Rotor unbalance state measurement system and identification method thereof

Abstract

The application relates to a rotor unbalance state measurement system and an identification method thereof, the measurement system adopts a two-dimension measurement mode to more completely and efficiently collect data, and simultaneously selects multiple measurement sections for detection, the formed shaft center motion track is more accurate, the identification method selects characteristic coordinates by dividing curvature regions, judges the unbalance state by calculating the eccentricity of the shaft center motion track, does not depend on complex dynamic parameter modeling, the feature extraction is intuitive and effective, and the identification can be simply and quickly completed on the basis of meeting a higher accuracy.

Description

Technical Field

[0001] This invention relates to the field of bearing and rotor system testing, and more specifically to a rotor imbalance state measurement system and identification method. Background Technology

[0002] As the core component of rotating machinery, the rotor system can cause a deviation between the rotor's center of mass and geometric center during operation due to uneven material distribution, component detachment, or rotor component wear. During rotor rotation, this center of mass shift will cause an unbalanced force on the rotor. This unbalanced force reacts on the rotor, causing it to vibrate unbalancedly, thus leading to an unbalanced operating state. This can easily cause shaft system failures and significantly affect equipment stability and working efficiency.

[0003] Traditional methods for identifying unbalanced rotor systems typically use single-dimensional vibration signals to measure the raw data of the unbalanced state. However, these methods rely excessively on expert experience, resulting in low accuracy and poor adaptability. In contrast, the shaft center trajectory of a rotor system contains multi-dimensional motion information, and the unbalanced motion state of the rotor system can be obtained through analysis and calculation of this trajectory.

[0004] Existing technologies include rotor fault diagnosis and identification methods based on shaft center trajectory analysis. However, current mainstream technologies mostly focus on the overall shape identification of the shaft center trajectory. Although the shape of the shaft center trajectory can reflect typical fault characteristics, these existing identification methods classify based on known patterns according to the trajectory shape, which has a weak ability to identify unknown faults. When multiple faults occur coupled, the shape of the shaft center trajectory becomes complex and ambiguous, resulting in low identification accuracy. In addition, existing technologies rely only on single-section or bidirectional displacement sensors and do not achieve multi-measurement point collaborative analysis, making it difficult to distinguish between local disturbances and systemic imbalances. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a rotor imbalance state measurement system and its identification method, which solves the problems of limited information dimensions and low identification accuracy in traditional identification methods.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is as follows: A rotor imbalance measurement system is disclosed. The system simultaneously measures multiple cross-sections of the rotor along the radial direction. Two sensors are arranged in the plane of each measurement cross-section, with the line connecting the two sensors to the shaft center forming a 90° angle, and the sensors are not in contact with the rotor. The sensors are eddy current displacement sensors. Each sensor is connected to a preamplifier, which converts the minute eddy current signals collected by the sensors into voltage signals and aggregates them to an EL3102 analog input module. The EL3102 analog input module can convert the voltage signals into digital signals. The EL3102 analog input module is connected to an EtherCAT bus via a bus coupler EK110 and communicates with TwinCAT. TwinCAT is used to write periodic data acquisition programs, set the sampling frequency, and export the acquired vibration displacement signals in CSV format.

[0007] A method for identifying rotor imbalance states, the method comprising the following steps: (1) Set up the measurement system and collect data: Set up the measurement system on the rotor shaft system. For a two-point support shaft system, select measurement sections and arrange sensors on the outer side of the two-point support and the middle part of the rotor respectively. At the same time, collect the vibration displacement signals of the three measurement sections. Based on the collected data, synthesize the shaft center motion trajectory. (2) Preprocessing the axis center trajectory: The synthesized axis center trajectory is centered and normalized to make the axis center trajectory closer to the origin and in a position where Within the range; (3) Calculate characteristic indicators: This includes three steps: dividing the curvature region, selecting characteristic coordinates, and calculating the eccentricity. (3.1) Dividing the curvature region: Construct vectors by discrete points on the axis trajectory, calculate the included angle using the cosine theorem, and distinguish between large curvature points and general curvature points based on the set angle threshold β, thereby realizing the division of the curvature region of the trajectory shape; (3.2) Select feature coordinates, and extract key points using an adaptive interval strategy based on the curvature region division results. Dense sampling is performed in areas with large curvature, and sparse sampling is performed in areas with general curvature. Finally, a concise and representative feature point sequence is generated, which provides a basis for subsequent eccentricity calculation. (3.3) Calculate the eccentricity. Based on the extracted feature coordinate sequence, calculate the Euclidean distance from each point to the origin to generate the eccentricity sequence H, which is used to quantify the degree of geometric asymmetry and unbalance of the rotor system. (4) Set an eccentricity threshold and determine the rotor state: Combine the characteristics of the rotor system and find the standard to set a specific eccentricity threshold. When the eccentricity at a certain point of a certain measurement section exceeds the specified threshold, it can be determined that the measurement section is in an unbalanced state.

[0008] Furthermore, in step (2), the specific steps for centering the axis trajectory are as follows: first, the original data is labeled according to the time series; the original data... and Divided into Independent single samples and , As shown in equation (1): (1) Centralized processing is shown in formula (2). (2) in, and This indicates data that has undergone centralized processing. and This represents the mean of a single sample.

[0009] Furthermore, in step (2), the specific steps for normalizing the axis trajectory are as follows: To ensure the consistency of each sample data, the single sample that has undergone centering processing is normalized. and After performing min-max normalization, as shown in formula (3) (3) in, and This indicates data that has undergone normalization.

[0010] Further, step (3.1) specifically involves: using the preprocessed single-sample data... and Synthetic axis trajectory ,in The axis trajectory consists of a series of coordinate points; take any point... Take the second data point at intervals of m points. and the third data point The three points form vectors with the origin respectively. , and Thus, we obtain the vector. and The angle between the two vectors can be calculated using the law of cosines. Specifically, it is shown in the following formulas (4), (5) and (6): (4) (5) (6) in, Representing vectors and The included angle, Representing vectors The model; Set angle threshold By comparing angles, the type of point B can be determined. By classifying point B as a point of high curvature and others as points of normal curvature, and performing the same calculation on all coordinate points in the axis trajectory R, ​​the regions of high curvature and normal curvature can be divided.

[0011] Further, step (3.2) uses an adaptive interval strategy to extract key points, specifically: the axis trajectory graphic generally follows the principle of selecting feature coordinates with equal intervals, but different intervals are implemented for different curvature regions; the feature point selection principle with interval L is implemented for large curvature regions, and the feature point selection principle with interval multiple L is implemented for general curvature regions, so as to realize the selection of adaptive interval feature points based on the curvature of the region.

[0012] Furthermore, the interval L is determined based on the number of coordinate points n of the axis trajectory R. When the number of coordinate points n of the axis trajectory is 6000, the interval L is determined to be 5. Generally, the curvature region follows the feature point selection principle of interval 2L, i.e., 10. After completing the feature point interval selection, a series of feature coordinate sequences are obtained. M represents the number of feature coordinates.

[0013] Furthermore, the specific method for calculating the eccentricity in step (3.3) is as follows: the characteristic coordinate sequence can be obtained from step (3.2). ,in M is the number of characteristic coordinates. The distance between each coordinate point and the origin, i.e., the eccentricity h, is calculated as shown in formula (7): (7) All feature coordinates can be calculated using formula (7) to obtain a data sequence, namely the eccentricity sequence. The eccentricity directly reflects the magnitude of the unbalanced force in the rotor system; the larger the value, the more severe the unbalance in the rotor system.

[0014] Furthermore, in step (4), the identification of the unbalanced state of the entire rotor system needs to be achieved through comprehensive calculation of the rotor system unbalance, as follows: Rotor system imbalance = Σ(cross-sectional imbalance value) (Section weight); Hazard value = Single-section threshold Adjustment factor; Among them, the unbalanced value of the cross section is the maximum value of the eccentricity calculated for the cross section. The cross section weight is allocated according to the deformation of the cross section location. The greater the deformation, the higher the weight. The sum of the weights of all cross sections is 1. The single-section threshold is the eccentricity threshold determined by the standard. Considering the engineering safety margin criterion, a conservative adjustment coefficient is set for this single-section threshold. Finally, the calculated rotor system imbalance is compared with the danger value. If it is determined that the imbalance exceeds the danger value, the rotor system is considered to be in an unbalanced state.

[0015] Furthermore, for the shaft system supported at both ends, the weighting coefficient of the three selected measurement sections is 0.2 for the sections at both ends and 0.6 for the section at the middle of the rotor; the adjustment coefficient is 0.8.

[0016] Beneficial effects: 1. Compared with the traditional identification method that only uses one-dimensional vibration signals, the present invention uses a two-dimensional measurement method to collect data more completely and efficiently. Furthermore, the present invention selects multiple measurement sections simultaneously in a two-dimensional manner for detection, so that the resulting axis motion trajectory will be more accurate and will also be beneficial for subsequent efficient analysis and identification.

[0017] 2. The method for extracting axial motion trajectory features proposed in this invention selects feature coordinates and calculates eccentricity by dividing curvature regions, thereby effectively extracting features of complex motion trajectories. This method has strong generalization ability and the feature extraction is intuitive and effective. Compared with the traditional method of fault identification based on trajectory shape, this invention achieves a deepening from "overall shape recognition" to "local geometric analysis", realizing a paradigm leap from "looking at the shape" to "calculating curvature", which can simplify the recognition process and improve the recognition accuracy.

[0018] 3. Compared with most current methods that use complex algorithms to identify the unbalanced state of a rotor system, this invention determines the unbalanced state by calculating the eccentricity of the shaft center trajectory. It does not rely on complex dynamic parameter modeling and can complete the identification simply and quickly while maintaining high accuracy, which has great advantages.

[0019] 4. The rotor system imbalance determination of this invention adopts a dual threshold mechanism of "weighted synthesis + safety margin", uses multiple sensors to collect data at multiple sections, and proposes a weighted synthesis eccentricity. By analyzing the data from multiple sections in a collaborative manner, it can quantify the contribution of different positions to the overall vibration, comprehensively evaluate the system imbalance state, and effectively distinguish between local disturbances and systemic imbalance. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the rotor imbalance state measurement system of the present invention. In the figure, numbers 1-6 all represent sensor probes.

[0021] Figure 2 This is a flowchart of the rotor imbalance state identification method of the present invention.

[0022] Figure 3 This is a set of raw data curves collected in an embodiment of the present invention.

[0023] Figure 4 This is a schematic diagram of the normalized axis trajectory and the distribution of its boundary extreme value coordinates.

[0024] Figure 5 A schematic diagram illustrating the curvature division and the selected feature coordinates.

[0025] Figure 6 This is the eccentricity curve calculated at the end of the embodiment of the present invention. Detailed Implementation

[0026] The present invention will now be described in further detail with reference to the accompanying drawings and specific embodiments.

[0027] The following examples illustrate the rotor imbalance measurement system using a two-point support shaft system.

[0028] The rotor imbalance state measurement system of the present invention measures the vibration displacement signal of the rotor using multiple eddy current displacement sensors, specifically as follows: Figure 1 As shown, in this embodiment, the rotor system is a two-point supported shaft system, the rotor system speed can reach up to 1200 r / min, and the sampling frequency is set to 16 kHz.

[0029] The measurement system simultaneously measures multiple cross-sections of the rotor along the radial direction. Two sensors are arranged in the plane of each measurement cross-section, with the line connecting the two sensors to the rotor axis forming a 90° angle. The sensors are non-contact with the rotor, ensuring no disruption to the rotor system's operation. The system is defined as follows: the rotor axis direction is the Z-axis; the direction perpendicular to the ground upwards is the Y-axis; and the direction along the radial direction of the cross-section at a 90° angle to the Y-axis is the X-axis. Figure 1 In the embodiment shown, two sensors within the same measurement section are respectively located on the X-axis and Y-axis.

[0030] When deploying sensors, multiple typical characteristic sections are selected based on the operating characteristics of each part of the rotor for simultaneous multi-section measurement. Since this embodiment uses a two-point support shaft system, sensors are deployed at two measurement sections on the outer sides of the two supports to measure the vibration displacement signals at the support sections. In addition, to verify whether there is local imbalance in the rotor, sensors are also deployed at a measurement section in the middle of the rotor. This simultaneous measurement of three typical characteristic sections makes the measurement data more comprehensive and the identification results more accurate.

[0031] In this embodiment, three measurement sections were selected, with two sensors arranged at each section. The entire measurement system uses six eddy current displacement sensors, i.e., a total of six probes (e.g., Figure 1 As shown in reference numerals 1-6, each sensor probe is connected to a preamplifier, which is connected to an EL3102 analog input module. The EL3102 analog input module is connected to the EtherCAT bus via a bus coupler EK110, and then communicates with TwinCAT.

[0032] When using this measurement system for signal acquisition, firstly, the preamplifier converts the tiny eddy current signal collected by the probe into a standard stable voltage signal (0-10V); then, it is aggregated to the EL3102 analog input module, which converts the voltage signal into a digital signal; next, the EL3102 analog input module is connected to the EtherCAT bus via the bus coupler EK110, enabling it to communicate with TwinCAT; finally, a periodic data acquisition program is written using TwinCAT, the sampling frequency is set, and the vibration displacement signal is exported in CSV format.

[0033] Based on the above rotor imbalance state measurement system, this embodiment of the invention also provides a rotor imbalance state identification method based on shaft center motion trajectory. Figure 2 The flowchart of the identification method in this embodiment of the invention includes the following steps 1-4. It should be noted that three measurement sides are selected in this embodiment. In the following embodiments, the calculation and identification of the cross section located in the middle of the rotor is taken as an example. The calculation method of other cross sections is the same.

[0034] Step 1: Set up the measurement system and collect data according to Figure 1 As shown, a measurement system is arranged on the rotor system. The vibration displacement data of the rotor is acquired through the software system of this measurement system. The raw data acquired at the intermediate section is shown in the figure. Figure 3 As shown, the collected data is then used to further synthesize the shaft center motion trajectory. Specifically, for each measurement section, two sensors collect data in the X and Y directions respectively. Therefore, at each moment, the data collected by the two sensors will correspond to a coordinate point in the XY coordinate system. All the data collected in the same section can form multiple such coordinate points. The set of these discrete coordinate points is the shaft center motion trajectory of the measurement section. In this way, the trajectory constructed with two-dimensional data is closer to the actual running trajectory of the rotor shaft.

[0035] Step 2: Preprocess the axis trajectory During the data acquisition process, the position and angle of the acquisition device cannot be guaranteed to be absolutely fixed, and the distribution and amplitude range of the axis trajectory in different unbalanced states are also different in the two-dimensional plane. In order to eliminate the influence of this phenomenon on feature coordinate extraction, it is necessary to center the axis trajectory.

[0036] Before preprocessing, the raw data and Labeled according to time series Independent single samples and , As shown in equation (1).

[0037] (1) The centralized processing method is shown in formula (2).

[0038] (2) in, and This indicates data that has undergone centralized processing. and This represents the mean of a single sample, and n represents the total number of data points.

[0039] Furthermore, to ensure the consistency of data across all samples, the single sample that has undergone centralization... and The minimum-maximum normalization process is performed, as shown in formula (3).

[0040] (3) in, and This indicates data that has undergone normalization; after centering and normalization, the axis trajectory converges towards the origin and is within... Within the range.

[0041] Based on the data collected in this embodiment, and according to the above methods and formulas, the data is first centered, and the results are calculated. , Substituting into formula (2), the result after centering can be obtained.

[0042] Next, normalization is performed, and the result can be obtained through calculation. = -0.30119、 0.313455; -0.7934 Substituting 0.7072 into formula (3) yields the normalized result, such as... Figure 4 As shown.

[0043] Step 3: Calculate feature indicators The main process of this step includes three steps: dividing the curvature region, selecting feature coordinates, and calculating the eccentricity.

[0044] (1). Dividing curvature regions: Construct vectors from discrete points on the axis trajectory, calculate the included angle using the cosine theorem, and distinguish between large curvature points and general curvature points based on the set angle threshold β, thereby realizing the division of curvature regions of the trajectory shape.

[0045] From preprocessed single sample data and Synthetic axis trajectory ,in It can be seen that the axis trajectory is composed of a series of coordinate points; take any point... Take the second data point at intervals of m points. and the third data point The three points form vectors with the origin respectively. , and Thus, we obtain the vector. and The angle between the two vectors is calculated using the law of cosines. The calculation process is shown in formulas (4), (5) and (6).

[0046] (4) (5) (6) in, Representing vectors and The included angle, Representing vectors The mold will Convert to This is to facilitate comparisons using the same standard within the same coordinate system.

[0047] Set angle threshold By comparing angles, the type of point B can be determined. The point is identified as a point of high curvature, and otherwise as a point of general curvature. The same calculation is performed on all coordinate points in the axisymmetric trajectory R to divide the region into regions of high curvature and regions of general curvature.

[0048] In this embodiment, observation and The distribution patterns of their respective data sequences are analyzed, and then, by filtering the boundary extreme value coordinates of the axis trajectory, taking the coordinate set of two cycles as an example, it is concluded that the axis motion trajectory is formed by periodic cycles. These boundary extreme value coordinates are used as a reference, such as... Figure 4 As shown, we will further select specific coordinate data for verification.

[0049] We can initially assume that the regions roughly located near the two sets of coordinates in the second and fourth quadrants are regions of high curvature, while the remaining regions (mainly distributed in the first and third quadrants) belong to regions of general curvature. Next, we will perform calculations to verify this assumption.

[0050] First, we set the interval m=1000. Then, we select two sets of data that roughly cover the large curvature region and the general curvature region and substitute them into the above formula for calculation. The specific process is as follows: The first set of data was selected. , , The three points are: The 300th point: X: -0.16083, Y: -0.04936; 1300th point: X: -0.45378, Y: 0.365117; 2300th point: X: -0.208, Y: 0.43813.

[0051] We can obtain the result through coordinate vector calculation. , , , , Substituting the above data into the formula, we get... , , .

[0052] The second set of data was selected. , , The three points are: The 43rd point: X: -0.38823, Y: 0.415758; The 1043rd point has the following values: X: -0.14047, Y: 0.393794. The 2043rd point has the following values: X: 0.270157, Y: -0.04855.

[0053] We can obtain the result through coordinate vector calculation. , , , , Substituting the above data into the formula, we get... , , .

[0054] Finally, based on the trajectory image features, an angle threshold β = 60° is set for judgment. , Therefore, the region encompassed by the first set of data—that is, the region near the two sets of coordinates in the second quadrant—is a region of high curvature, while the region of general curvature is the region of moderate curvature, such as... Figure 5 As shown, since the elliptical trajectory is symmetric about the origin, the hypothesis can be verified.

[0055] (2). Selecting feature coordinates After region division, the identification of all coordinate points in the axisymmetric trajectory R is achieved. The axisymmetric trajectory graphic generally follows the principle of selecting feature coordinates at equal intervals L, but different intervals are applied for different curvature regions. The feature point selection principle of interval L is applied to regions with large curvature, while the feature point selection principle of interval multiples L is applied to regions with general curvature. This achieves adaptive selection of feature points based on the curvature of the region.

[0056] Typically, the interval L is determined based on the number of coordinate points n of the axis trajectory R. In this embodiment, the number of coordinate points n of the axis trajectory is 6000. To balance data volume and representation effect, the interval L is set to 5. Generally, the curvature region follows the principle of selecting feature points with an interval of 2L, i.e., an interval of 10. After completing the selection of feature point intervals, a series of feature coordinate sequences are obtained. M represents the number of feature coordinates.

[0057] Because the trajectory of the axis of motion is formed in a periodic cycle, the difference between the position of each coordinate point in each cycle and the coordinate point in the next cycle is negligible. Therefore, to reduce unnecessary repetition, only the coordinate points in the second cycle are selected. Based on the above method, it is known that feature points in regions of high curvature are selected at intervals of L=5, while those in regions of general curvature are selected at intervals of 2L (i.e., 10).

[0058] Depend on Figure 5 By selecting the median coordinates between the boundary extreme coordinates, the regions with large curvature are: 1090-1390, 1684-2041; the regions with general curvature are: 2041-2284, 1390-1684. Finally, the characteristic coordinate sequence is synthesized. That is, the final number of feature coordinates selected is 161, and the specific distribution is as follows: Figure 5 As shown.

[0059] (3). Calculate the eccentricity Calculating the eccentricity is an important basis for judging whether the rotor system is balanced. From the previous step, we can obtain... ,in Calculate the distance between each coordinate point and the origin, i.e., the eccentricity h, as shown in formula (7).

[0060] (7) All feature coordinates can be calculated using formula (7) to obtain a data sequence, namely the eccentricity sequence. This parameter directly reflects the magnitude of the unbalanced force in the rotor system; the larger the value, the more severe the unbalance in the rotor system.

[0061] In this embodiment, ,in Substitute into formula (7) to calculate the eccentricity h of each coordinate point, and finally obtain an eccentricity sequence H, as shown in the figure. Figure 6 As shown.

[0062] Step 4: Set the eccentricity threshold Finally, based on the characteristics of this rotor system and by referring to standards, a specific eccentricity threshold is set. When the eccentricity at a certain point of a certain measurement section exceeds the specified threshold, it can be determined that the measurement section is in an unbalanced state.

[0063] In this embodiment, considering the application background of the rotor system—a two-point support shaft system with a rotational speed of 1200 r / min—the eccentricity threshold is determined by referring to the standard GB / T9239.1-2006 / ISO1940-1:2003, "Mechanical Vibration—Constant State (Rigid) Rotor Balance Quality Requirements, Part 1: Specification and Inspection of Balance Tolerances". According to JB / T9752.3-2014, JB / T9601-2013, and ISO1940-1, Grade G is the core quantitative indicator of balance quality, and G6.3 is recommended for conventional rigid rotors. The allowable residual unbalance and eccentricity mentioned in the standards are numerically the same, only the units are defined differently. Therefore, the final eccentricity threshold is determined to be 0.02 mm.

[0064] Comparison Figure 6 From the eccentricity sequence calculated in the middle section, it is not difficult to find that the eccentricity measured at the middle section far exceeds the eccentricity threshold. Therefore, it can be determined that the section corresponding to the rotor system is in an unbalanced state.

[0065] The above provides a method for judging the imbalance of the measured cross section. Furthermore, since the measurement system of this invention measures three cross sections simultaneously, to avoid interference from local disturbances on the system imbalance, this invention further proposes a method for identifying the imbalance state of the entire rotor system. This method requires comprehensive calculation of the rotor system imbalance, as shown in the following formula: Rotor system imbalance = Σ(cross-sectional imbalance value) (Section weight); Hazard value = Single-section threshold Adjustment coefficient.

[0066] Among them, the cross-sectional unbalance value is the maximum value of the eccentricity calculated for that cross-section, while the single cross-section threshold is the threshold determined by the standard mentioned above; the rotor system unbalance is obtained by weighted summation of the unbalance values ​​of each cross-section.

[0067] Based on engineering experience, the rotor in the middle position has greater flexibility and deformation, resulting in a larger eccentricity than at the support points; therefore, it receives a higher weight. Conversely, the rotor at the support points has greater rigidity, smaller deformation, and a smaller eccentricity, thus receiving a lower weight. Furthermore, considering engineering safety margin criteria, a conservative adjustment coefficient needs to be set. Finally, the calculated rotor system imbalance is compared to the critical value. If it exceeds the critical value, the rotor system is determined to be unbalanced.

[0068] In this embodiment, the section weighting coefficients at both ends of the support are all taken as 0.2, and the section weighting coefficient at the middle of the rotor is taken as 0.6; the adjustment coefficient is taken as 0.8. After calculating the unbalance value of each section in the above manner and performing a weighted sum, it is compared with the danger value to comprehensively judge the unbalance state of the rotor system. If the calculated unbalance of the rotor system is greater than the danger value, the entire rotor system is considered to be in an unbalanced state. This method can avoid interference caused by local disturbances.

[0069] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A rotor imbalance state measurement system, characterized in that, The measurement system simultaneously measures multiple sections of the rotor along the radial direction. Two sensors are arranged in the plane of each measurement section, with the two sensors forming a 90° angle with the line connecting them to the shaft center. The sensors are not in contact with the rotor. The sensors are eddy current displacement sensors. Each sensor is connected to a preamplifier. The preamplifier converts the tiny eddy current signals collected by the sensors into voltage signals and summarizes them to the EL3102 analog input module. The EL3102 analog input module can convert voltage signals into digital signals. The EL3102 analog input module is connected to the EtherCAT bus and communicates with TwinCAT through the bus coupler EK110. TwinCAT is used to write periodic data acquisition programs, set the sampling frequency, and export the acquired vibration displacement signals in CSV format.

2. A method for identifying rotor imbalance, characterized in that, The identification method includes the following steps: (1) Set up the measurement system and collect data: Set up the measurement system as described in claim 1 on the rotor shaft system. For a two-point support shaft system, select measurement sections and arrange sensors on the outer side of the two-point support and the middle part of the rotor respectively. At the same time, collect the vibration displacement signals of the three measurement sections. Based on the collected data, synthesize the shaft center motion trajectory. (2) Preprocessing the axis center trajectory: The synthesized axis center trajectory is centered and normalized to make the axis center trajectory closer to the origin and in a position where Within the range; (3) Calculate characteristic indicators: This includes three steps: dividing the curvature region, selecting characteristic coordinates, and calculating the eccentricity. (3.1) Dividing the curvature region: Construct vectors by discrete points on the axis trajectory, calculate the included angle using the cosine theorem, and distinguish between large curvature points and general curvature points based on the set angle threshold β, thereby realizing the division of the curvature region of the trajectory shape; (3.2) Select feature coordinates, and extract key points using an adaptive interval strategy based on the curvature region division results. Dense sampling is performed in areas with large curvature, and sparse sampling is performed in areas with general curvature. Finally, a concise and representative feature point sequence is generated, which provides a basis for subsequent eccentricity calculation. (3.3) Calculate the eccentricity. Based on the extracted feature coordinate sequence, calculate the Euclidean distance from each point to the origin to generate the eccentricity sequence H, which is used to quantify the degree of geometric asymmetry and unbalance of the rotor system. (4) Set an eccentricity threshold and determine the rotor state: Combine the characteristics of the rotor system and find the standard to set a specific eccentricity threshold. When the eccentricity at a certain point of a certain measurement section exceeds the specified threshold, it can be determined that the measurement section is in an unbalanced state.

3. The rotor imbalance state identification method according to claim 2, characterized in that, In step (2), the specific steps for centering the axis trajectory are as follows: First, the original data is labeled according to the time series. and Divided into Independent single samples and , As shown in equation (1): (1) Centralized processing is shown in formula (2). (2) in, and This indicates data that has undergone centralized processing. and This represents the mean of a single sample.

4. The rotor imbalance state identification method according to claim 3, characterized in that, In step (2), the specific steps for normalizing the axis trajectory are as follows: To ensure the consistency of each sample data, the single sample after centering is normalized... and After performing min-max normalization, as shown in formula (3) (3) in, and This indicates data that has undergone normalization.

5. The rotor imbalance state identification method according to claim 4, characterized in that, Step (3.1) specifically involves: using the preprocessed single-sample data... and Synthetic axis trajectory ,in ; The axis trajectory consists of a series of coordinate points; take any point... Take the second data point at intervals of m points. and the third data point The three points form vectors with the origin respectively. , and Thus, we obtain the vector. and The angle between the two vectors can be calculated using the law of cosines. Specifically, it is shown in the following formulas (4), (5) and (6): (4) (5) (6) in, Representing vectors and The included angle, Representing vectors The model; Set angle threshold By comparing angles, the type of point B can be determined. By classifying point B as a point of high curvature and others as points of normal curvature, and performing the same calculation on all coordinate points in the axis trajectory R, ​​the regions of high curvature and normal curvature can be divided.

6. The rotor imbalance state identification method according to claim 2, characterized in that, Step (3.2) uses an adaptive interval strategy to extract key points. Specifically, the axis trajectory graphic generally follows the principle of selecting feature coordinates with equal intervals, but different intervals are implemented for different curvature regions. The feature point selection principle with interval L is implemented for large curvature regions, and the feature point selection principle with interval multiple L is implemented for general curvature regions. This is to achieve the selection of adaptive interval feature points based on the curvature of the region.

7. The rotor imbalance state identification method according to claim 6, characterized in that, The interval L is determined based on the number of coordinate points n on the axis trajectory R. When the number of coordinate points n on the axis trajectory is 6000, the interval L is determined to be 5. Generally, for curvature regions, the feature points are selected according to the principle of interval 2L, i.e., 10. After completing the selection of feature point intervals, a series of feature coordinate sequences are obtained. M represents the number of feature coordinates.

8. The rotor imbalance state identification method according to claim 2, characterized in that, The specific method for calculating the eccentricity in step (3.3) is as follows: the characteristic coordinate sequence can be obtained from step (3.2). ,in M is the number of characteristic coordinates. The distance between each coordinate point and the origin, i.e., the eccentricity h, is calculated as shown in formula (7): (7) All feature coordinates can be calculated using formula (7) to obtain a data sequence, namely the eccentricity sequence. The eccentricity directly reflects the magnitude of the unbalanced force in the rotor system; the larger the value, the more severe the unbalance in the rotor system.

9. The rotor imbalance state identification method according to claim 2, characterized in that, In step (4), the identification of the unbalanced state of the entire rotor system needs to be achieved through comprehensive calculation of the rotor system unbalance, as follows: Rotor system imbalance = Σ(cross-sectional imbalance value) (Section weight); Hazard value = Single-section threshold Adjustment factor; Among them, the unbalanced value of the cross section is the maximum value of the eccentricity calculated for the cross section. The cross section weight is allocated according to the deformation of the cross section location. The greater the deformation, the higher the weight. The sum of the weights of all cross sections is 1. The single-section threshold is the eccentricity threshold determined by the standard. Considering the engineering safety margin criterion, a conservative adjustment coefficient is set for this single-section threshold. Finally, the calculated rotor system imbalance is compared with the danger value. If it is determined that the imbalance exceeds the danger value, the rotor system is considered to be in an unbalanced state.

10. A rotor imbalance state identification method according to claim 9, characterized in that, For the shaft system with supports at both ends, the weighting coefficient of the three selected measurement sections is 0.2 for the sections at both ends and 0.6 for the section in the middle of the rotor; the adjustment coefficient is 0.8.

Images