of cycle nodes), TZi represents the i-th characteristic data value, and SMi represents the i-th life data value. represents the average value of the characteristic data group. represents the average value of the life data group.

[0078] After taking the absolute value of the characteristic correlation value R of each characteristic data and comparing it with the preset threshold, the characteristic data corresponding to the characteristic correlation value R greater than the preset threshold is taken as the life-related characteristic data of each gear.

[0079] In this embodiment, the life-related characteristic data of the two gears in each gear group are compared, and the same life-related characteristic data is marked as the same life-related characteristic. The number of the same life-related characteristics obtained by comparing each gear group is counted and ratio-processed with the total number of life-related characteristics of the gear group to obtain the life characteristic similarity value of each gear group. Among them, the total number of life-related characteristics corresponding to the gear group is obtained by summing up all the life-related characteristic cells corresponding to the gear group after removing duplicates.

[0080] According to the life characteristic similarity value of each gear group, the life characteristic similarity value of each gear group is compared with the preset life characteristic similarity threshold, and the gear group with the life characteristic similarity value greater than or equal to the life characteristic similarity threshold is taken as the gear group with high similarity of the life-related characteristics of the two gears included. Thus, through the third round of screening, each gear group with high similarity of the life-related characteristics of the two gears included is screened out.

[0081] In this embodiment, the life characteristic similarity value reflects the similarity degree of the corresponding life-related characteristics between the gears in the gear group. The advantage of understanding the similarity degree of the corresponding life-related characteristics between the gears is that it is beneficial to judge whether the gears included in the gear group can be combined and their lives can be evaluated simultaneously in the subsequent process, thereby improving the efficiency of gear life evaluation.

[0082] In this embodiment, finally, the life data and several life-related characteristic data of each gear in each gear group obtained by the third round of screening are used as the gear life evaluation dataset.

[0083] Embodiment 2

[0084] This embodiment discloses a method for evaluating the gear life. According to the method for constructing a data set for gear life evaluation described in Embodiment 1, a data set for gear life evaluation is constructed.

[0085] Then, using the life data of each gear and a number of life-related characteristic data in the data set for gear life evaluation, a neural network model is trained to obtain a life evaluation prediction model.

[0086] Finally, the life-related characteristic data of the target gear is input into the life evaluation prediction model, and the life of the target gear is predicted and evaluated by the life evaluation prediction model.

[0087] According to the real-time operation data of the gear during the historical operation cycle and the actual life record of the gear in this embodiment, the life-related characteristics of the gear are extracted, and through the comparison and analysis of the life-related characteristics of the gears in the gear set, the gear set is further screened. The life evaluation model is constructed through the life-related characteristics of the gears in the gear set after screening. The present invention improves the efficiency of gear life evaluation and improves the accuracy of gear combination evaluation through the comparison of life-related characteristics.

[0088] Embodiment 3

[0089] As Figure 2 shown, this embodiment discloses a gear life evaluation system, including a combined preliminary screening module, a secondary screening module, a final screening module, and a life evaluation module, where:

[0090] After the combined preliminary screening module divides the gears into gear sets, it performs a similarity analysis on the operation data of the gears in the gear set to perform the first screening in the method for constructing a data set for gear life evaluation described in Embodiment 1.

[0091] The secondary screening module obtains the service life data of the gears in the gear set retained in the first screening during the historical operation cycle, and based on the comparison and analysis results of the service life data of the gears during the historical operation cycle, performs the second screening in the method for constructing a data set for gear life evaluation described in Embodiment 1.

[0092] The final screening module extracts the life-related characteristics of the gears according to the real-time operation data of the gears in the gear set after the second round of screening during the historical operation cycle and the actual record of the gear life. Through the comparison and analysis of the life-related characteristics, the third screening in the method for constructing a data set for gear life evaluation described in Embodiment 1 is performed, thereby obtaining a data set for gear life evaluation.

[0093] The lifespan evaluation module generates a neural network model, trains the neural network model with the dataset for gear lifespan evaluation obtained by the final screening module to obtain a lifespan evaluation prediction model, and the lifespan evaluation module obtains the lifespan-related feature data of the target gear and inputs it into the lifespan evaluation prediction model, thereby predicting and evaluating the lifespan of the target gear.

[0094] The preferred embodiments of the present invention have been described in detail above in conjunction with the accompanying drawings. The embodiments described in the present invention are only descriptions of the preferred embodiments of the present invention, and do not limit the concept and scope of the present invention. Among the various specific technical features described in the above specific embodiments, they can be combined in any suitable manner without contradiction. As long as such a combination does not violate the idea of the present invention, it should also be regarded as the content disclosed in this disclosure. To avoid unnecessary repetition, the present invention does not separately describe various possible combination methods.

[0095] The present invention is not limited to the specific details in the above embodiments. Within the scope of the technical concept of the present invention and without departing from the design idea of the present invention, various modifications and improvements made by those skilled in the art to the technical solutions of the present invention should fall within the protection scope of the present invention. The technical content claimed by the present invention has been fully recorded in the claims.

Claims

1. A method for constructing a dataset for gear life assessment, characterized in that The process is as follows: Arbitrarily combine any one of the multiple gears in the gearbox with the other gears in pairs to form multiple gear sets; obtain multiple types of operation data sets of each of the two gears in each gear set during the historical operation period, and perform similarity analysis based on the multiple types of operation data sets of the two gears in each gear set during the historical operation period to screen out each gear set with high running similarity between the two gears included, thus completing the first round of screening; Obtain the usage duration data of each time of the two gears in each gear set screened in the first round during the historical operation period, and perform usage similarity analysis based on the usage duration data of each time of the two gears in each gear set screened in the first round during the historical operation period to screen out each gear set with high usage similarity between the two gears included, thus completing the second round of screening; Obtain the life data of each of the two gears in each gear set screened in the second round during the historical operation period, and multiple characteristic data of each of the two gears in each gear set screened in the second round during the historical operation period. Perform correlation analysis based on the life data and multiple characteristic data of each gear in each gear set screened in the second round during the historical operation period. Screen out several life-related characteristic data from the multiple characteristic data of each gear in each gear set according to the correlation analysis results. Then perform life-related characteristic similarity analysis based on the life-related characteristic data of the two gears in each gear set to screen out each gear set with high life-related characteristic similarity between the two gears included, thus completing the third round of screening; Finally, use the life data and several life-related characteristic data of each gear in each gear set obtained from the third round of screening as the data set for gear life assessment.

2. The method for constructing a data set for gear life evaluation according to claim 1, wherein, During the first round of screening, compare the operation data in the multiple types of operation data sets of the two gears in each gear set in the same category, mark the corresponding operation data sets with equal operation data as the operation same groups, and mark the corresponding operation data sets with unequal operation data as the operation deviation groups; Count the number of operation same groups, and perform a ratio process on the number of operation same groups and the number of operation data sets to obtain the same operation value of each gear set; After standardizing the operation data in each operation deviation group, perform a difference process on the operation data in the standardized operation deviation group and take the absolute value to obtain the operation data difference corresponding to each operation deviation group. Then sum and average the operation data differences of all operation deviation groups to obtain the deviation operation value of each gear set; Based on the same operation value and deviation operation value of each gear set, calculate the operation similarity value of each gear set; Finally, according to the operation similarity values of each gear set, screen out each gear set with high running similarity between the two gears included.

3. The method for constructing a data set for evaluating the gear life according to claim 2, wherein, Compare the operation similarity value of each gear set with a preset operation similarity threshold, and select the gear sets with the operation similarity value greater than or equal to the operation similarity threshold as the gear sets with high running similarity between the two gears included.

4. The method for constructing a data set for evaluating the gear life according to claim 1, characterized in that, During the second-round screening, sum up the usage durations of each pair of gears in each gear set over the historical operation period to obtain the total usage duration of the two gears in the historical operation period. After performing a difference operation on the total usage durations of the two gears in the historical operation period and taking the absolute value, obtain the total usage duration deviation of the two gears. Perform a ratio operation on the total usage duration deviation of the two gears and the duration corresponding to the historical operation period to obtain the total usage deviation value of the corresponding gear set. Perform a ratio operation on the total usage duration of the two gears in each gear set over the historical operation period and the total number of times the two gears are used in the historical operation period to obtain the average usage duration of the two gears in the historical operation period. After performing a difference operation on the average usage durations of the two gears in the historical operation period and taking the absolute value, obtain the average usage duration difference of the two gears. Perform a ratio operation on the average usage duration difference of the two gears and the duration corresponding to the historical operation period to obtain the average usage deviation value of the corresponding gear set. Perform a difference operation on the start time point of each use of the two gears in each gear set and the end time point of the previous use of the two gears to obtain the single-use interval duration of the two gears. Sum up the single-use interval durations of the two gears and take the average value to obtain the average interval duration of the two gears. Perform a difference operation on the average interval durations of the two gears and take the absolute value of the difference to obtain the average interval duration difference of the two gears. Perform a ratio operation on the average interval duration difference of the two gears and the duration corresponding to the historical operation period of the two gears to obtain the usage interval deviation value of the corresponding gear set. Calculate the usage similarity value of each gear set based on the total usage deviation value, average usage deviation value, and usage interval deviation value of each gear set. Finally, based on the usage similarity values of each gear set, screen out each gear set with a high similarity in the life-related characteristics of the two gears it contains.

5. The method for constructing a data set for gear life evaluation according to claim 4, characterized in that, Compare the usage similarity values of each gear set with a preset usage similarity threshold respectively. If the usage similarity values of a gear set are all less than the usage similarity threshold, then this gear set is regarded as a gear set with a high usage similarity of the two gears it contains.

6. The method for constructing a dataset for gear life assessment according to claim 1, wherein During the third-round screening, compare the life-related characteristic data of the two gears in each gear set, and mark the same life-related characteristic data as the same life-related characteristics. Count the number of the same life-related characteristics obtained by comparing each gear set, and perform a ratio operation with the total number of life-related characteristics of this gear set to obtain the life characteristic similarity value of each gear set. Finally, based on the life characteristic similarity values of each gear set, screen out each gear set with a high similarity in the life-related characteristics of the two gears it contains.

7. The method for constructing a data set for gear life assessment according to claim 6, wherein Compare the life characteristic similarity values of each gear set with a preset life characteristic similarity threshold, and select the gear sets with life characteristic similarity values greater than or equal to the life characteristic similarity threshold as the gear sets with a high similarity in the life-related characteristics of the two gears they contain.

8. The method for constructing a gear life evaluation data set according to claim 1 or 6, characterized in that In the third round of screening, the Pearson correlation coefficient method is used to perform correlation analysis and calculation on the life data and each characteristic data of each gear in each gear group during the historical operation period of the gear, so as to obtain the characteristic correlation value of each characteristic data of the gear. After taking the absolute value of the characteristic correlation value of each characteristic data, it is compared with a preset threshold, and the characteristic data corresponding to the characteristic correlation value greater than the preset threshold is taken as the life-related characteristic data of the gear.

9. A method for evaluating the life of a gear, characterized in that, According to the method for constructing a data set for gear life evaluation described in any one of claims 1-8, construct a data set; then use the data set to train a neural network model to obtain a life evaluation prediction model; finally, input the life-related characteristic data of the target gear into the life evaluation prediction model, and the life evaluation prediction model predicts and evaluates the life of the target gear.

10. A gear life evaluation system, characterized in that, It includes a combined preliminary screening module, a secondary screening module, a final screening module, and a life evaluation module, where: The combined preliminary screening module performs the first screening in the method for constructing a data set for gear life evaluation described in any one of claims 1-8; The secondary screening module performs the second screening in the method for constructing a data set for gear life evaluation described in any one of claims 1-8; The final screening module performs the third screening in the method for constructing a data set for gear life evaluation described in any one of claims 1-8 and obtains a data set; The life evaluation module generates a neural network model, uses the data set to train the neural network model to obtain a life evaluation prediction model, and the life evaluation module obtains the life-related characteristic data of the target gear and inputs it into the life evaluation prediction model, thereby predicting and evaluating the life of the target gear.

Images