Gear fault diagnosis method and system

Abstract

The invention relates to the technical field of gear fault monitoring, in particular to a gear fault diagnosis method and system. Comprising the following steps: establishing a plurality of working condition categories and a plurality of wear nodes according to historical parameters, and generating a primary diagnosis model of each working condition category; acquiring a monitoring signal and a working condition parameter of the gear according to a preset monitoring time node, and setting a secondary diagnosis model according to the working condition parameter; generating a fault risk value of the gear according to the secondary diagnosis model and the monitoring signal, and judging whether a maintenance instruction is generated or not according to the fault risk value; based on historical parameters of the gear, a plurality of working condition categories are set, and corresponding primary diagnosis models are set according to different working condition categories, so that accurate diagnosis of gear faults is realized, and early warning is timely performed on potential fault risks in the gear operation process. Through a mode of combining a working condition category and a wear node, accurate early warning of potential faults of the gear is realized, and the fault probability of the gear is reduced. And stable operation of the system is ensured.

Description

Technical Field

[0001] The present application relates to the technical field of gear fault monitoring, and in particular to a gear fault diagnosis method and system. Background Art

[0002] Gears are one of the key transmission components of industrial equipment and are widely used in modern large-scale equipment systems such as power, metallurgy, and mining. Due to long-term operation in harsh environments, gears are inevitably prone to wear, fatigue, cracks, and other faults, which seriously affect the performance of the entire mechanical system.

[0003] The alternating stresses generated by the meshing of gear pairs during operation can easily cause fatigue damage to the gear surface, gradually evolving into faults such as pitting on the tooth surface and cracks at the tooth root. Without early warning, these problems can quickly progress to severe damage such as spalling and tooth breakage, easily leading to safety accidents. Summary of the Invention

[0004] The purpose of this application is: to solve the above technical problems, this application provides a gear fault diagnosis method and system, aiming to improve the early warning efficiency of potential gear failure risks and ensure the production efficiency of industrial systems.

[0005] In some embodiments of the present application, multiple operating condition categories are set based on the historical parameters of the gears, and corresponding first-level diagnostic models are set according to different operating condition categories to achieve accurate diagnosis of gear faults and timely early warning of potential failure risks during gear operation.

[0006] In some embodiments of the present application, multiple wear nodes are set according to the expected working life of the gear, and corresponding correction sub-modules are set according to the real-time status of each wear node to avoid fluctuations in the monitoring index signal due to natural wear of the gear, which may distort the diagnostic results. By combining the working condition category with the wear node, accurate early warning of potential gear failures can be achieved, thereby reducing the probability of gear failure.

[0007] In some embodiments of the present application, a gear fault diagnosis method is provided, comprising: Establish multiple operating condition categories and multiple wear nodes based on historical parameters, and generate a first-level diagnostic model for each operating condition category; Acquire the gear monitoring signal and operating parameters according to the preset monitoring time nodes, and set the secondary diagnosis model based on the operating parameters; Generate a gear fault risk value based on the secondary diagnostic model and monitoring signals, and determine whether to generate a maintenance instruction based on the fault risk value; Among them, when establishing multiple working condition categories, including: Establish a working condition category sequence A, A=(a1,a2…a i …an ), where a i is the i-th operating condition category; n is the number of operating condition categories.

[0008] In some embodiments of the present application, generating a primary diagnostic model for each operating condition category includes: Preset multiple monitoring indicators; According to the working condition category sequence A, the i-th working condition category is set as the target working condition category; Traverse historical data, extract characteristic signals of each monitoring indicator under the target working condition category, and generate training data packets; Generate a first-level diagnostic model for the target operating condition category based on the iterative results of the training data; Generate the first-level diagnostic model of each working condition category in turn, and establish the first-level diagnostic model sequence B, B=(b1, b2…b i …b n ), where b i is the first-level diagnostic model for the i-th operating condition category.

[0009] In some embodiments of the present application, generating multiple wear nodes includes: Establish multiple wear nodes based on the expected life of the gear; Establish the wear node sequence T, T=(t1,t2… t i …t m ), where t i is the i-th wear node; m is the number of wear nodes of the gear within its expected life; Select the target working condition category; According to the first-level diagnostic model sequence B, set b i It is a target-level diagnostic model; Generate a modified sub-model of the target first-level diagnostic model at each wear node; Establish the modified sub-model sequence C of the target first-level diagnostic sub-model, C=(c1,c2…c i …c m ), where c i is the modified sub-model of the target first-level diagnostic model at the i-th wear node; Generate the corrected sub-models of all the first-level diagnostic models at each wear node in sequence.

[0010] In some embodiments of the present application, when setting the secondary diagnostic model according to the operating condition parameters, it includes: Get the real-time working condition parameters of the current monitoring time node; Generate similarity evaluation values between real-time working condition parameters and each working condition category, and establish similarity evaluation value series K, K=(k1, k2…k i …k n), where k i is the similarity evaluation value between the real-time working condition parameters and the i-th working condition category; Set the maximum value k in the similarity evaluation value sequence K max The first-level diagnostic model of the corresponding working condition category is the second-level diagnostic model.

[0011] In some embodiments of the present application, when generating a gear failure risk value based on the secondary diagnostic model and the monitoring signal, the process includes: Generate the primary standard value of each monitoring indicator based on the secondary diagnostic model; Establish a first-level standard value sequence G1, G1=(g'1, g'2…g' i …g' r ), where g 1i is the first-level standard value of the i-th monitoring indicator at the current monitoring time node; r is the number of monitoring indicators; Generate the wear status of the gear at the current monitoring time node based on the preset simulation model; Select the target wear node according to the wear state, and set the correction sub-model corresponding to the target wear node in the secondary diagnosis model as the target correction sub-model; Generate compensation coefficients for various monitoring indicators based on the correction sub-model; Generate real-time reference values of various monitoring indicators at the current monitoring time node based on the monitoring signal; Generate the fault risk value f of the current monitoring time node.

[0012] In some embodiments of the present application, generating the fault risk value f of the current monitoring time node includes: f=e1*Q1*[ β i *(g i -g' i ) 2 ]+e2*Q2*[ β i *(g i -η i *g' i ) 2 ]; Wherein, e1 is the preset first weight coefficient; e2 is the preset second weight coefficient; Q1 is the preset first fixed coefficient; Q2 is the preset second fixed coefficient; β i is the impact factor of the i-th monitoring indicator; g i is the compensation coefficient of the i-th primary standard value; η i is the real-time reference value of the i-th monitoring indicator at the current monitoring time node.

[0013] In some embodiments of the present application, when determining whether to generate a maintenance instruction based on the fault risk value, it includes: Presetting a first fault risk value threshold F1 and a second fault risk value threshold F2; If f < F1, no maintenance instruction is generated at the current monitoring time node; If F1 < f < F2, a first-level maintenance instruction is generated at the current monitoring time node; If f > F2, a second-level monitoring instruction is generated at the current monitoring time node, and an expected fault category is generated based on the monitoring signal and the second-level diagnostic model at the current monitoring time node.

[0014] In some embodiments of the present application, a gear fault diagnosis system is provided, including: A central control unit for establishing multiple working condition categories and multiple wear nodes based on historical parameters, and generating a first-level diagnostic model for each working condition category; A monitoring unit for collecting monitoring signals and working condition parameters of the gear according to preset monitoring time nodes; The central control unit includes: A first processing module for establishing a working condition category sequence A, A = (a1, a2…a i …a n ), where a i is the i-th working condition category; n is the number of working condition categories; The first processing module is further configured to preset multiple monitoring indicators; Sequentially setting the i-th working condition category as the target working condition category according to the working condition category sequence A; Traversing historical data, extracting characteristic signals of each monitoring indicator under the target working condition category, and generating a training data packet; Generating a first-level diagnostic model for the target working condition category according to the iterative result of the training data; Sequentially generating first-level diagnostic models for each working condition category, and establishing a first-level diagnostic model sequence B, B = (b1, b2…b i …b n ), where bi is the first-level diagnostic model for the i-th working condition category; A second processing module for establishing a wear node sequence T, T = (t1, t2…t i …t m ), where t i is the i-th wear node; m is the number of wear nodes of the gear within the expected service life; Selecting the target working condition category; Sequentially setting bi as the target first-level diagnostic model according to the first-level diagnostic model sequence B; Generating a correction sub-model for each wear node of the target first-level diagnostic model; Establish the modified sub-model sequence C of the target first-level diagnostic sub-model, C=(c1,c2…c i …c m ), where c i is the modified sub-model of the target first-level diagnostic model at the i-th wear node; Generate the modified sub-models of all the first-level diagnostic models at each wear node in sequence; The third processing module is used to set a secondary diagnosis model according to the working condition parameters; The fourth processing module is used to generate a fault risk value of the gear according to the secondary diagnostic model and the monitoring signal, and determine whether to generate a maintenance instruction according to the fault risk value.

[0015] In some embodiments of the present application, the third processing module is further configured to: Get the real-time working condition parameters of the current monitoring time node; Generate similarity evaluation values between real-time working condition parameters and each working condition category, and establish similarity evaluation value series K, K=(k1, k2…k i …k n ), where k i is the similarity evaluation value between the real-time working condition parameters and the i-th working condition category; Set the maximum value k in the similarity evaluation value sequence K max The first-level diagnostic model of the corresponding working condition category is the second-level diagnostic model.

[0016] In some embodiments of the present application, the fourth processing module is further configured to: Generate the primary standard value of each monitoring indicator based on the secondary diagnostic model; Establish a first-level standard value sequence G1, G1=(g'1, g'2…g' i …g' r ), where g 1i is the first-level standard value of the i-th monitoring indicator at the current monitoring time node; r is the number of monitoring indicators; Generate the wear status of the gear at the current monitoring time node based on the preset simulation model; Select the target wear node according to the wear state, and set the correction sub-model corresponding to the target wear node in the secondary diagnosis model as the target correction sub-model; Generate compensation coefficients for various monitoring indicators based on the correction sub-model; Generate real-time reference values of various monitoring indicators at the current monitoring time node based on the monitoring signal; Generate the fault risk value f of the current monitoring time node; f=e1*Q1*[ β i *(g i -g'i ) 2 +e2*Q2* β i *(g i -η i *g' i ) 2 ; Where, e1 is a preset first weight coefficient; e2 is a preset second weight coefficient; Q1 is a preset first fixed coefficient; Q2 is a preset second fixed coefficient; β i is the influence factor of the i-th monitoring index; g i is the compensation coefficient of the i-th first-level standard value; η i is the real-time reference value of the i-th monitoring index at the current monitoring time node; Preset the first fault risk value threshold F1 and the second fault risk value threshold F2; If f < F1, no maintenance instruction is generated at the current monitoring time node; If F1 < f < F2, a first-level maintenance instruction is generated at the current monitoring time node; If f > F2, a second-level monitoring instruction is generated at the current monitoring time node, and the expected fault category is generated according to the monitoring signal at the current monitoring time node and the second-level diagnosis model.

[0017] Compared with the prior art, the gear fault diagnosis method and system in the embodiments of the present application have the following beneficial effects: Based on the historical parameters of the gear, multiple working condition categories are set, and corresponding first-level diagnosis models are set according to different working condition categories, so as to achieve accurate diagnosis of gear faults and timely warn of potential fault risks during the operation of the gear.

[0018] According to the expected working life of the gear, multiple wear nodes are set, and corresponding correction sub-modules are set according to the real-time states of each wear node, so as to avoid the monitoring index signal fluctuation caused by the natural wear of the gear and make the diagnosis result distorted. By combining the working condition category and the wear node, accurate early warning of potential gear faults is achieved, and the fault probability of the gear is reduced. BRIEF DESCRIPTION OF THE DRAWINGS<00002​​​​​​​​​​In the description of this application, it should be understood that the terms "center", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application.

[0022] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature specified as "first" or "second" may explicitly or implicitly include one or more of such features. Throughout this application, unless otherwise specified, "plurality" means two or more.

[0023] In the description of this application, it should be noted that, unless otherwise expressly specified or limited, the terms "mounted," "connected," and "connected" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integral connections; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; and internal connections between two components. Those skilled in the art will understand the specific meanings of the above terms in this application based on the specific circumstances.

[0024] like Figure 1 As shown, a gear fault diagnosis method according to a preferred embodiment of the present application includes: S101: establishing multiple operating condition categories and multiple wear nodes based on historical parameters, and generating a first-level diagnostic model for each operating condition category; S102: Acquire a monitoring signal and operating parameters of the gear according to a preset monitoring time node, and set a secondary diagnosis model according to the operating parameters; S103: generating a fault risk value of the gear according to the secondary diagnosis model and the monitoring signal, and determining whether to generate a maintenance instruction according to the fault risk value; Among them, when establishing multiple working condition categories, including: Establish a working condition category sequence A, A=(a1,a2…a i …a n ), where a i is the i-th operating condition category; n is the number of operating condition categories.

[0025] Specifically, multiple operating condition indicators are generated based on historical data, including but not limited to operating speed, operating load, gear surface pressure and other parameters. Multiple value intervals are established for each operating condition indicator in turn, and multiple operating condition categories are generated based on random combinations of each value interval. The values of each operating condition indicator within any two operating condition categories are different.

[0026] Specifically, when generating the first-level diagnostic model for each operating condition category, it includes: Preset multiple monitoring indicators; According to the working condition category sequence A, the i-th working condition category is set as the target working condition category; Traverse historical data, extract characteristic signals of each monitoring indicator under the target working condition category, and generate training data packets; Generate a first-level diagnostic model for the target operating condition category based on the iterative results of the training data; Generate the first-level diagnostic model of each working condition category in turn, and establish the first-level diagnostic model sequence B, B=(b1, b2…b i …b n ), where b i is the first-level diagnostic model for the i-th operating condition category.

[0027] Specifically, the monitoring indicators include but are limited to parameters such as gear temperature, gear vibration signal, shock pulse intensity and frequency generated by the gear, and peak value changes at specific frequencies.

[0028] Specifically, the historical values of each monitoring indicator within a single working condition category are used to construct a training data package, and the characteristic parameters therein are extracted to generate the standard operating values of each monitoring indicator under the current working condition, thereby constructing the corresponding first-level diagnostic model.

[0029] It can be understood that in the above embodiment, multiple operating condition categories are set based on the historical parameters of the gear, and corresponding first-level diagnostic models are set according to different operating condition categories to achieve accurate diagnosis of gear faults and timely warning of potential failure risks during gear operation.

[0030] In a preferred embodiment of the present application, when generating multiple wear nodes, the method includes: Establish multiple wear nodes based on the expected life of the gear; Establish the wear node sequence T, T=(t1,t2…t i …t m ), where t i is the i-th wear node; m is the number of wear nodes of the gear within its expected life; Select the target working condition category; According to the first-level diagnostic model sequence B, set bi as the target first-level diagnostic model in sequence; Generate a modified sub-model of the target first-level diagnostic model at each wear node; Establish the modified sub-model sequence C of the target first-level diagnostic sub-model, C=(c1,c2…c i …c m ), where c i is the modified sub-model of the target first-level diagnostic model at the i-th wear node; Generate the corrected sub-models of all the first-level diagnostic models at each wear node in sequence.

[0031] Specifically, the real-time values of various monitoring indicators are collected when the gear is at different wear nodes under a single working condition category, so as to analyze the fluctuation of each monitoring indicator at different wear nodes and construct the corresponding correction sub-model.

[0032] In the preferred embodiment of the application of our company, when setting the secondary diagnosis model according to the working condition parameters, it includes: Get the real-time working condition parameters of the current monitoring time node; Generate similarity evaluation values between real-time working condition parameters and each working condition category, and establish similarity evaluation value series K, K=(k1, k2…k i …k n ), where k i is the similarity evaluation value between the real-time working condition parameters and the i-th working condition category; Set the maximum value k in the similarity evaluation value sequence K max The first-level diagnostic model of the corresponding working condition category is the second-level diagnostic model.

[0033] Specifically, the real-time values of various working condition indicators are generated according to the real-time working condition parameters, thereby generating similarity evaluation values with each working condition category. The larger the similarity evaluation value, the closer the working condition of the current gear is to the corresponding working condition category.

[0034] Specifically, when generating the gear failure risk value based on the secondary diagnostic model and monitoring signals, it includes: Generate the primary standard value of each monitoring indicator based on the secondary diagnostic model; Establish a first-level standard value sequence G1, G1=(g'1, g'2…g' i …g' r ), where g 1i is the first-level standard value of the i-th monitoring indicator at the current monitoring time node; r is the number of monitoring indicators; Generate the wear status of the gear at the current monitoring time node based on the preset simulation model; Select the target wear node according to the wear state, and set the correction sub-model corresponding to the target wear node in the secondary diagnosis model as the target correction sub-model; Generate the compensation coefficients for each monitoring index according to the correction sub-model; Generate the real-time reference values of each monitoring index at the current monitoring time node according to the monitoring signals; Generate the fault risk value f at the current monitoring time node.

[0035] Specifically, by establishing a simulation model, the operating state of the gear is simulated and the wear degree of the gear is generated according to the historical monitoring parameters of the gear, so as to select the corresponding wear nodes.

[0036] Specifically, when generating the fault risk value f at the current monitoring time node, it includes: f = e1 * Q1 * β i *(g i - g' i ) 2 + e2 * Q2 * β i *(g i - η i * g' i ) 2 ; Where, e1 is the preset first weight coefficient; e2 is the preset second weight coefficient; Q1 is the preset first fixed coefficient; Q2 is the preset second fixed coefficient; β i is the influence factor of the i-th monitoring index; g i is the compensation coefficient of the i-th first-level standard value; η i is the real-time reference value of the i-th monitoring index at the current monitoring time node.

[0037] Specifically, normalize all the data in the model by the preset first fixed coefficient and second fixed coefficient, so that each parameter is within the same value range.

[0038] It can be understood that in the above embodiment, according to the expected working life of the gear, multiple wear nodes are set, and the corresponding correction sub-modules are set according to the real-time states of each wear node, avoiding the signal fluctuation of the monitoring index caused by the natural wear of the gear and making the diagnosis result distorted. By combining the working condition category with the wear node, the accurate early warning of the potential fault of the gear is realized, and the fault probability of the gear is reduced.

[0039] In the preferred embodiment of the embodiment of the present application, when judging whether to generate a maintenance instruction according to the fault risk value, it includes: Preset the first fault risk value threshold F1 and the second fault risk value threshold F2; If f < F1, no maintenance instruction is generated at the current monitoring time node; If F1 < f < F2, a first-level maintenance instruction is generated at the current monitoring time node; If f>F2, the current monitoring time node generates a secondary monitoring instruction, and generates an expected fault category based on the monitoring signal of the current monitoring time node and the secondary diagnosis model.

[0040] Specifically, a Level 1 maintenance instruction indicates that the current gear may be operating abnormally, and a corresponding maintenance plan can be generated based on the diagnostic results of the remaining gears. A Level 2 maintenance instruction indicates that the current gear is at risk of failure and requires immediate maintenance to eliminate the fault and ensure safe operation of the system.

[0041] Based on another preferred embodiment of a gear fault diagnosis method in any of the above preferred embodiments, this preferred embodiment provides a gear fault diagnosis system, including: The central control unit is used to establish multiple operating condition categories and multiple wear nodes based on historical parameters, and generate a first-level diagnostic model for each operating condition category; A monitoring unit, used to collect monitoring signals and working condition parameters of the gear according to preset monitoring time nodes; Specifically, the monitoring unit is preferably various sensors, including but not limited to vibration sensors for collecting vibration signals of gears, and temperature sensors for collecting operating temperatures of gears.

[0042] The central control unit includes: The first processing module is used to establish the working condition category sequence A, A=(a1, a2…a i …a n ), where a i is the i-th operating condition category; n is the number of operating condition categories; The first processing module is also used to preset multiple monitoring indicators; According to the working condition category sequence A, the i-th working condition category is set as the target working condition category; Traverse historical data, extract characteristic signals of each monitoring indicator under the target working condition category, and generate training data packets; Generate a first-level diagnostic model for the target operating condition category based on the iterative results of the training data; Generate the first-level diagnostic model of each working condition category in turn, and establish the first-level diagnostic model sequence B, B=(b1, b2…b i …b n ), where b i is the first-level diagnostic model of the i-th operating condition category; The second processing module establishes the wear node sequence T, T=(t1, t2…t i …t m ), where t i is the i-th wear node; m is the number of wear nodes of the gear within its expected life; Select the target working condition category; According to the first-level diagnostic model sequence B, set bi as the target first-level diagnostic model in sequence; Generate a modified sub-model of the target first-level diagnostic model at each wear node; Establish the modified sub-model sequence C of the target first-level diagnostic sub-model, C=(c1,c2…c i …c m ), where c i is the modified sub-model of the target first-level diagnostic model at the i-th wear node; Generate the modified sub-models of all the first-level diagnostic models at each wear node in sequence; The third processing module is used to set a secondary diagnosis model according to the working condition parameters; The fourth processing module is used to generate a fault risk value of the gear according to the secondary diagnostic model and the monitoring signal, and determine whether to generate a maintenance instruction according to the fault risk value.

[0043] In a preferred embodiment of the present application, the third processing module is further configured to: Get the real-time working condition parameters of the current monitoring time node; Generate similarity evaluation values between real-time working condition parameters and each working condition category, and establish similarity evaluation value series K, K=(k1, k2…k i …k n ), where k i is the similarity evaluation value between the real-time working condition parameters and the i-th working condition category; Set the maximum value k in the similarity evaluation value sequence K max The first-level diagnostic model of the corresponding working condition category is the second-level diagnostic model.

[0044] In a preferred embodiment of the present application, the fourth processing module is further configured to: Generate the primary standard value of each monitoring indicator based on the secondary diagnostic model; Establish a first-level standard value sequence G1, G1=(g'1, g'2…g' i …g' r ), where g 1i is the first-level standard value of the i-th monitoring indicator at the current monitoring time node; r is the number of monitoring indicators; Generate the wear status of the gear at the current monitoring time node based on the preset simulation model; Select the target wear node according to the wear state, and set the correction sub-model corresponding to the target wear node in the secondary diagnosis model as the target correction sub-model; Generate compensation coefficients for various monitoring indicators based on the correction sub-model; Generate real-time reference values of various monitoring indicators at the current monitoring time node based on the monitoring signal; Generate the fault risk value f at the current monitoring time node; f = e1 * Q1 * β i *(g i - g' i ) 2 + e2 * Q2 * β i *(g i - η i * g' i ) 2 ; Where, e1 is the preset first weight coefficient; e2 is the preset second weight coefficient; Q1 is the preset first fixed coefficient; Q2 is the preset second fixed coefficient; β i is the influence factor of the i-th monitoring index; g i is the compensation coefficient of the i-th first-level standard value; η i is the real-time reference value of the i-th monitoring index at the current monitoring time node; Preset the first fault risk value threshold F1 and the second fault risk value threshold F2; If f < F1, no maintenance instruction is generated at the current monitoring time node; If F1 < f < F2, a first-level maintenance instruction is generated at the current monitoring time node; If f > F2, a second-level monitoring instruction is generated at the current monitoring time node, and the expected fault category is generated according to the monitoring signal and the second-level diagnosis model at the current monitoring time node.

[0045] According to the first concept of the present application, based on the historical parameters of the gear, multiple working condition categories are set, and the corresponding first-level diagnosis models are set according to different working condition categories, so as to achieve accurate diagnosis of gear faults and timely warn of potential fault risks during the operation of the gear.

[0046] According to the second concept of the present application, multiple wear nodes are set according to the expected working life of the gear, and the corresponding correction sub-modules are set according to the real-time states of each wear node, so as to avoid the distortion of the monitoring index signal caused by the natural wear of the gear, and the accurate warning of potential gear faults is realized by combining the working condition category and the wear node, and the fault probability of the gear is reduced.

[0047] The above are only the preferred embodiments of the present application. It should be noted that for those of ordinary skill in the art, without departing from the technical principle of the present application, several improvements and replacements can be made, and these improvements and replacements should also be regarded as the protection scope of the present application.

Claims

1. A gear fault diagnosis method, characterized in that: Including: Establish multiple working condition categories and multiple wear nodes based on historical parameters, and generate a first-level diagnostic model for each working condition category; Obtain the monitoring signals and working condition parameters of the gear according to the preset monitoring time nodes, and set a second-level diagnostic model according to the working condition parameters; Generate a fault risk value for the gear according to the second-level diagnostic model and the monitoring signals, and determine whether to generate a maintenance instruction according to the fault risk value; Among them, when establishing multiple working condition categories, it includes: Establish a working condition category sequence A, A=(a1,a2…a i …a n ), where a i is the i-th operating condition category; n is the number of operating condition categories.

2. The gear fault diagnosis method according to claim 1, wherein: When generating a first-level diagnostic model for each working condition category, it includes: Preset multiple monitoring indicators; Set the i-th working condition category as the target working condition category in sequence according to the working condition category sequence A; Traverse the historical data, extract the characteristic signals of each monitoring indicator under the target working condition category, and generate a training data packet; Generate a first-level diagnostic model for the target working condition category according to the iterative results of the training data; Generate the first-level diagnostic model of each working condition category in turn, and establish the first-level diagnostic model sequence B, B=(b1, b2…b i …b n ), where b i is the first-level diagnostic model for the i-th operating condition category.

3. The gear fault diagnosis method according to claim 2, wherein: When generating multiple wear nodes, it includes: Establish multiple wear nodes according to the expected life of the gear; Establish the wear node sequence T, T=(t1,t2… t i …t m ), where t i is the i-th wear node; m is the number of wear nodes of the gear within its expected life; Select the target working condition category; According to the first-level diagnostic model sequence B, set b i It is a target-level diagnostic model; Generate a correction sub-model for each wear node of the target first-level diagnostic model; Establish the modified sub-model sequence C of the target first-level diagnostic sub-model, C=(c1,c2…c i …c m ), where c i is the modified sub-model of the target first-level diagnostic model at the i-th wear node; Generate correction sub-models for each wear node of all first-level diagnostic models in sequence.

4. The gear fault diagnosis method according to claim 3, wherein: When setting the second-level diagnostic model according to the working condition parameters, it includes: Obtain the real-time working condition parameters of the current monitoring time node; Generate similarity evaluation values between real-time working condition parameters and each working condition category, and establish similarity evaluation value series K, K=(k1, k2…k i …k n ), where k i is the similarity evaluation value between the real-time working condition parameters and the i-th working condition category; Set the maximum value k in the similarity evaluation value sequence K max The first-level diagnostic model of the corresponding working condition category is the second-level diagnostic model.

5. The gear fault diagnosis method according to claim 4, characterized in that: When generating a fault risk value for the gear according to the second-level diagnostic model and the monitoring signals, it includes: Generate a first-level standard value for each monitoring indicator according to the second-level diagnostic model; Establish a first-level standard value sequence G1, G1=(g'1, g'2…g' i …g' r ), where g 1i is the first-level standard value of the i-th monitoring indicator at the current monitoring time node; r is the number of monitoring indicators; Generate the wear state of the gear at the current monitoring time node according to the preset simulation model; Select the target wear node according to the wear state, and set the correction sub-model corresponding to the target wear node of the second-level diagnostic model as the target correction sub-model; Generate a compensation coefficient for each monitoring indicator according to the correction sub-model; Generate a real-time reference value for each monitoring indicator at the current monitoring time node according to the monitoring signals; Generate a fault risk value f for the current monitoring time node.

6. The gear fault diagnosis method according to claim 5, characterized in that: When generating a fault risk value f for the current monitoring time node, it includes: f=e1*Q1*[ b i *(g i -g' i ) 2 ]+e2*Q2*[ b i *(g i -or i *g' i ) 2 ]; Wherein, e1 is the preset first weight coefficient; e2 is the preset second weight coefficient; Q1 is the preset first fixed coefficient; Q2 is the preset second fixed coefficient; β i is the impact factor of the i-th monitoring indicator; g i is the compensation coefficient of the i-th primary standard value; η i is the real-time reference value of the i-th monitoring indicator at the current monitoring time node.

7. The gear fault diagnosis method according to claim 5, characterized in that: When determining whether to generate a maintenance instruction according to the fault risk value, it includes: Preset a first fault risk value threshold F1 and a second fault risk value threshold F2; If f < F1, no maintenance instruction is generated at the current monitoring time node; If F1 < f < F2, a first-level maintenance instruction is generated at the current monitoring time node; If f > F2, a second-level monitoring instruction is generated at the current monitoring time node, and an expected fault category is generated according to the monitoring signals and the second-level diagnostic model at the current monitoring time node.

8. A gear fault diagnosis system, using the gear fault diagnosis method according to any one of claims 1 to 7, characterized in that: Including: A central control unit, used to establish multiple working condition categories and multiple wear nodes based on historical parameters, and generate a first-level diagnostic model for each working condition category; A monitoring unit, used to collect the monitoring signals and working condition parameters of the gear according to the preset monitoring time nodes; The central control unit includes: The first processing module is used to establish the working condition category sequence A, A=(a1, a2…a i …a n ), where a i is the i-th operating condition category; n is the number of operating condition categories; The first processing module is also used to preset multiple monitoring indicators; Set the i-th working condition category as the target working condition category in sequence according to the working condition category sequence A; Traverse the historical data, extract the characteristic signals of each monitoring indicator under the target working condition category, and generate a training data packet; Generate a first-level diagnostic model for the target working condition category according to the iterative results of the training data; Generate the first-level diagnostic model of each working condition category in turn, and establish the first-level diagnostic model sequence B, B=(b1, b2…b i …b n ), where bi is the first-level diagnostic model of the i-th operating condition category; The second processing module establishes the wear node sequence T, T=(t1, t2…t i …t m ), where t i is the i-th wear node; m is the number of wear nodes of the gear within its expected life; Select the target working condition category; Set bi as the target first-level diagnostic model in sequence according to the first-level diagnostic model sequence B; Generate the correction sub - models of the target first - level diagnosis model at each wear node; Establish the modified sub-model sequence C of the target first-level diagnostic sub-model, C=(c1,c2…c i …c m ), where c i is the modified sub-model of the target first-level diagnostic model at the i-th wear node; Generate the correction sub - models of all first - level diagnosis models at each wear node in sequence; The third processing module is used to set the second - level diagnosis model according to the working condition parameters; The fourth processing module is used to generate the fault risk value of the gear according to the second - level diagnosis model and the monitoring signal, and judge whether to generate a maintenance instruction according to the fault risk value.

9. The gear fault diagnosis system according to claim 8, characterized in that: The third processing module is further used for: Obtain the real - time working condition parameters of the current monitoring time node; Generate similarity evaluation values between real-time working condition parameters and each working condition category, and establish similarity evaluation value series K, K=(k1, k2…k i …k n ), where k i is the similarity evaluation value between the real-time working condition parameters and the i-th working condition category; Set the maximum value k in the similarity evaluation value sequence K max The first-level diagnostic model of the corresponding working condition category is the second-level diagnostic model.

10. The gear fault diagnosis system according to claim 9, wherein: The fourth processing module is further used for: Generate the first - level standard values of each monitoring index according to the second - level diagnosis model; Establish a first-level standard value sequence G1, G1=(g'1, g'2…g' i …g' r ), where g 1i is the first-level standard value of the i-th monitoring indicator at the current monitoring time node; r is the number of monitoring indicators; Generate the wear state of the gear at the current monitoring time node according to the preset simulation model; Select the target wear node according to the wear state, and set the correction sub - model corresponding to the target wear node of the second - level diagnosis model as the target correction sub - model; Generate the compensation coefficients of each monitoring index according to the correction sub - model; Generate the real - time reference values of each monitoring index at the current monitoring time node according to the monitoring signal; Generate the fault risk value f at the current monitoring time node; f=e1*Q1*[ b i *(g i -g' i ) 2 ]+e2*Q2*[ b i *(g i -or i *g' i ) 2 ]; Wherein, e1 is the preset first weight coefficient; e2 is the preset second weight coefficient; Q1 is the preset first fixed coefficient; Q2 is the preset second fixed coefficient; β i is the impact factor of the i-th monitoring indicator; g i is the compensation coefficient of the i-th primary standard value; η i is the real-time reference value of the i-th monitoring indicator at the current monitoring time node; Preset the first fault risk value threshold F1 and the second fault risk value threshold F2; If f < F1, no maintenance instruction is generated at the current monitoring time node; If F1 < f < F2, a first - level maintenance instruction is generated at the current monitoring time node; If f > F2, a second - level monitoring instruction is generated at the current monitoring time node, and the expected fault category is generated according to the monitoring signal and the second - level diagnosis model at the current monitoring time node.

Images