Gear Random Fatigue Load Processing Method and System Based on GCM Counting Method

Abstract

The present invention relates to the technical field of load spectrum analysis, and discloses a method and system for processing random fatigue loads of gears based on the GCM counting method. The method includes: collecting actual operating load data and actual operating environment data of a gear transmission system, and converting torque into instantaneous contact stress; dividing it into a number of discrete small intervals based on discretization processing; performing cluster analysis on all discrete small intervals; constructing a two-dimensional gear counting matrix; converting non-symmetric cyclic stress into equivalent symmetric cyclic stress to determine the total gear life; comparing the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method to determine an adjustment coefficient, and determining a warning threshold according to the total gear life and the adjustment coefficient; comparing the total fatigue damage value with the warning threshold to determine the warning level. The present invention realizes the joint statistics of stress amplitude, mean stress and environmental characteristics, and overcomes the deficiencies of the rainflow counting method in dealing with meshing intermittent loads and lacking environmental adaptability.

Description

Technical Field

[0001] The present invention relates to the technical field of load spectrum analysis, and in particular to a method and system for processing random fatigue loads of gears based on a GCM counting method. Background Art

[0002] Gears, as core components in mechanical transmission systems, are often subjected to complex and variable random loads during their service life. This is particularly true in applications such as vehicles, wind power plants, and construction machinery. These gears are subject to long-term operation under multiple alternating operating conditions, non-constant torque, and frequent starts and stops, which can easily lead to contact fatigue failure. Accurately predicting the fatigue life of gears under these random loads is crucial for improving gear system reliability and enabling intelligent maintenance.

[0003] In the existing technology, the raindrop counting method is commonly used to count gear stress loads. This method is based on the peak-to-valley pairs of the stress-time curve. By identifying complete cycles and counting the stress amplitude and frequency, fatigue life is estimated in combination with the material SN curve. The raindrop counting method is suitable for components with continuous and regular stress changes. However, for gear pairs with intermittent meshing processes, the actual force on the gears during meshing is periodic and intermittent. The raindrop counting method treats stress as a continuous waveform, which easily overestimates the frequency of fatigue damage and leads to deviations in life assessment. It also does not include the actual impact of environmental variables such as temperature and humidity on the fatigue strength of the material, resulting in a lack of environmental adaptability in life prediction. It is unable to effectively separate and model the stress effects under different temperature and humidity environments, making it difficult to achieve working condition-sensitive life prediction and graded warning.

[0004] Therefore, it is necessary to design a gear random fatigue load processing method and system based on the GCM counting method to solve the problems existing in the current technology. Summary of the Invention

[0005] In view of this, the present invention proposes a gear random fatigue load processing method based on the GCM counting method, aiming to solve the current problems of ignoring the discreteness of the gear meshing process, not considering the operating environment factors, and lacking a partition modeling mechanism for multi-working condition fatigue damage.

[0006] In one aspect, the present invention proposes a method for processing random fatigue loads of gears based on the GCM counting method, comprising:

[0007] Collecting actual operating load data and actual operating environment data of the gear transmission system, wherein the actual operating load data is a torque-time variation curve, and the actual operating environment data includes temperature and humidity-time variation curves, and converting the torque in the torque-time variation curve into instantaneous contact stress on the gear tooth surface;

[0008] Based on discretization, the continuous contact stress is divided into several discrete small intervals, and each of the discrete small intervals corresponds to a time segment and corresponds to a temperature and humidity data;

[0009] Perform clustering analysis on all the discrete small intervals according to the temperature and humidity data to determine several clustering sets;

[0010] Collect the gear rotation speed and obtain the number of gear rotations within each of the clustering sets according to the gear rotation speed, record it as the occurrence frequency of this stress level, and construct a two-dimensional gear counting matrix, where the two-dimensional gear counting matrix includes stress amplitude, mean stress, and corresponding frequency;

[0011] Use the Goodman life curve to convert the asymmetric cyclic stress into an equivalent symmetric cyclic stress, and determine the total gear life according to the S-N curve of the gear material and the equivalent symmetric cyclic stress; Compare the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method, determine the adjustment coefficient according to the comparison result, and determine the warning threshold according to the total gear life and the adjustment coefficient;

[0012] Obtain the total fatigue damage value under the action of all levels of equivalent stress according to the two-dimensional gear counting matrix, compare the total fatigue damage value with the warning threshold, and determine the warning level according to the comparison result.

[0013] Further, when converting the torque in the torque-time change curve into the instantaneous contact stress of the gear tooth surface, it includes:

[0014] ;

[0015] where, represents the instantaneous contact stress of the tooth surface at time t, represents the circumferential force, represents the pitch diameter, represents the contact tooth width, represents the contact stress size coefficient, represents the elastic influence coefficient, represents the helix angle correction coefficient, represents the contact ratio correction coefficient, represents the applied load coefficient, represents the dynamic load coefficient, represents the load distribution coefficient in the tooth direction.

[0016] Further, when dividing the continuous contact stress into several discrete small intervals based on discretization, it includes:

[0017] The time interval of the discretization process is inversely proportional to the sampling frequencies of the torque sensor, temperature sensor, and humidity sensor.

[0018] Further, when performing clustering analysis on all the discrete small intervals according to the temperature and humidity data to determine several clustering sets, it includes:

[0019] Extract the temperature and humidity characteristics of each discrete small interval to form a two-dimensional point set;

[0020] Determine the initial neighborhood radius through the k-distance graph and set MinPts to 4; determine ε through the k-distance graph method;

[0021] Randomly select an unclassified point, find all the points within its ε neighborhood. If the number of points within the neighborhood is greater than or equal to MinPts, form a new class and continuously expand it; if the number of points within the neighborhood is less than MinPts, mark this point as an isolated point until all points are classified or marked.

[0022] Further, after performing clustering analysis on all the discrete small intervals according to the temperature and humidity data to determine several clustering sets, it further includes:

[0023] If the number of instantaneous contact stresses in the clustering set is at least two, determine the representative instantaneous contact stress of this clustering set based on kernel density estimation;

[0024] ;

[0025] Wherein, represents the representative instantaneous contact stress, represents the number of instantaneous contact stresses in the clustering set, represents the smoothing bandwidth, represents the value of a certain frequency point to be estimated, represents the i-th instantaneous contact stress in the clustering set.

[0026] Further, the smoothing bandwidth is calculated by the following formula:

[0027] ;

[0028] Wherein, represents the smoothing bandwidth, represents the standard deviation of the data in the clustering set, represents the data skewness, represents the data kurtosis, represents the adjustment coefficient.

[0029] Further, when converting the non-symmetric cyclic stress to the equivalent symmetric cyclic stress by using the Goodman life curve, it includes:

[0030] ;

[0031] Among them, represents the instantaneous contact stress characterizing the clustering set, represents the average stress of the clustering set, represents the ultimate tensile strength of the gear material, represents the corrected equivalent symmetric stress.

[0032] Furthermore, when comparing the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method and determining the adjustment coefficient according to the comparison result, and determining the warning threshold according to the total gear life and the adjustment coefficient, it includes:

[0033] Obtain a ratio coefficient according to the equivalent symmetric cyclic stress and the cyclic stress under the rainflow counting method, and the ratio coefficient is the ratio of the equivalent symmetric cyclic stress to the cyclic stress under the rainflow counting method;

[0034] When the ratio coefficient is less than or equal to 1, determine the adjustment coefficient according to the ratio coefficient. The adjustment coefficient is in a direct proportional relationship with the ratio coefficient, and the value range of the adjustment coefficient is (0.7, 0.9);

[0035] When the ratio coefficient is greater than 1, determine the adjustment coefficient according to the ratio coefficient. The adjustment coefficient is in a direct proportional relationship with the ratio coefficient, and the value range of the adjustment coefficient is (0, 0.7).

[0036] Furthermore, when comparing the total fatigue damage value with the warning threshold and determining the warning level according to the comparison result, it includes:

[0037] Obtain the remaining life of the gear according to the total fatigue damage value and the total gear life, and obtain the life difference according to the remaining life of the gear and the warning threshold. The life difference is the difference between the remaining life of the gear and the warning threshold;

[0038] Determine the warning level according to the life difference, and the warning level is in an inverse proportional relationship with the life difference.

[0039] Compared with the prior art, the beneficial effects of the present invention are as follows: By means of the gear random fatigue load processing method based on the GCM counting method, the actual operating load is combined with environmental factors (temperature, humidity), and discrete and clustering analysis means are used to perform environmental perception modeling on stress segments under different working conditions, realizing the joint statistics of stress amplitude, mean stress and environmental characteristics, and inversely calculating the fatigue frequency under each clustering working condition through the gear speed to establish a multi-dimensional gear counting matrix, thereby overcoming the deficiencies of the rainflow counting method in dealing with meshing intermittent loads and lacking environmental adaptability; By correcting the asymmetric cyclic stress through the Goodman life curve, evaluating the total gear life in combination with the S-N curve, and then forming an adjustment coefficient by comparing with the results of the rainflow method, a dynamic and refined warning threshold is set, making the fatigue damage assessment more in line with the actual working conditions and having higher reliability and intelligent warning capabilities.

[0040] On the other hand, the present application also provides a gear random fatigue load processing system based on the GCM counting method for applying the above-mentioned gear random fatigue load processing method based on the GCM counting method, including:

[0041] An acquisition unit configured to acquire the actual operating load data and actual operating environment data of the gear transmission system, the actual operating load data being a torque-time change curve, and the actual operating environment data including temperature and humidity-time change curves, and converting the torque in the torque-time change curve into the instantaneous contact stress of the gear tooth surface;

[0042] A first processing unit configured to divide the continuous contact stress into a plurality of discrete small intervals based on discretization processing, each of the discrete small intervals corresponding to a time segment and corresponding to a temperature and humidity data;

[0043] An analysis unit configured to perform clustering analysis on all the discrete small intervals according to the temperature and humidity data to determine a plurality of clustering sets;

[0044] A second processing unit acquires the gear speed and obtains the number of gear rotations within each clustering set according to the gear speed, records it as the occurrence frequency of the stress level of the clustering set, and constructs a two-dimensional gear counting matrix, the two-dimensional gear counting matrix including stress amplitude, mean stress and corresponding frequency;

[0045] A third processing unit configured to convert the asymmetric cyclic stress into an equivalent symmetric cyclic stress by using the Goodman life curve, determine the total gear life according to the S-N curve of the gear material and the equivalent symmetric cyclic stress; compare the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method, determine an adjustment coefficient according to the comparison result, and determine a warning threshold according to the total gear life and the adjustment coefficient;

[0046] An early warning unit, configured to obtain the total fatigue damage value under the action of all levels of equivalent stress according to the two-dimensional gear counting matrix, compare the total fatigue damage value with the early warning threshold, and determine the early warning level according to the comparison result.

[0047] It can be understood that the above-mentioned gear random fatigue load processing method and system based on the GCM counting method have the same beneficial effects, which will not be elaborated here. Description of the Drawings

[0048] By reading the following detailed description of the preferred embodiments, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of showing the preferred embodiments and are not considered as limiting the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0049] Figure 1 is a flowchart of a gear random fatigue load processing method based on the GCM counting method provided by an embodiment of the present invention;

[0050] Figure 2 is a functional block diagram of a gear random fatigue load processing system based on the GCM counting method provided by an embodiment of the present invention. Detailed Embodiments

[0051] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0052] In some embodiments of the present application, referring to Figure 1 as shown, a gear random fatigue load processing method based on the GCM counting method includes:

[0053] S100: Collect the actual operating load data and actual operating environment data of the gear transmission system. The actual operating load data is a torque-time change curve, and the actual operating environment data includes temperature and humidity-time change curves. Convert the torque in the torque-time change curve into the instantaneous contact stress on the gear tooth surface.

[0054] S200: Based on discretization processing, divide the continuous contact stress into a number of discrete small intervals. Each discrete small interval corresponds to a time segment and corresponds to a temperature and humidity data.

[0055] S300: Perform cluster analysis on all discrete small intervals according to the temperature and humidity data to determine several clustering sets.

[0056] S400: Collect the gear rotation speed and obtain the number of gear rotation cycles within each clustering set according to the gear rotation speed. Record it as the occurrence frequency of the stress level of the clustering set and construct a two-dimensional gear counting matrix. The two-dimensional gear counting matrix includes stress amplitude, mean stress, and corresponding frequency.

[0057] S500: Use the Goodman life curve to convert the asymmetric cyclic stress into an equivalent symmetric cyclic stress, and determine the total gear life according to the S-N curve of the gear material and the equivalent symmetric cyclic stress. Compare the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method, determine the adjustment coefficient according to the comparison result, and determine the warning threshold according to the total gear life and the adjustment coefficient.

[0058] S600: Obtain the total fatigue damage value under the action of all levels of equivalent stress according to the two-dimensional gear counting matrix, compare the total fatigue damage value with the warning threshold, and determine the warning level according to the comparison result.

[0059] Specifically, in step S100, the actual operating load data and operating environment data of the gear transmission system are collected. The operating load data is a torque-time variation curve, and the environment data includes temperature-time and humidity-time variation curves. Through the gear contact stress conversion model, the time-varying torque is converted into the instantaneous contact stress of the gear tooth surface, forming the contact stress time history. In step S200, the contact stress time series is discretized. The continuous stress history is divided into multiple small intervals, and each small interval is associated with an independent time segment, and then bound to the temperature and humidity data at the corresponding time points, preparing the data basis for subsequent environmental clustering. In step S300, according to the temperature and humidity data corresponding to each time segment, a two-dimensional environmental feature vector is constructed. The density-based clustering analysis algorithm (such as DBSCAN) is used to cluster all small intervals, and the clustering sets with similar environmental characteristics are identified. Each set represents a typical operating environment condition. In step S400, the rotational speed information of the gear is collected in real time, and the number of gear rotations within each clustering set is calculated according to the rotational speed and the length of each time segment. The clustering result is combined with the contact stress amplitude and the mean stress to construct a two-dimensional gear count matrix (Gear Count Matrix, GCM), where the matrix contains each stress amplitude, mean stress, and their corresponding frequencies. In step S500, the Goodman life curve is used to correct the asymmetric stress cycle, convert it into an equivalent symmetric cycle stress, and based on the constructed equivalent stress and the S-N curve (stress-life curve) of the gear material, determine the total life of the gear under the current load spectrum. At the same time, the equivalent stress obtained by the GCM method is compared with the stress cycle parameters obtained by the traditional rainflow counting method, the differences between the two are analyzed, and an adjustment coefficient is obtained to correct the environmental or modeling deviation in the life assessment result, and then determine the fatigue damage warning threshold. In step S600, according to the two-dimensional gear count matrix, the fatigue damage values under each stress level are statistically calculated, and summed up according to the Miner linear cumulative damage theory to obtain the total fatigue damage value. The total damage value is compared with the adjusted warning threshold, and then the fatigue risk level of the current gear state is determined, and the corresponding warning result is output to realize the monitoring of the fatigue deterioration trend and the management of the remaining life.

[0060] It can be understood that by combining the load data actually collected during the gear service process with environmental information such as temperature and humidity, and using the clustering analysis and the frequency statistics method based on the rotational speed, a two-dimensional gear count matrix adaptable to multiple working conditions is constructed, which truly reflects the stress evolution characteristics under different environments. Through the adjustment coefficient obtained by the Goodman life correction, the comparison between GCM and the rainflow counting method, the dynamic correction and adaptive optimization of the fatigue life assessment model are realized, and the life prediction accuracy and warning rationality of the gear system under random loads and variable working conditions are improved.

[0061] In some embodiments of the present application, when converting the torque in the torque-time change curve into the instantaneous contact stress of the gear tooth surface, it includes:

[0062] .

[0063] Wherein, represents the instantaneous contact stress of the tooth surface at time t, represents the circumferential force, represents the pitch diameter, represents the contact tooth width, represents the contact stress size factor, represents the elastic influence coefficient, represents the helix angle correction coefficient, represents the contact ratio correction coefficient, represents the application load coefficient, represents the dynamic load coefficient, represents the load distribution coefficient in the tooth direction.

[0064] It can be understood that, wherein, , T(t) represents the torque at time t. The contact stress size factor reflects the influence of the gear size on the stress distribution; the elastic influence coefficient is related to parameters such as the elastic modulus and Poisson's ratio of the meshing gear pair; the helix angle correction coefficient is used to correct the deviation of the meshing state of helical gears (this value is 1 for spur gears); the contact ratio correction coefficient represents the adjustment effect of the actual number of meshing teeth on the contact area; the application load coefficient represents the correction caused by additional load fluctuations under actual working conditions; the dynamic load coefficient reflects the influence of fluctuations during the dynamic transmission of the load; the load distribution coefficient in the tooth direction is used to correct the uniformity of the load distribution along the tooth width. In this embodiment, by combining the structural geometry, material properties, and load correction coefficients, the contact stress time history borne by the gear during actual meshing can be more realistically reconstructed. Compared with the traditional method of simplifying the torque into a constant stress input, it can improve the physical accuracy and time resolution of the stress spectrum, provide more realistic basic data for subsequent fatigue life prediction, and further improve the accuracy and applicability of the entire GCM counting method processing system under complex and variable load conditions. It is applicable to gear fatigue modeling scenarios in typical high-load and variable-load environments such as automotive final drives and wind power main shaft gearboxes.

[0065] In some embodiments of the present application, when dividing the continuous contact stress into several discrete small intervals based on discretization processing, it includes: The time interval of the discretization processing is inversely proportional to the sampling frequencies of the torque sensor, temperature sensor, and humidity sensor.

[0066] Specifically, during the operation of the gear system, the contact stress, as a function derived from the torque change, needs to be discretized in its continuous time history and divided into multiple small segments for subsequent stress frequency statistics and correlation with environmental data. In practice, to ensure that the contact stress changes within the discrete intervals can be approximately considered constant, and to ensure that each small interval accurately corresponds to a set of real-time temperature and humidity data. The discretization time step of the stress spectrum is adjusted according to the sampling capabilities of various sensors, so that the sampled load-environmental data has a consistent time resolution, facilitating the construction of an accurate "stress-temperature-humidity" three-dimensional joint data matrix, providing a unified data structure support for subsequent cluster analysis and counting modeling.

[0067] It can be understood that driving the discrete time interval with the sensor sampling frequency can effectively coordinate the synchronization accuracy of load data and environmental data in the time dimension, avoid time deviation or stress-environment pairing errors caused by inconsistent sampling rates of different sensors, and thus improve the consistency of data processing and modeling accuracy.

[0068] In some embodiments of the present application, cluster analysis is performed on all discrete cells based on temperature and humidity data to determine a number of cluster sets, including:

[0069] The temperature and humidity characteristics of each discrete interval are extracted to form a two-dimensional point set.

[0070] The initial neighborhood radius is determined by the k-distance graph, and MinPts is determined to be 4. ε is determined by the k-distance graph method.

[0071] Randomly select an unclassified point and find all the points in its ε neighborhood. If the number of points in the neighborhood is greater than or equal to MinPts, form a new class and continue to expand. If the number of points in the neighborhood is less than MinPts, mark the point as an isolated point until all points are classified or marked.

[0072] In some embodiments of the present application, after cluster analysis is performed on all discrete small intervals based on temperature and humidity data and several cluster sets are determined, it also includes: if the number of instantaneous contact stresses in the cluster set is at least two, determining the characterizing instantaneous contact stress of the cluster set based on kernel density estimation.

[0073] .

[0074] in, represents the instantaneous contact stress, represents the number of instantaneous contact stresses in the cluster set, represents the smoothing bandwidth, Indicates the value of a frequency point to be estimated, Represents the instantaneous contact stress at the i-th moment in the clustering set.

[0075] In some embodiments of the present application, the smoothing bandwidth is calculated as follows:

[0076] .

[0077] Where, Represents the smoothing bandwidth, Represents the standard deviation of the data in the clustering set, Represents the data skewness, Represents the data kurtosis, Represents the adjustment coefficient.

[0078] Specifically, . . Where, Represents the mean value of the data in the clustering set.

[0079] It can be understood that the minimum point number parameter MinPts = 4 is set; the Euclidean distance from each point to its 4th nearest neighbor is calculated, and the 4th neighbor distances of all points are sorted in ascending order to form a k-distance curve; the neighborhood radius ε is determined according to the "inflection point position" of the curve. Subsequently, the clustering process is executed: a non-classified point is randomly selected; if the number of points contained in its ε-neighborhood is ≥ MinPts, then this point is used as a core point to create a new cluster and recursively expand this cluster; if the number of neighborhood points < MinPts, then this point is marked as an isolated point; this process is repeated until all points are classified or marked, and finally several clustering sets are output. Through the density-based clustering method and the kernel density estimation model, the influence of different temperature and humidity environments on the gear stress spectrum can be effectively distinguished, and statistically representative contact stress characteristics can be extracted in each clustering environment, avoiding stress estimation deviation caused by the interference of extreme values or outliers, and realizing more stable and accurate environmental condition stress modeling. When facing complex non-Gaussian distributed stress data, the flexibility and adaptability of stress density estimation are improved through the adaptive bandwidth control mechanism.

[0080] In some embodiments of the present application, when converting the asymmetric cyclic stress to the equivalent symmetric cyclic stress using the Goodman life curve, it includes:

[0081] .

[0082] Where, Represents the representative instantaneous contact stress of the clustering set, Represents the average stress of the clustering set, Represents the tensile ultimate strength of the gear material, Represents the corrected equivalent symmetric stress.

[0083] In some embodiments of the present application, when comparing the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method and determining the adjustment coefficient according to the comparison result, and determining the warning threshold according to the total gear life and the adjustment coefficient, it includes: obtaining a ratio coefficient according to the equivalent symmetric cyclic stress and the cyclic stress under the rainflow counting method, where the ratio coefficient is the ratio of the equivalent symmetric cyclic stress to the cyclic stress under the rainflow counting method.

[0084] Specifically, when the ratio coefficient is less than or equal to 1, the adjustment coefficient is determined according to the ratio coefficient. The adjustment coefficient is directly proportional to the ratio coefficient, and the value range of the adjustment coefficient is (0.7, 0.9). When the ratio coefficient is greater than 1, the adjustment coefficient is determined according to the ratio coefficient. The adjustment coefficient is directly proportional to the ratio coefficient, and the value range of the adjustment coefficient is (0, 0.7).

[0085] It can be understood that when the ratio coefficient is less than or equal to 1, it indicates that the stress predicted by the rainflow counting method is greater and the result is more conservative, and the adjustment coefficient can appropriately relax the warning threshold. When the ratio coefficient is greater than 1, it indicates that the stress predicted by the GCM method is greater and the life is shorter, and it is necessary to improve the warning sensitivity and set the adjustment coefficient to tighten the warning threshold. By correcting the asymmetric load stress to the equivalent symmetric stress, it is ensured that the stress parameters used in the fatigue life estimation are unified in the effective domain of the S-N curve, and the physical rationality of the stress-life matching is improved. On this basis, the comparative analysis with the result obtained by the rainflow counting method is introduced, and the dynamic adjustment coefficient is determined based on the stress ratio, so that the final warning threshold can take into account the differences and respective advantages of the two load modeling methods, and realize a more adaptive and working condition sensitive life warning mechanism. The accuracy of the fatigue assessment of the gear under multiple load sources and multiple working conditions is improved, and the response ability of the warning strategy to the actual damage evolution process is enhanced.

[0086] In some embodiments of the present application, when comparing the total fatigue damage value with the warning threshold and determining the warning level according to the comparison result, it includes: obtaining the remaining life of the gear according to the total fatigue damage value and the total gear life, obtaining the life difference according to the remaining life of the gear and the warning threshold, where the life difference is the difference between the remaining life of the gear and the warning threshold. The warning level is determined according to the life difference, and the warning level is inversely proportional to the life difference.

[0087] It is understandable that the life difference is obtained by comparing the remaining life with the set fatigue warning threshold. The fatigue warning threshold is usually calculated based on historical experience, safety margin or dynamic adjustment coefficient, which represents the minimum remaining life required to trigger the warning. On this basis, the warning level is divided according to the life difference, and it is preferred to set the warning level inversely proportional to the life difference. By introducing the life difference, a dynamic quantitative safety margin evaluation mechanism is established between the fatigue damage accumulation and the life threshold. It enhances the perception sensitivity and response flexibility of the gear fatigue risk, which can effectively extend the gear service life cycle and reduce the risk of failure downtime.

[0088] In the above-mentioned embodiment, a random fatigue load processing method for gears based on the GCM counting method combines actual operating loads with environmental factors (temperature and humidity). Discretization and cluster analysis are used to perform environmentally aware modeling of stress fragments under different operating conditions. This achieves joint statistics of stress amplitude, average stress, and environmental characteristics. The fatigue frequency under each clustered operating condition is inferred from gear speed, and a multidimensional gear counting matrix is established. This overcomes the shortcomings of the raindrop counting method in handling intermittent meshing loads and its lack of environmental adaptability. Asymmetric cyclic stresses are corrected using the Goodman life curve, and the total gear life is estimated using the SN curve. An adjustment coefficient is then generated by comparing the results with the raindrop method, enabling dynamic and refined early warning threshold setting. This makes fatigue damage assessment more consistent with actual operating conditions, providing higher reliability and intelligent early warning capabilities.

[0089] In another preferred embodiment based on the above embodiment, refer to Figure 2 As shown, this embodiment provides a gear random fatigue load processing system based on the GCM counting method, which is used to apply the above-mentioned gear random fatigue load processing method based on the GCM counting method, including:

[0090] The acquisition unit is configured to collect actual operating load data and actual operating environment data of the gear transmission system. The actual operating load data is a torque-time variation curve, and the actual operating environment data includes a temperature and humidity-time variation curve. The torque in the torque-time variation curve is converted into the instantaneous contact stress of the gear tooth surface.

[0091] The first processing unit is configured to divide the continuous contact stress into a plurality of discrete small intervals based on discretization processing, where each discrete small interval corresponds to a time segment and corresponds to a temperature and humidity data.

[0092] The analysis unit is configured to perform cluster analysis on all discrete cells according to the temperature and humidity data to determine a number of cluster sets.

[0093] The second processing unit collects the gear rotation speed and obtains the number of gear rotation cycles within each clustering set according to the gear rotation speed, records it as the occurrence frequency of the stress level of the clustering set, and constructs a two-dimensional gear counting matrix, where the two-dimensional gear counting matrix includes stress amplitude, mean stress, and corresponding frequency.

[0094] The third processing unit is configured to convert the asymmetric cyclic stress into an equivalent symmetric cyclic stress by using the Goodman life curve, and determine the total gear life according to the S-N curve of the gear material and the equivalent symmetric cyclic stress. Compare the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method, determine the adjustment coefficient according to the comparison result, and determine the warning threshold according to the total gear life and the adjustment coefficient.

[0095] The warning unit is configured to obtain the total fatigue damage value under the action of all levels of equivalent stress according to the two-dimensional gear counting matrix, compare the total fatigue damage value with the warning threshold, and determine the warning level according to the comparison result.

[0096] It can be understood that through the gear random fatigue load processing method based on the GCM counting method, the actual operating load is combined with environmental factors (temperature, humidity), and the stress segments under different working conditions are modeled for environmental perception by means of discretization and clustering analysis, realizing the joint statistics of stress amplitude, mean stress and environmental characteristics, and inversely deducing the fatigue frequency under each clustering working condition through the gear rotation speed, and establishing a multi-dimensional gear counting matrix, thus overcoming the deficiencies of the rainflow counting method in dealing with meshing intermittent loads and lacking environmental adaptability. The asymmetric cyclic stress is corrected by the Goodman life curve, and the total gear life is evaluated in combination with the S-N curve, and then the adjustment coefficient is formed by comparing with the results of the rainflow method, realizing the dynamic and refined setting of the warning threshold, making the fatigue damage assessment more in line with the actual working conditions, and having higher reliability and intelligent warning capabilities.

[0097] Those skilled in the art should understand that the embodiments of the present application can be provided as methods, systems, or computer program products. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0098] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of flows and / or blocks in the flowchart and / or block diagram can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate a device for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or a device for implementing the functions specified in multiple blocks.

[0099] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or a device for implementing the functions specified in multiple blocks.

[0100] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one flow Figure 1 one flow or multiple flows and / or blocks Figure 1 or a device for implementing the functions specified in multiple blocks.

[0101] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the specific implementation manners of the present invention can still be modified or equivalently replaced, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A method for processing random fatigue loads of gears based on the GCM counting method, characterized in that, Including: Collecting the actual operating load data and actual operating environment data of the gear transmission system, where the actual operating load data is a torque-time change curve, and the actual operating environment data includes temperature and humidity-time change curves, and converting the torque in the torque-time change curve into the instantaneous contact stress of the gear tooth surface; Based on discretization processing, dividing the continuous contact stress into several discrete small intervals, each of the discrete small intervals corresponding to a time segment and corresponding to a temperature and humidity data; Performing clustering analysis on all the discrete small intervals according to the temperature and humidity data to determine several clustering sets; Collecting the gear rotation speed and obtaining the number of gear rotations within each clustering set according to the gear rotation speed, recording it as the occurrence frequency of the stress level of the clustering set, and constructing a two-dimensional gear counting matrix, where the two-dimensional gear counting matrix includes stress amplitude, mean stress, and corresponding frequency; Converting the asymmetric cyclic stress into an equivalent symmetric cyclic stress by using the Goodman life curve, and determining the total gear life according to the S-N curve of the gear material and the equivalent symmetric cyclic stress; comparing the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method, determining an adjustment coefficient according to the comparison result, and determining an early warning threshold according to the total gear life and the adjustment coefficient; Obtaining the total fatigue damage value under the action of all levels of equivalent stress according to the two-dimensional gear counting matrix, comparing the total fatigue damage value with the early warning threshold, and determining the early warning level according to the comparison result.

2. The gear random fatigue load processing method based on the GCM counting method according to claim 1, wherein, When converting the torque in the torque-time change curve into the instantaneous contact stress of the gear tooth surface, it includes: ； Among them, represents the instantaneous contact stress of the tooth surface at time t, represents the circumferential force, represents the pitch diameter, represents the contact tooth width, represents the contact stress size factor, represents the elastic influence coefficient, represents the helix angle correction coefficient, represents the contact ratio correction coefficient, represents the application load factor, represents the dynamic load factor, represents the load distribution factor in the tooth direction.

3. The method for processing random fatigue load of gears based on the GCM counting method according to claim 1, characterized in that When dividing the continuous contact stress into several discrete small intervals based on discretization processing, it includes: The time interval of the discretization processing is inversely proportional to the sampling frequencies of the torque sensor, temperature sensor, and humidity sensor.

4. The method for processing random fatigue load of gears based on the GCM counting method according to claim 1, wherein When performing clustering analysis on all the discrete small intervals according to the temperature and humidity data to determine several clustering sets, it includes: Extracting the temperature and humidity characteristics of each discrete small interval to form a two-dimensional point set; Determining the initial neighborhood radius through the k-distance graph and determining MinPts as 4; determining ε through the k-distance graph method; Randomly selecting an unclassified point, finding all the points within its ε neighborhood. If the number of points within the neighborhood is greater than or equal to MinPts, forming a new class and continuously expanding; if the number of points within the neighborhood is less than MinPts, marking the point as an isolated point until all points are classified or marked.

5. The gear random fatigue load processing method based on the GCM counting method according to claim 4, wherein After performing clustering analysis on all the discrete small intervals according to the temperature and humidity data to determine several clustering sets, it further includes: If the number of instantaneous contact stresses in the clustering set is at least two, determining the representative instantaneous contact stress of the clustering set based on kernel density estimation; ； Among them, represents the instantaneous contact stress, represents the number of instantaneous contact stresses in the clustering set, represents the smoothing bandwidth, represents the value of a certain frequency point to be estimated, represents the i-th instantaneous contact stress in the clustering set.

6. The gear random fatigue load processing method based on the GCM counting method according to claim 5, wherein The smoothing bandwidth is calculated by the following formula: ； Among them, represents the smoothing bandwidth, represents the standard deviation of the data in the clustering set, represents the data skewness, represents the data kurtosis, represents the adjustment coefficient.

7. The method for processing random fatigue load of gears based on the GCM counting method according to claim 6, wherein, When converting the asymmetric cyclic stress into an equivalent symmetric cyclic stress by using the Goodman life curve, it includes: ； Among them, represents the instantaneous contact stress of the clustering set, represents the average stress of the clustering set, represents the ultimate tensile strength of the gear material, represents the corrected equivalent symmetric stress.

8. The method for processing random fatigue load of gears based on the GCM counting method according to claim 7, wherein, When comparing the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method and determining the adjustment coefficient according to the comparison result, and determining the warning threshold according to the total gear life and the adjustment coefficient, it includes: Obtaining a ratio coefficient according to the equivalent symmetric cyclic stress and the cyclic stress under the rainflow counting method, where the ratio coefficient is the ratio of the equivalent symmetric cyclic stress to the cyclic stress under the rainflow counting method; When the ratio coefficient is less than or equal to 1, determining the adjustment coefficient according to the ratio coefficient. The adjustment coefficient is in a direct proportional relationship with the ratio coefficient, and the value range of the adjustment coefficient is (0.7, 0.9); When the ratio coefficient is greater than 1, determining the adjustment coefficient according to the ratio coefficient. The adjustment coefficient is in a direct proportional relationship with the ratio coefficient, and the value range of the adjustment coefficient is (0, 0.7).

9. The method for processing random fatigue load of gears based on the GCM counting method according to claim 8, wherein When comparing the total fatigue damage value with the warning threshold and determining the warning level according to the comparison result, it includes: Obtaining the remaining life of the gear according to the total fatigue damage value and the total gear life, and obtaining the life difference according to the remaining life of the gear and the warning threshold. The life difference is the difference between the remaining life of the gear and the warning threshold; Determining the warning level according to the life difference. The warning level is in an inverse proportional relationship with the life difference.

10. A gear random fatigue load processing system based on the GCM counting method, which is used to apply the gear random fatigue load processing method based on the GCM counting method according to any one of claims 1-9, and is characterized in that, It includes: A collection unit configured to collect the actual operating load data and actual operating environment data of the gear transmission system. The actual operating load data is a torque-time change curve, and the actual operating environment data includes temperature and humidity-time change curves. Convert the torque in the torque-time change curve into the instantaneous contact stress of the gear tooth surface; A first processing unit configured to divide the continuous contact stress into a number of discrete small intervals based on discretization processing. Each discrete small interval corresponds to a time segment and corresponds to a temperature and humidity data; An analysis unit configured to perform cluster analysis on all the discrete small intervals according to the temperature and humidity data to determine a number of cluster sets; A second processing unit collects the gear speed and obtains the number of gear rotations in each cluster set according to the gear speed, records the occurrence frequency of the stress level of the cluster set, and constructs a two-dimensional gear counting matrix. The two-dimensional gear counting matrix includes stress amplitude, mean stress, and corresponding frequency; A third processing unit configured to convert the asymmetric cyclic stress into an equivalent symmetric cyclic stress by using the Goodman life curve, and determine the total gear life according to the S-N curve of the gear material and the equivalent symmetric cyclic stress; compare the equivalent symmetric cyclic stress with the cyclic stress under the rainflow counting method, determine the adjustment coefficient according to the comparison result, and determine the warning threshold according to the total gear life and the adjustment coefficient; A warning unit configured to obtain the total fatigue damage value under the action of all levels of equivalent stress according to the two-dimensional gear counting matrix, compare the total fatigue damage value with the warning threshold, and determine the warning level according to the comparison result.

Images