Real-time gearbox gear wear prediction method and system based on motor current signal

Abstract

The application discloses a gearbox gear wear real-time prediction method based on motor current signals, single-phase stator current signals are collected and time-domain statistical feature extraction is carried out; a health score is obtained according to a health baseline model and weighted fusion; trend tracking and threshold judgment are carried out to realize gear wear early warning and residual life prediction. The method solves the problems of traditional fault diagnosis, such as invasive installation, high cost and difficulty in embedded real-time operation, thereby realizing non-invasive, low-cost and online monitoring without disassembling the gearbox. The method can adapt to new energy automobile high-speed, electric engineering machinery heavy load impact and other variable working condition environments, has strong robustness, wide application range, can be directly integrated into the existing ECU and CAN bus system, and supports remote monitoring and predictive maintenance.

Description

Technical Field

[0001] This invention relates to the field of equipment fault diagnosis technology, and in particular to a method and system for real-time prediction of gear wear in a gearbox based on motor current signals. Background Technology

[0002] With the rapid development of new energy vehicles and electric construction machinery, gearboxes, as core transmission components, are increasingly facing major risks affecting system efficiency, durability, and safety due to gear wear, fatigue cracks, and other defects. Specifically, new energy vehicle gearboxes need to adapt to the high-speed motor drive characteristics, while electric construction machinery gearboxes are subjected to high torque of hundreds of Newton-meters and complex alternating loads for extended periods. These operating conditions easily lead to premature gear wear, resulting in decreased transmission efficiency, increased noise and vibration, and even sudden failures, significantly increasing maintenance costs.

[0003] Currently, traditional gearbox fault diagnosis methods mainly rely on vibration sensors, torque sensors, or acoustic emission signals for detection. While these methods are effective to some extent, they suffer from the following significant drawbacks: sensor installation typically requires invasive modifications, resulting in high system costs, and they are sensitive to ambient noise and electromagnetic interference. At the signal processing level, frequency domain analysis-based methods, such as the Fast Fourier Transform, have high computational complexity, making real-time implementation difficult in vehicle controllers or embedded systems in engineering machinery. Furthermore, most existing technologies can only achieve post-detection or offline analysis, making it difficult to predict early trends in gear wear and estimate remaining service life online. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention provides a method and system for real-time prediction of gearbox gear wear based on motor current signals. By collecting single-phase stator current signals of the motor and extracting lightweight time-domain statistical features to obtain a health score, and combining it with exponentially weighted moving average trend tracking to obtain the remaining life, it achieves early wear warning and remaining life estimation with millisecond-level response and non-intrusive embedded deployment.

[0005] On the one hand, a method for real-time prediction of gearbox gear wear based on motor current signals is provided, including: Obtain the single-phase stator current signal of the motor running the gearbox; The time-domain statistical characteristics of the single-phase stator current signal of the motor are obtained by preprocessing. The time-domain statistical features are compared with the baseline model, the deviations of each feature are calculated, and the weighted fusion is used to obtain the health score. The health score sequence is smoothly tracked. When the health score shows an upward trend for N consecutive windows, an early warning is triggered, and the remaining lifespan is estimated by linear extrapolation. It outputs early warning information and prediction results, including health scores and estimated remaining life expectancy.

[0006] Furthermore, the single-phase stator current signal of the motor running in the gearbox is obtained, including: obtaining the single-phase stator current signal of the motor through a current sensor, wherein the current sensor is a non-contact current sensor, and a Hall effect clamp or shunt sensor is used.

[0007] Furthermore, the single-phase stator current signal of the motor is preprocessed to obtain time-domain statistical characteristics, including: The acquired current signal is segmented into sliding windows, and time-domain statistical characteristics are calculated within each window.

[0008] Furthermore, the time-domain statistical features include four aspects: amplitude change rate, variance, kurtosis, and mean value at zero crossing time.

[0009] Furthermore, the time-domain statistical features are compared with the baseline model, the biases of each feature are calculated, and weighted fusion is performed to obtain a health score, including: A baseline model is pre-established using multiple sets of data under healthy conditions. The baseline model includes the mean and standard deviation of the features. The extracted time-domain statistical features are compared with the baseline model in real time, the deviation of each feature is calculated and weighted and fused to obtain a health score.

[0010] Furthermore, the health score sequence is smoothly tracked. When the health score shows an upward trend for N consecutive windows, an early warning is triggered, and the remaining life expectancy is estimated through linear extrapolation, including: An exponentially weighted moving average algorithm is used to smoothly track the health score sequence. When the health score shows an upward trend for multiple consecutive windows, it is determined to be in the wear development stage, and the remaining useful life is estimated by linear extrapolation.

[0011] On the other hand, a real-time gear wear prediction system based on motor current signals is provided, including: Acquisition module: Acquires the single-phase stator current signal of the motor running in the gearbox; Feature extraction module: preprocesses the single-phase stator current signal of the motor to obtain time-domain statistical features; Health score calculation module: compares time-domain statistical features with the baseline model, calculates the deviation of each feature, and weights and fuses them to obtain the health score; Trend prediction module: smoothly tracks the health score sequence, triggers an alert when the health score shows an upward trend for N consecutive windows, and estimates the remaining life expectancy through linear extrapolation; Early warning output module: Outputs early warning information and prediction results, including health score and estimated remaining lifespan.

[0012] Furthermore, an electronic device is also provided, including: Memory, used for non-transitory storage of computer-readable instructions; and Processor, for executing the computer-readable instructions, When the computer-readable instructions are executed by the processor, they perform the method described in the first aspect above.

[0013] In another aspect, a storage medium is also provided for non-transitory storage of computer-readable instructions, wherein when the non-transitory computer-readable instructions are executed by a computer, the method described in the first aspect is performed.

[0014] In another aspect, a computer program product is also provided, including a computer program that, when run on one or more processors, is used to implement the method described in the first aspect above.

[0015] The above technical solution has the following advantages or beneficial effects: Obtain the single-phase stator current signal of the motor running the gearbox; The time-domain statistical characteristics of the single-phase stator current signal of the motor are obtained by preprocessing. The time-domain statistical features are compared with the baseline model, the deviations of each feature are calculated, and the weighted fusion is used to obtain the health score. The health score sequence is smoothly tracked. When the health score shows an upward trend for N consecutive windows, an early warning is triggered, and the remaining lifespan is estimated by linear extrapolation. It outputs early warning information and prediction results, including health scores and estimated remaining life expectancy.

[0016] This invention provides a real-time prediction method for gearbox gear wear based on motor current signals, comprising the following steps: acquiring single-phase stator current signals and extracting time-domain statistical features; obtaining a health score based on a health baseline model and weighted fusion; and performing trend tracking and threshold judgment to achieve gear wear early warning and remaining life prediction. This method solves the problems of traditional fault diagnosis methods, such as invasive installation, high cost, and difficulty in embedded real-time operation, thus achieving non-invasive, low-cost, and gearbox-free online monitoring. This method is adaptable to varying operating conditions such as high-speed operation of new energy vehicles and heavy-load impact of electric construction machinery, exhibiting strong robustness and wide applicability. Furthermore, it can be directly integrated into existing ECU and CAN bus systems, supporting remote monitoring and predictive maintenance, demonstrating strong engineering feasibility and significant economic and social benefits. Attached Figure Description

[0017] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0018] Figure 1 This is a schematic diagram of the overall power connection and current signal acquisition of the electric motor and gearbox as described in Example 1; Figure 2 The flowchart is shown in Embodiment 1, which describes a method for real-time prediction of gear wear in a gearbox based on motor current signals. Figure 3 This is a health rating chart of the electric logistics vehicle transmission described in Example 1; Figure 4 This is a comparison diagram between the real-time gear wear prediction method based on motor current signal of the present invention described in Example 1 and the traditional gear wear prediction method. Detailed Implementation

[0019] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0020] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the exemplary embodiments of the invention. The terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion, for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0021] In this embodiment of the invention, "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Furthermore, in the description of this invention, "multiple" refers to two or more.

[0022] Furthermore, to facilitate a clear description of the technical solutions of the embodiments of the present invention, the terms "first" and "second" are used in the embodiments of the present invention to distinguish identical or similar items with essentially the same function and effect. Those skilled in the art will understand that the terms "first" and "second" do not limit the quantity or execution order, and the terms "first" and "second" are not necessarily different.

[0023] Where there is no conflict, the embodiments and features in the embodiments of the present invention can be combined with each other.

[0024] All data acquisition in this embodiment is carried out in accordance with laws and regulations and with user consent, and the data is used legally.

[0025] Example 1 A schematic diagram of the overall power connection and current signal acquisition between the electric motor and the gearbox, as shown below. Figure 1 As shown, the vehicle controller is connected to the energy source via an energy management unit. The energy source is the system's power source, typically a high-voltage lithium battery pack providing direct current (DC). The energy management unit is primarily responsible for battery charging and discharging management, energy distribution, and safety monitoring. The vehicle controller, acting as the system's brain, receives sensor signals, issues control commands, and integrates the fault prediction algorithm module of this invention. A DC-AC inverter converts the battery's DC power into three-phase AC power to drive the motor, while simultaneously acquiring the motor's current signal. The DC-AC converter is the core acquisition component of this invention, installed on the U-phase (or any one phase) power line from the inverter output to the motor, used for non-intrusive acquisition of single-phase stator current signals. The motor is a permanent magnet synchronous motor or an asynchronous motor, converting electrical energy into mechanical energy, which drives the gearbox via the output shaft. The gearbox is the monitoring object of this invention, comprising gears, bearings, shafts, etc. Torque pulsations caused by gear wear are reflected back to the motor through mechanical connections, thus appearing in the current signal.

[0026] The gearbox is rigidly connected to the motor's output shaft via mechanical connections such as splines, couplings, or flanges. When the gearbox gears wear, minute torque pulsations are generated. These pulsations are transmitted to the motor rotor through the mechanical shaft, causing subtle changes in the motor stator current, such as amplitude, variance, and kurtosis. In this embodiment, a non-contact current sensor is installed on the motor power line. The sensor can be directly clipped to the existing cable, utilizing the existing wiring of the motor driver, allowing for real-time prediction of gearbox gear wear without ever opening the gearbox.

[0027] This embodiment provides a method for real-time prediction of gearbox gear wear based on motor current signals, such as... Figure 2 As shown, it includes: Obtain the single-phase stator current signal of the motor running the gearbox; The time-domain statistical characteristics of the single-phase stator current signal of the motor are obtained by preprocessing. The time-domain statistical features are compared with the baseline model, the deviations of each feature are calculated, and the weighted fusion is used to obtain the health score. The health score sequence is smoothly tracked. When the health score shows an upward trend for N consecutive windows, an early warning is triggered, and the remaining lifespan is estimated by linear extrapolation. It outputs early warning information and prediction results, including health scores and estimated remaining life expectancy.

[0028] Therefore, the real-time prediction method for gearbox gear wear based on motor current signal provided in this embodiment includes the following steps: S1: Obtain the single-phase stator current signal of the motor running in the gearbox.

[0029] This embodiment can acquire the single-phase stator current signal of the motor driving the gearbox by using a current sensor installed at the output terminal of the motor driver, preferably the U-phase current. .

[0030] The current sensor is a non-contact current sensor, which can use Hall effect clamps or shunt sensors, etc., without the need for additional vibration, torque or multiphase fusion sensors, reducing installation and modification costs by more than 90%.

[0031] S2: Preprocess the single-phase stator current signal of the motor to obtain time-domain statistical characteristics.

[0032] The preprocessing described in this embodiment includes: performing sliding window segmentation on the acquired current signal, wherein the window length is 1 second and the overlap rate is 50%, and calculating the time-domain statistical features within each window.

[0033] Preferably, the time-domain statistical features include four: amplitude change rate, variance, kurtosis, and mean value at zero crossing time.

[0034] The formula for calculating the rate of change of amplitude is: (1) in, The average peak value pre-statistically calculated under healthy conditions. This indicates the maximum value of the acquired current signal; This represents the minimum value of the acquired current signal; the amplitude change rate is used to reflect the degree of change in the peak-to-peak value of the current signal.

[0035] The formula for calculating sample variance is: (2) in, is the number of sampling points within the sliding window; i(k) is the current value of the kth sampling point; The average current within the window is represented by the sample variance, which reflects the intensity and dispersion of the current signal fluctuations.

[0036] The formula for calculating kurtosis is: (3) (4) in, The standard deviation of the current window signal; The failure threshold is used to reflect the severity of the impact components (pulses, spikes) in the signal, and it is one of the most sensitive indicators of early wear.

[0037] Formula for calculating the zero-crossing time mean: (5) in, This is the sign function. The zero-crossing time mean is used to reflect the degree of periodicity of the current waveform.

[0038] Preferably, single-window computation can be completed on an embedded processor, such as an STM32 or a vehicle ECU, with a latency of <100ms, which can meet the real-time monitoring requirements of automotive and engineering machinery.

[0039] S3: Compare the time-domain statistical features with the baseline model, calculate the deviation of each feature, and weight and fuse them to obtain a health score.

[0040] Specifically, a baseline model is pre-established using multiple sets of data under healthy conditions. The baseline model includes the mean and standard deviation of features. The extracted time-domain statistical features are compared with the baseline model in real time, and the deviations of each feature are calculated and weighted to obtain a health score.

[0041] The specific steps to obtain a health score include: S31: For each sliding window, calculate the relative deviation of the four features relative to the healthy baseline: Amplitude change rate deviation: ; Variance bias: ; Kurtosis: ; Zero-crossing time mean deviation ; The subscript "O" indicates the baseline value of the corresponding characteristic in a healthy state, which is obtained in advance through a large amount of health data statistics.

[0042] S32: Weighted fusion of the various feature deviations: The weighting formula used is: ; The weights used vary depending on the application scenario. The weights used in this embodiment are as follows: .

[0043] This embodiment can also automatically optimize the baseline according to different vehicle models / operating conditions and adaptively adjust the threshold of the baseline model. It is effective for high-speed dynamic operating conditions of new energy vehicles and high-load impact scenarios of construction machinery, thus improving diagnostic robustness.

[0044] S4: Smoothly track the health score sequence. When the health score shows an upward trend for N consecutive windows, trigger an early warning and estimate the remaining lifespan through linear extrapolation.

[0045] Specifically, this embodiment uses the exponentially weighted moving average (EWMA) algorithm to smoothly track the health score sequence. When the health score shows an upward trend for multiple consecutive windows, it is determined to be in the wear development stage, and the remaining useful life is estimated by linear extrapolation.

[0046] The steps to determine the remaining useful life include: S41: Determine key parameters: Current health score after EWMA smoothing Failure threshold It is usually set to 1.25~1.5, or calibrated according to experiments; recently A sequence of health scores in a sliding window.

[0047] S42: Calculate the trend slope : ,in for The actual running time (hours) or number of cycles corresponding to each window.

[0048] S43: Perform linear extrapolation calculation: The final output is the remaining running time (hours) or the remaining number of loops.

[0049] This embodiment introduces EWMA trend tracking and linear extrapolation to achieve a leap from detection to prediction. It can provide early warning in the early stages of wear, i.e., when the kurtosis increases by 5.2% and the variance increases by 4.9%, which significantly reduces the risk of sudden downtime and maintenance costs.

[0050] S5: Outputs early warning information and prediction results, including health scores and estimated remaining life expectancy.

[0051] The actual health score range is generally between 0.95 and 1.60. Under normal health conditions, the baseline value of the health score is H(t)≈1.00. If the health score is less than 1.0, it may be noise or slight fluctuation, which is considered normal. If the health score is between 1.00 and 1.05, it is considered a slight deviation. If the health score is between 1.05 and 1.25, it is considered moderate wear. If the health score is greater than 1.30, it is considered severe wear.

[0052] Furthermore, this embodiment uses a combination of tiered thresholds and trend judgment to determine the wear development stage, as shown in Table 1 below. When the health score does not exceed 1.03, the transmission is in a healthy stage, and the system monitors normally. When the health score is (1.03, 1.08), the transmission is in an early warning stage, and the system enters the early warning zone and records the trend. When the health score is (1.08, 1.20), the transmission is in a mild wear stage, and the system begins to trigger a level one warning, suggesting attention. When the health score is (1.20, 1.35), the transmission is in a moderate wear stage, and the system begins to trigger a level two alarm, predicting RUL (Range, Underwear, and Lift). When the health score is greater than 1.35, the transmission is in a severe wear stage, triggering a level three alarm, and it is immediately recommended to stop operation.

[0053] Table 1: Wear Development Stages and Corresponding System Actions

[0054] The warning information and prediction results are output to the vehicle instrument panel or construction machinery monitoring system via the CAN bus.

[0055] To verify the effectiveness and feasibility of the real-time prediction method for gearbox gear wear based on motor current signal provided in this embodiment under the actual operating conditions of new energy vehicles, a real-time prediction experiment was conducted on the gearbox gear wear process of a certain type of electric logistics vehicle.

[0056] The experimental subject is the gearbox of a 2S310 electric logistics vehicle, equipped with a permanent magnet synchronous motor with a rated power of 60kW, a rated speed of 3000rpm, and a peak speed of up to 8000rpm. In the experimental platform, a Hall current sensor is used to collect current signals. This sensor has a measurement accuracy of ±1%, a sampling frequency of 5kHz, and only collects the U-phase current. The embedded processor uses an STM32F4 series chip to run the real-time prediction method described in this embodiment.

[0057] First, baseline data was collected with the transmission in a healthy state to obtain a baseline model. Under normal transmission operation, multiple sets of current signal data were collected within a typical operating range of 500–2000 rpm and 20–60 Nm load torque. Based on the collected data, the baseline mean and standard deviation of four time-domain statistical characteristic parameters were calculated. Specifically, the baseline mean of kurtosis under healthy conditions was approximately 2.8, and the baseline mean of variance was approximately 0.015.

[0058] Subsequently, a fault injection experiment was conducted to simulate the early fault state of slight wear on the gearbox gears. Specifically, a wear amount of 0.1–0.3 mm was artificially created on the gear tooth surface. Current signals under the wear state were re-acquired under the same operating conditions, and real-time preprocessing was performed. A sliding window method was used for preprocessing, with a window length of 1 second and an overlap rate of 50% between adjacent windows. Within each time window, key time-domain statistical features reflecting the signal state were calculated: amplitude change rate, variance, kurtosis, and mean zero-crossing time. Comparing the feature changes between the healthy state and the wear state, the results showed that the amplitude change rate increased by approximately 4.6%, the variance increased by approximately 4.9%, the kurtosis increased by approximately 5.2%, and the mean zero-crossing time decreased by approximately 5% under the wear state compared to the healthy state.

[0059] Further health score calculation and remaining service life prediction were performed. A weighted fusion strategy was used to comprehensively score the four feature parameters mentioned above, with the weights of kurtosis, variance, amplitude change rate, and zero-crossing time mean set to 0.4, 0.3, 0.2, and 0.1, respectively. The fused health score signal was smoothed using the exponentially weighted moving average (EWMA) method, with a smoothing coefficient α of 0.2. The early warning criterion for health score deviation was set as follows: when the deviation of the health score from the baseline threshold for four consecutive sliding windows exceeds 1.05 times, the system triggers an early wear warning. A linear extrapolation method was used to predict the remaining service life, i.e., based on the slope of the health score change trend over time, the remaining running time required for the gear to progress from its current state to the failure threshold was estimated. Experimental results show that this method can provide effective early warnings approximately 150 to 300 hours before the actual remaining service life of the gear.

[0060] The experimental results are as follows: Figure 3 The health score chart of the electric logistics vehicle's gearbox is displayed, showing the upward trend of the health score with the degree of wear. The detection accuracy is >92%, the false positive rate is <3%, and the average calculation delay is 65ms, which fully demonstrates the effectiveness and engineering applicability of the method proposed in this embodiment in early wear warning and remaining life prediction.

[0061] To verify the effectiveness and feasibility of the real-time prediction method for gearbox gear wear based on motor current signal provided in this embodiment under the actual operating conditions of an electric excavator, a real-time prediction experiment was conducted on the gearbox gear wear process of a certain model of 20-ton electric excavator.

[0062] The experimental subject was a planetary gearbox of a 20-ton electric excavator, equipped with an asynchronous drive motor with a rated power of 120kW, a rated speed of 1500 rpm, and a high-load torque of 400~600Nm. In the experimental platform, an external shunt current sensor was used to collect current signals. This sensor had a measurement accuracy of ±0.5% and a sampling frequency of 5kHz, collecting the U-phase current. An industrial-grade embedded board was used to run the real-time prediction method described in this embodiment.

[0063] First, baseline data is collected when the transmission is in a healthy state to obtain a baseline model. Data is collected under no-load to medium-load conditions to establish a baseline model, where the kurtosis baseline is approximately 3.1 and the variance baseline is approximately 0.028. Under high load, the kurtosis weight is automatically increased to 0.5.

[0064] Subsequently, a fault injection experiment was conducted to simulate the early fault state of slight wear on the gearbox gears. Specifically, 0.15mm of wear and localized cracks were injected into the gear surface. Current signals under the wear state were acquired under high-load excavation cycle conditions, and real-time preprocessing was performed. A sliding window method was used for preprocessing, with a window length of 1 second and a 50% overlap between adjacent windows. Within each time window, key time-domain statistical features reflecting the signal state were calculated: amplitude change rate, variance, kurtosis, and mean zero-crossing time. Comparing the characteristic changes between the healthy and worn states, the results showed that kurtosis is sensitive to impact, increasing by approximately 6.1% after wear.

[0065] Further health score calculations and remaining service life predictions were performed. A weighted fusion strategy was used to comprehensively score the four feature parameters mentioned above, with the weights for kurtosis, variance, amplitude change rate, and zero-crossing time mean set to 0.4, 0.3, 0.2, and 0.1, respectively. The fused health score signal was smoothed using the exponentially weighted moving average (EWMA) method, with a smoothing coefficient α of 0.2. The early warning criterion for health score deviation was set as follows: when the health score shows an upward trend for five consecutive sliding windows, the system triggers an early wear warning. A linear extrapolation method was used to predict the remaining service life, i.e., based on the slope of the health score change trend over time, the remaining operating time required for the gear to progress from its current state to the failure threshold was calculated. Experimental results show that this method can provide effective early warnings approximately 80 to 200 hours before the actual remaining service life of the gear.

[0066] The experimental results show that the robustness of the method is better than that of the traditional vibration method in high-noise environments, the early warning time is significantly longer, and the detection accuracy is >88%. Figure 4The comparison between the real-time gearbox gear wear prediction method based on motor current signal described in this invention and the traditional gearbox gear wear prediction method is shown, which fully demonstrates the effectiveness and engineering applicability of the method proposed in this embodiment in early wear warning and remaining life prediction.

[0067] The two experiments described above demonstrate that the real-time prediction method for gearbox gear wear based on motor current signals described in this embodiment is applicable to single-speed / multi-speed gearboxes in new energy vehicles and planetary gear / parallel shaft gearboxes in engineering machinery. Especially under variable speed and load conditions, by utilizing the sensitivity of kurtosis to impact components and the smoothing of trends by EWMA, early detection (tooth surface wear at the 0.1mm level) can be achieved, with a prediction lead time of several days to several weeks.

[0068] Example 2 This embodiment provides a real-time gear wear prediction system based on motor current signals, including: Acquisition module: Acquires the single-phase stator current signal of the motor running in the gearbox; Feature extraction module: preprocesses the single-phase stator current signal of the motor to obtain time-domain statistical features; Health score calculation module: compares time-domain statistical features with the baseline model, calculates the deviation of each feature, and weights and fuses them to obtain the health score; Trend prediction module: smoothly tracks the health score sequence, triggers an alert when the health score shows an upward trend for N consecutive windows, and estimates the remaining life expectancy through linear extrapolation; Early warning output module: Outputs early warning information and prediction results, including health score and estimated remaining lifespan.

[0069] The descriptions of each embodiment in the above embodiments have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0070] The proposed system can be implemented in other ways. For example, the system embodiments described above are merely illustrative, and the division of modules described above is only a logical functional division. In actual implementation, there may be other division methods. For example, multiple modules may be combined or integrated into another system, or some features may be ignored or not executed.

[0071] Example 3 This embodiment also provides an electronic device, including: one or more processors, one or more memories, and one or more computer programs; wherein, the processor is connected to the memory, and the one or more computer programs are stored in the memory. When the electronic device is running, the processor executes the one or more computer programs stored in the memory to cause the electronic device to perform the method described in Embodiment 1.

[0072] It should be understood that in this embodiment, the processor can be a central processing unit (CPU), or it can be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or any conventional processor, etc.

[0073] Memory may include read-only memory and random access memory, and provides instructions and data to the processor. A portion of memory may also include non-volatile random access memory. For example, memory may also store information about the device type.

[0074] In the implementation process, each step of the above method can be completed by the integrated logic circuits in the processor hardware or by software instructions.

[0075] The method in Embodiment 1 can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor. The software modules can reside in readily available storage media in the art, such as random access memory, flash memory, read-only memory, programmable read-only memory, electrically erasable programmable memory, or registers. This storage medium is located in memory; the processor reads information from the memory and, in conjunction with its hardware, completes the steps of the above method. To avoid repetition, a detailed description is not provided here.

[0076] Those skilled in the art will recognize that the units and algorithm steps described in connection with the various examples of this embodiment can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this invention.

[0077] Example 4 This embodiment also provides a storage medium for storing computer instructions, which, when executed by a processor, complete the method described in Embodiment 1.

[0078] Example 5 This embodiment also provides a computer program product, including a computer program that, when run on one or more processors, implements the method described in Embodiment 1.

[0079] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for real-time prediction of gearbox gear wear based on motor current signals, characterized in that, include: Obtain the single-phase stator current signal of the motor running the gearbox; The time-domain statistical characteristics of the single-phase stator current signal of the motor are obtained by preprocessing. The time-domain statistical features are compared with the baseline model, the deviations of each feature are calculated, and the weighted fusion is used to obtain the health score. The health score sequence is smoothly tracked. When the health score shows an upward trend for N consecutive windows, an early warning is triggered, and the remaining lifespan is estimated by linear extrapolation. It outputs early warning information and prediction results, including health scores and estimated remaining life expectancy.

2. The method for real-time prediction of gearbox gear wear based on motor current signal according to claim 1, characterized in that, Acquiring the single-phase stator current signal of the motor during gearbox operation includes: acquiring the single-phase stator current signal of the motor through a current sensor, wherein the current sensor is a non-contact current sensor, using a Hall effect clamp or a shunt sensor.

3. The method for real-time prediction of gearbox gear wear based on motor current signal according to claim 1, characterized in that, Preprocessing the single-phase stator current signal of the motor yields time-domain statistical characteristics, including: The acquired current signal is segmented into sliding windows, and time-domain statistical characteristics are calculated within each window.

4. The method for real-time prediction of gearbox gear wear based on motor current signal according to claim 3, characterized in that, The time-domain statistical features include four components: amplitude change rate, variance, kurtosis, and mean value at zero crossing time.

5. The method for real-time prediction of gearbox gear wear based on motor current signal according to claim 1, characterized in that, The time-domain statistical features are compared with the baseline model, the deviations of each feature are calculated, and the weighted fusion is used to obtain a health score, including: A baseline model is pre-established using multiple sets of data under healthy conditions. The baseline model includes the mean and standard deviation of the features. The extracted time-domain statistical features are compared with the baseline model in real time, the deviation of each feature is calculated and weighted and fused to obtain a health score.

6. The method for real-time prediction of gearbox gear wear based on motor current signal according to claim 1, characterized in that, The health score sequence is smoothly tracked. When the health score shows an upward trend for N consecutive windows, an early warning is triggered, and the remaining life expectancy is estimated by linear extrapolation, including: An exponentially weighted moving average algorithm is used to smoothly track the health score sequence. When the health score shows an upward trend for multiple consecutive windows, it is determined to be in the wear and tear development stage, and the remaining useful life is estimated by linear extrapolation.

7. A real-time prediction system for gearbox gear wear based on motor current signals, characterized in that, include: Acquisition module: Acquires the single-phase stator current signal of the motor running in the gearbox; Feature extraction module: preprocesses the single-phase stator current signal of the motor to obtain time-domain statistical features; Health score calculation module: compares time-domain statistical features with the baseline model, calculates the deviation of each feature, and weights and fuses them to obtain the health score; Trend prediction module: smoothly tracks the health score sequence, triggers an alert when the health score shows an upward trend for N consecutive windows, and estimates the remaining life expectancy through linear extrapolation; Early warning output module: Outputs early warning information and prediction results, including health score and estimated remaining lifespan.

8. An electronic device, characterized in that, include: Memory is used to store computer-readable instructions in a non-transitory manner. as well as Processor, for executing the computer-readable instructions, When the computer-readable instructions are executed by the processor, they perform the real-time prediction method for gearbox gear wear based on motor current signals as described in any one of claims 1-6.

9. A storage medium, characterized in that, Non-transitory storage of computer-readable instructions, wherein, when executed by a computer, the non-transitory computer-readable instructions perform the real-time prediction method for gearbox gear wear based on motor current signals as described in any one of claims 1-6.

10. A computer program product, characterized in that, The method includes a computer program that, when running on one or more processors, implements the real-time prediction method for gearbox gear wear based on motor current signals as described in any one of claims 1-6.

Images