A rotor life prediction method and system based on big data

Abstract

The present application relates to the technical field of life prediction, more particularly, the present application relates to a rotor life prediction method and system based on big data, the rotor life prediction method comprises: collecting the operating parameters and working condition parameters of the rotor, and clustering the working condition parameters to obtain multiple working conditions; for any working condition, the target feature vector of each time is constructed, the target feature vector includes: operating parameters, the health state of the rotor and the health attenuation index of the motor. The present application realizes the comprehensive characterization of the rotor degradation state by constructing the comprehensive health characteristics of the rotor vibration and the electrical state of the motor (three-phase current, power factor). At the same time, the clustering algorithm is used to divide the operating conditions, and a special prediction model is trained for each working condition, which greatly improves the prediction accuracy and generalization ability of the model under complex and variable conditions, making the life prediction more accurate and reliable.

Description

Technical Field

[0001] This invention relates to the field of lifespan prediction technology. More specifically, this invention relates to a rotor lifespan prediction method and system based on big data. Background Technology

[0002] The rotor is the core load-bearing component of various rotating machinery (such as steam turbines, aero engines, and electric motors), and its health status directly affects the safe and reliable operation of the entire equipment. During long-term service, rotors gradually suffer damage due to fatigue, wear, corrosion, and other factors, which accumulates and eventually leads to sudden failure, causing serious safety accidents and huge economic losses. Therefore, monitoring the health of rotors and accurately predicting their remaining service life is of paramount importance for achieving predictive maintenance, ensuring equipment safety, and reducing operating costs.

[0003] Existing technologies typically employ machine learning algorithms such as Random Forest to construct predictive models by fusing information from multiple sensors. However, these methods have inherent limitations in application: Firstly, they mostly rely solely on mechanical signals such as rotor vibration and temperature, neglecting the coupled influence of the drive system's (e.g., motor) health status on rotor load and dynamic characteristics, resulting in incomplete feature information in the input model. Secondly, rotors often experience various changing operating conditions (such as different speeds and loads) during actual operation, and the degradation patterns and data characteristics under different operating conditions vary significantly. If a single model is used to uniformly process and predict data under all operating conditions, it is difficult to capture the degradation patterns under specific operating conditions, leading to poor generalization ability and difficulty in guaranteeing accuracy of the predictive model. Summary of the Invention

[0004] This invention provides a rotor life prediction method and system based on big data, aiming to solve the problem in related technologies that if a single model is used to uniformly process and predict data under all working conditions, it is difficult to capture the degradation pattern under specific working conditions, which leads to poor generalization ability and low accuracy of the prediction model.

[0005] In a first aspect, the present invention provides a rotor life prediction method based on big data, comprising: collecting rotor operating parameters and working condition parameters, and clustering the working condition parameters to obtain multiple working conditions; for any working condition, constructing a target feature vector at each moment, the target feature vector including: operating parameters, rotor health status, and motor health degradation index; wherein, the rotor health status reflects the energy proportion of the power spectrum of the rotor vibration signal at that moment within the set of healthy feature frequency bands; the method for obtaining the set of healthy feature frequency bands includes: selecting a health reference moment based on the motor health index, and constructing a reference power spectrum using its vibration signal; then performing peak detection in the reference power spectrum, identifying peaks greater than a threshold, constructing narrowbands based on the characteristic frequencies of the peaks, and merging them to obtain the set of healthy feature frequency bands under that working condition, wherein, the motor health index reflects the degree of imbalance of the three-phase current of the motor and the degree of deviation of the power factor, and the degree of deviation of the motor health index relative to its health reference value is used as the health degradation index of the motor at that moment; inputting the target feature vector at the current moment into the remaining service life prediction model of the corresponding working condition, and outputting the predicted remaining service life of the rotor. By combining the health status of the drive motor (quantified by three-phase current imbalance and power factor deviation) with the rotor vibration status, a more comprehensive target feature vector is constructed. This is the first time that the impact of the drive system on rotor health has been captured at the system level, solving the problem of incomplete feature information. In addition, this method improves the accuracy and reliability of rotor remaining service life prediction by clustering operating parameters and constructing a dedicated prediction model for each operating condition.

[0006] Furthermore, a reference power spectrum is constructed using the vibration signal, including: performing a Fourier transform on the vibration signal at each healthy reference time to obtain the power spectrum; and calculating the median of the power spectrum at each frequency point to obtain the reference power spectrum. Constructing the reference power spectrum by calculating the "median" of the power spectrum at each frequency point across multiple healthy times effectively suppresses accidental noise interference and instantaneous fluctuations, providing a solid foundation for improving the accuracy of rotor health status assessment.

[0007] Furthermore, the calculation method for the motor health index includes multiplying the reciprocal of the three-phase current imbalance by the reciprocal of the deviation between the actual power factor and the rated power factor of the motor, and using this product as the motor health index. Compared to a single evaluation index, this index can more comprehensively and accurately reflect the overall health level of the motor, providing a more reliable input for the subsequent construction of system-level feature vectors.

[0008] Furthermore, the method for obtaining the health benchmark value includes: obtaining the motor health index during the early operation phase of the motor, and selecting the largest motor health index as the health benchmark value.

[0009] Furthermore, it also includes: determining the current working condition, the distance between the current working condition parameters and the cluster centers of all clusters, and selecting the working condition corresponding to the nearest cluster as the current working condition.

[0010] Furthermore, the operating parameters are clustered to obtain multiple operating conditions, including: using the K-Means algorithm to cluster the operating parameters to obtain multiple clusters corresponding to multiple operating conditions, wherein the elbow method is used to determine the number of clusters.

[0011] Furthermore, the method also includes smoothing the time series of the motor health index before calculating the health degradation index at each time point. This method can effectively filter out short-term fluctuations caused by measurement noise or minor fluctuations in operating conditions, thereby extracting a stable signal that better reflects the true, long-term degradation trend of the equipment. This makes the calculated health degradation index more monotonic and stable, which is beneficial to improving the robustness of life prediction.

[0012] Furthermore, a narrowband is constructed based on the characteristic frequencies of the peak values, including: constructing a narrowband with a width of ±1Hz centered on each characteristic frequency.

[0013] Furthermore, the early operation phase is the phase within a preset time period after the motor's first operation.

[0014] In a second aspect, the present invention also provides a rotor life prediction system based on big data, comprising a processor and a memory, wherein the memory stores a computer program, and the processor executes the computer program to implement the rotor life prediction method based on big data as described above.

[0015] Beneficial effects: By constructing a comprehensive health feature that integrates rotor vibration and motor electrical conditions (three-phase current, power factor), a comprehensive characterization of rotor degradation is achieved. Simultaneously, by using clustering algorithms to classify operating conditions and training a dedicated prediction model for each condition, the model's prediction accuracy and generalization ability under complex and variable operating conditions are greatly improved, making life prediction more accurate and reliable. Attached Figure Description

[0016] Figure 1 This is a flowchart illustrating rotor life prediction according to an embodiment of the present invention;

[0017] Figure 2 This is a flowchart illustrating the calculation of the motor health degradation index at various times according to an embodiment of the present invention. Detailed Implementation

[0018] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0019] like Figure 1 As shown, S101: Collects rotor operating data and divides operating conditions.

[0020] Specifically, various sensors deployed on rotating machinery collect operating parameters and condition parameters covering the entire lifecycle of the rotor. These operating parameters include, but are not limited to, vibration signal characteristics, main bearing temperature, gearbox oil temperature, and rotor temperature; the condition parameters include rotational speed and load. To eliminate the influence of differences in dimensions and numerical ranges among different features, the collected raw data needs to be standardized. As a preferred approach, this embodiment employs the Z-score standardization method to ensure that all features are on the same comparable scale.

[0021] Next, the operating conditions of the rotor are categorized. Specifically, unsupervised clustering algorithms are used to cluster the operating parameters, resulting in multiple clusters corresponding to various operating conditions. For example, the K-Means algorithm can be used to divide historical data into K independent clusters, each cluster representing a specific operating condition. It should be noted that the number of clusters K can be determined using the elbow method.

[0022] S102: Construct the target feature vector under each working condition.

[0023] For each working condition identified in the above steps, a target feature vector for model training is independently constructed. Taking time t under the k-th working condition as an example, the target feature vector at time t is constructed. , , For the first The rotor temperature at time t under the given operating condition. For the first The temperature of the main bearing at time t under the given operating condition. For the first The ambient temperature of the rotor at time t under the given operating conditions. Indicates the first The health status of the rotor at time t under various operating conditions. For the first The motor health degradation index at time t under various operating conditions.

[0024] like Figure 2 As shown, S1021: Calculate the health status of the rotor at each moment.

[0025] Specifically, the rotor's health status is an indicator that quantifies the rotor's current health level. The logic behind constructing this health status is as follows: In a healthy rotor, the vibration energy is mainly concentrated at the rotational frequency and a few characteristic frequencies such as its low harmonics; however, when the rotor experiences early faults (such as microcracks or localized wear), it will trigger a nonlinear response, causing the vibration energy to diffuse into a wider frequency band or non-harmonic frequencies. Therefore, the rotor's health status can be assessed by quantifying the degree of energy concentration.

[0026] In this embodiment, the formula for calculating the rotor's health status is as follows: In the formula, For the first The health status of the rotor at time t under various operating conditions. For the first The power spectral density function of the vibration signal at time t under the given working condition. For the first The set of frequency bands corresponding to health characteristics under various operating conditions, which includes multiple narrow bands. This is the lower limit of narrowband. The upper limit of narrowband, For the first The sum of vibrational energy at time t under various operating conditions within the set of healthy characteristic frequency bands. For the first The total energy of the vibration signal at time t under a certain working condition.

[0027] As can be seen from the above formula, when the rotor is in a very healthy state, its vibration energy is highly concentrated in the healthy characteristic frequency band set. At this point, the numerator is approximately equal to the denominator. The value approaches 1. As equipment degrades, energy leaks to other frequency bands, causing molecules to shrink, thus... The value then decreases. Therefore, The degree of reduction directly reflects the degree of deterioration in the rotor's health condition.

[0028] No. The method for obtaining the set of healthy characteristic frequency bands corresponding to various operating conditions includes: Under the k-th operating condition, calculating the motor health index of the rotor at each moment, and selecting the N moments with the largest motor health index as the health reference moments, where N ranges from 50 to 100. Then, treating the motor at the health reference moments as a motor operating under normal conditions, statistically analyzing the rotational speed at each health reference moment under the k-th operating condition and calculating its corresponding fundamental frequency. Since it is the same operating condition, these fundamental frequencies are relatively similar. Performing a Fourier transform on the vibration signal at each moment to obtain the power spectrum. Then, calculating the median of these power spectra at each frequency point to construct a reference power spectrum representing the vibration characteristics of a healthy rotor under that operating condition.

[0029] Finally, peak detection is performed on the reference power spectrum to identify all significant and stable energy peaks. For example, the peak detection threshold can be set to a point where the peak's apex is more than three times the median of the entire reference power spectrum. For each detected peak, its frequency position is determined as a characteristic frequency, and a narrow band is constructed around this characteristic frequency; for example, a narrow band with a width of ±1 Hz can be constructed. All narrow bands are merged to obtain the set of healthy characteristic frequency bands for the k-th operating condition. .

[0030] The formula for calculating the motor health index of the rotor at each moment is as follows: In the formula, This represents the motor health index at time t. , and Let represent the three-phase currents of the motor at time t, respectively. This represents the average value of the three-phase current of the motor at time t. The actual power factor of the motor at time t. This represents the rated power factor under load at time t. These are preset parameters to avoid a denominator of 0.

[0031] In the formula, The deviation of the three-phase current from the average value was calculated, reflecting the degree of current imbalance. The larger the value, the greater the current imbalance. The smaller the value, the better. The absolute deviation between the actual power factor of the motor and the rated power factor under the load is calculated, which reflects whether the motor's performance is normal at the current operating time. The larger the deviation between the actual power factor and the rated value, the smaller this value, and the worse the motor performance.

[0032] S1022: Calculate the motor health degradation index at each time point.

[0033] First, determine the health baseline value of the motor. Specifically, this is determined by finding the maximum value of the motor health index during the early operation phase of the motor. The early operation phase can be a period within a preset time after the first operation of the motor, which can be 10 days or one month, etc.

[0034] It is important to note that, considering the potential fluctuations in instantaneous motor health index due to minor variations in operating conditions or measurement noise, smoothing of the time series of the motor health index is necessary to more accurately capture long-term, stable degradation trends. One suitable smoothing method is the exponential moving average, a commonly used method that will not be elaborated upon here.

[0035] Then, based on the baseline health value and the smoothed health index, the motor health degradation index at each time point can be calculated. This is based on the assumption that the degradation index should be the normalized deviation of the smoothed health index from the baseline health value. In the early stages of motor operation, its health state is close to optimal. The smoothed health index at each time point is closer to the motor's baseline health value, and at this point, the motor health degradation index is close to 0, indicating almost no health degradation. As operating time increases, motor performance gradually deteriorates, causing its health index to show a downward trend. The smoothed health index decreases accordingly, resulting in the motor's health degradation index monotonically increasing from 0 to 1. When the motor completely fails, the smoothed health index approaches 0, and at this moment, the motor's health degradation index approaches 1.

[0036] S103: Construct a prediction model for the remaining useful life under various operating conditions.

[0037] An independent remaining useful life prediction model is trained for each operating condition. In this embodiment, the random forest algorithm is selected as the prediction model. Random forest is an ensemble learning algorithm that improves the accuracy and robustness of the model by constructing multiple decision trees and integrating their prediction results (such as voting or averaging). It is also suitable for the high-dimensional, non-linear features constructed in this embodiment.

[0038] First, construct the training dataset. For each working condition, the training dataset consists of label pairs {( , Composed of )} This represents the target feature vector at time t under the k-th working condition in history. This represents the rotor's actual remaining service life at historical time t. In the formula, This refers to the time of failure or replacement of the rotor in the historical log. A timestamp for a specific historical moment.

[0039] Next, the constructed training dataset is divided into a training set and a validation set. Then, the key hyperparameters of the random forest are set, including: the number of decision trees, the maximum depth of each tree, the maximum number of features considered when splitting nodes, and the minimum number of samples for leaf nodes. The training data is input into the model, and the mean squared error or mean absolute error is used as the loss function to optimize the splitting strategy for each tree. Finally, when the predetermined number of decision trees in the random forest have been constructed based on the training dataset, and the performance evaluation on the validation set meets the expected metrics, the model training is considered complete, and the remaining useful life prediction model is obtained. At this point, the obtained remaining useful life prediction model can be used to input the target feature vector at any time and output the corresponding predicted remaining useful life value.

[0040] S104: Based on the remaining service life prediction model, predict the current rotor life.

[0041] After completing the dedicated model training for all operating conditions, the system can be used to predict the rotor's remaining service life in real time. The specific process is as follows: First, the current operating condition parameters (speed, load) and running parameters are collected in real time, and a target feature vector for the current moment is constructed according to the steps described above. Second, the current operating condition parameters are compared with the K cluster centers obtained in step S101, and the cluster whose distance to all cluster centers is closest to the current operating condition parameters is selected as the operating condition to which the current moment belongs. Then, the target feature vector for the current moment is input into the remaining service life prediction model, and the predicted remaining service life of the rotor at the current moment is output.

[0042] The present invention also provides a rotor life prediction system based on big data. The system includes a processor and a memory, the memory storing computer program instructions. When the computer program instructions are executed by the processor, they implement the rotor life prediction method based on big data according to the first aspect of the present invention.

[0043] The system also includes other components well known to those skilled in the art, such as communication buses and communication interfaces, the settings and functions of which are known in the art and therefore will not be described in detail here.

[0044] In this invention, the aforementioned memory can be any tangible medium containing or storing a program that can be used or combined with an instruction execution system, apparatus, or device. For example, a computer-readable storage medium can be any suitable magnetic or magneto-optical storage medium, such as Resistive Random Access Memory (RRAM), Dynamic Random Access Memory (DRAM), Static Random Access Memory (SRAM), Enhanced Dynamic Random Access Memory (EDRAM), High-Bandwidth Memory (HBM), Hybrid Memory Cube (HMC), etc., or any other medium that can be used to store desired information and can be accessed by an application, module, or both. Any such computer storage medium can be part of a device or accessible to or connected to a device. Any application or module described in this invention can be implemented using computer-readable / executable instructions stored or otherwise maintained on such a computer-readable medium.

[0045] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention.

Claims

1. A big data based rotor life prediction method, characterized in that, The rotor life prediction method comprises: collecting operating parameters and working condition parameters of the rotor, and clustering the working condition parameters to obtain multiple working conditions, comprising: clustering the working condition parameters by using a K-Means algorithm to obtain multiple clustering clusters corresponding to multiple working conditions, wherein the number of clustering clusters is determined by using an elbow method; For any working condition, a target feature vector at each time is constructed, the target feature vector comprising: operating parameters, a health state of the rotor, and a health degradation index of the motor; wherein the health state of the rotor reflects the energy proportion of the power spectrum of the rotor vibration signal at the time in the health feature band set; The health feature band set acquisition method comprises: selecting a health reference time according to a motor health index, and constructing a reference power spectrum using the vibration signal thereof; then performing peak value detection in the reference power spectrum, identifying peak values greater than a threshold, and constructing narrow bands according to the characteristic frequencies of the peak values to obtain the health feature band set of the working condition, wherein the motor health index reflects the imbalance degree of the motor three-phase current and the deviation degree of the power factor, and the deviation degree of the motor health index relative to the health reference value is taken as the health degradation index of the motor at the time; An independent remaining useful life prediction model is trained for each working condition, and the target feature vector at the current time is input into the remaining useful life prediction model corresponding to the working condition to output the predicted remaining useful life of the rotor.

2. The big data based rotor life prediction method of claim 1, wherein, The reference power spectrum is constructed using the vibration signal, comprising: Performing Fourier transform on the vibration signal of each health reference time to obtain a power spectrum; Calculating the median of the power spectrum at each frequency point to obtain the reference power spectrum.

3. The big data based rotor life prediction method of claim 1, wherein, The motor health index calculation method comprises: Taking the product of the negative exponential function of the three-phase current imbalance degree and the reciprocal of the deviation between the actual power factor of the motor and the rated power factor as the motor health index.

4. The big data based rotor life prediction method of claim 1, wherein, The health reference value acquisition method comprises: Obtaining the motor health index of the motor in the early running stage, and selecting the maximum motor health index as the health reference value.

5. The big data based rotor life prediction method of claim 1, wherein, Further comprising: Determining the working condition to which the current time belongs, wherein the working condition corresponding to the nearest clustering cluster is selected as the working condition to which the current time belongs based on the distance between the current working condition parameter and the clustering center of all clustering clusters.

6. The big data based rotor life prediction method of claim 1, wherein, Further comprising: Before calculating the health degradation index of the motor at each time, the time series of the motor health index is smoothed.

7. The big data based rotor life prediction method of claim 1, wherein, The narrow band is constructed according to the characteristic frequency of the peak value, comprising: Centering each characteristic frequency, a narrow band with a width of ±1 Hz is constructed.

8. The big data based rotor life prediction method of claim 4, wherein, The early running stage is a stage within a preset time period after the first running of the motor. 9.A rotor life prediction system based on big data, comprising a processor and a memory, characterized in that, The memory stores a computer program, and the processor executes the computer program to implement the rotor life prediction method based on big data according to any one of claims 1-8.

Images