A brush motor rotor detection method and system

Abstract

This application belongs to the field of motor testing technology and discloses a brushed motor rotor testing method and system. By acquiring the spatial position sampling data of the rotor shaft, the original surface contour signal and the original force detection signal in real time, and combining the determination of the instantaneous rotation axis, follow-up compensation processing and interference suppression processing, the influence of environmental factors and mechanical vibration on the test results can be effectively eliminated. Thus, the true geometric contour features and inherent centrifugal force feature signals of the rotor shaft can be accurately obtained, and the accurate determination of the centroid distribution data and the roundness of the rotor shaft of the brushed motor can be achieved, which significantly improves the testing accuracy and reliability.

Description

Technical Field

[0001] This application relates to the field of motor testing technology, and more specifically, to a method and system for testing the rotor of a brushed motor. Background Technology

[0002] In modern industrial production, the brushed motor rotor is a core component, and its quality directly affects the motor's operating performance and lifespan. Therefore, rigorous quality inspection before the rotor leaves the factory is an essential step. Typically, a specialized rotor inspection system is used to accomplish this task. This system can comprehensively evaluate the rotor's dynamic balance performance and dimensional accuracy. This inspection system usually includes a rotating support with a precision pressure sensor, a rotating drive mechanism, and a laser sensor. In actual operation, the two ends of the rotor shaft to be inspected are placed in two rotating support seats. When the rotating drive mechanism drives the rotor to rotate, the system can calculate the radial offset of the rotor's center of mass and the circumferential angle requiring counterweight by finely analyzing the fluctuation signals output by the pressure sensor. Simultaneously, the laser sensor performs a high-frequency scan of the rotating shaft's cross-section, acquiring a series of discrete point data, and then reconstructing the shaft's contour curve. By mathematically analyzing this contour curve, the system can determine the roundness of the shaft.

[0003] However, under long-term continuous production conditions, fine particulate matter such as metal dust, fiber debris, and lubricating oil mist, which permeate the factory environment, slowly and continuously accumulates on the surfaces of various components of the detection system. These particles gradually accumulate inside the V-grooves of the two V-shaped rotating supports, forming an uneven "pad" layer containing tiny abrasive grains. When the rotor shaft is placed on such a support and begins to rotate, the contact interface between the shaft and the support is no longer a smooth metal-to-metal contact, but rather a contact between the shaft and this uneven layer of particulate matter. This directly leads to additional, irregular, minute frictional forces between the shaft and the support during rotation, causing slight "vibration" or "micro-slippage" of the shaft within the V-grooves.

[0004] The additional high-frequency noise introduced by friction and vibration, mixed with the real signal of rotor imbalance, makes it difficult for the system to accurately separate pure centrifugal force fluctuations from the complex signal when analyzing pressure sensor data. This leads to a significant decrease in the measurement accuracy of the rotor's radial offset and circumferential angle, resulting in unstable and poor repeatability. A further problem is that this particulate-based "shield" causes continuous microscopic wear on the surface of the V-shaped support, resulting in subtle and irreversible changes to the original precise geometry of the V-groove. This geometric change means that the rotor's axis of rotation no longer remains stably in the ideal position as initially; the shaft may exhibit slight "wobbling" or "drifting," and its center of rotation is no longer fixed. This decrease in support stability not only further exacerbates the noise problem in the pressure sensor signal, making the measurement of imbalance more difficult and inaccurate, but also severely affects the geometric measurements of the laser sensor. The laser sensor measures the shaft profile based on the assumption that the center of rotation of the shaft is fixed. Once this premise is broken, even if the laser sensor is working properly, it will be unable to obtain a true and stable cross-sectional profile of the shaft, because the "section" it scans is constantly changing in minute ways.

[0005] To address the aforementioned issues, existing technologies urgently need improvement. Summary of the Invention

[0006] This invention discloses a brushed motor rotor detection method and system, which aims to solve the problem that in the existing brushed motor rotor detection system, the wear of the support seat caused by the deposition of environmental particulate matter during long-term operation leads to unstable rotation of the rotor shaft, severe interference with the measurement signal, and ultimately a decrease in the accuracy of rotor centroid distribution and roundness measurement.

[0007] In a first aspect, the present invention provides a method for detecting the rotor of a brushed motor, the method comprising the steps of: A1. With the brushed motor rotor shaft supported by two rotating supports at both ends, acquire spatial position sampling data, original surface contour signals, and original force detection signals of the brushed motor rotor during rotation; the spatial position sampling data includes the spatial coordinates of multiple surface measuring points of the shaft collected from different axial and circumferential positions; the original surface contour signals include the spatial coordinates of circumferential measuring points of the shaft collected from preset detection positions; the original force detection signals include the pressure of the shaft on the two rotating supports; A2. Based on the spatial position sampling data, determine the instantaneous rotation axis of the rotating shaft at each sampling moment; A3. Using the instantaneous rotation axis as a reference, perform follow-up compensation processing on the original surface contour signal to obtain the true geometric contour features of the rotating shaft; A4. Extract the motion characteristics of the instantaneous rotation axis, and use the motion characteristics to perform interference suppression processing on the original force detection signal to obtain the inherent centrifugal force characteristic signal of the rotating shaft; A5. Based on the actual geometric contour features and the inherent centrifugal force feature signals, determine the centroid distribution data of the brushed motor rotor and the roundness of the shaft.

[0008] Preferably, step A1 includes: A101. By deploying at least three non-contact displacement sensors at different circumferential and axial positions of the rotating shaft, the spatial coordinates of the surface measuring points of the rotating shaft are collected in real time as the spatial position sampling data; A102. The cross-sectional profile of the rotating shaft is scanned from a preset detection position using a laser sensor to obtain the spatial coordinates of multiple circumferential measuring points, which serve as the original surface profile signal. A103. The pressure of the rotating shaft on the two rotating support seats is detected by a pressure sensor and used as the original force detection signal; A104. Perform time alignment processing on the spatial location sampling data, the original surface contour signal, and the original force detection signal.

[0009] Preferably, step A2 includes: A201. For each sampling time, perform linear fitting on the spatial coordinates of the surface measuring points corresponding to that sampling time in the spatial location sampling data to construct the instantaneous rotation axis of the rotating shaft at that sampling time.

[0010] Preferably, step A2 further includes: A202. Obtain the nominal radius of the shaft; A203. Perform inverse geometric fitting on the original surface contour signal based on the nominal radius to determine the reference axis; A204. Compare the instantaneous rotation axis with the reference axis to identify systematic measurement drift in the spatial position sampling data; A205. Dynamically compensate the instantaneous rotation axis based on the systematic measurement drift.

[0011] Preferably, step A3 includes: A301. Calculate the radial distance from each circumferential measuring point in the original surface profile signal to the corresponding instantaneous rotation axis; A302. Calculate the local rate of change of the radial distance along the circumferential direction of the rotating shaft; A303. A dynamic threshold is set according to the nominal geometric parameters of the rotating shaft, and the circumferential measuring points whose local rate of change exceeds the dynamic threshold are identified as abnormal reflection points; the nominal geometric parameters include the nominal radius and roundness tolerance; A304. The radial distance of the abnormal reflection point is removed or interpolated for repair, and the radial distance of each circumferential measuring point after processing is used as the true geometric contour feature.

[0012] Preferably, step A4 includes: A401. Obtain the instantaneous displacement information of the instantaneous rotation axis relative to the preset ideal rotation axis, as the motion feature; the instantaneous displacement information includes radial offset and circumferential offset angle; A402. Identify and remove the non-periodic vibration components related to the instantaneous displacement information from the original force detection signal to extract the inherent centrifugal force characteristic signal of the shaft.

[0013] Preferably, step A402 includes: Using the instantaneous displacement information as a reference input and the original force detection signal as the desired output, the mechanical vibration noise component in the original force detection signal is estimated by an adaptive filtering algorithm and used as the non-periodic vibration component. The inherent centrifugal force characteristic signal is extracted by subtracting the mechanical vibration noise component from the original force detection signal.

[0014] Preferably, before step A401, the method further includes: A400a. Obtain the environmental vibration reference signal on the rotary support; A400b. Calculate the cross-power spectral density of the environmental vibration reference signal and the original force detection signal; A400c. Identify the common-mode interference component introduced by the external environment in the original force detection signal based on the cross-power spectral density; A400d. Filter out the common-mode interference component from the original force detection signal.

[0015] Preferably, step A5 includes: A501. Perform statistical analysis or geometric fitting on the actual geometric contour features to determine the roundness of the rotating shaft; A502. Perform amplitude and phase analysis on the inherent centrifugal force characteristic signal to determine the centroid distribution data; the centroid distribution data includes the radial offset and circumferential angle of the centroid of the brushed motor rotor.

[0016] Secondly, this application provides a brushed motor rotor detection system, including a host computer, two rotating support seats, a rotating drive device, at least three non-contact displacement sensors, a laser sensor, and two pressure sensors; the rotating drive device, the non-contact displacement sensors, the laser sensor, and the pressure sensor are all electrically connected to the host computer; The two rotating support seats are respectively used to support the two ends of the rotating shaft of the brushed motor rotor; The rotary drive device is used to drive the brushed motor rotor to rotate. At least three of the non-contact displacement sensors are distributed along different circumferential and axial positions of the rotating shaft and are used to collect the spatial coordinates of surface measuring points of the rotating shaft; The laser sensor is used to scan the cross-sectional profile of the shaft when the brushed motor rotor rotates, so as to obtain the spatial coordinates of multiple circumferential measuring points. The two pressure sensors are respectively installed at the two rotating support seats and are used to detect the pressure of the rotating shaft on the two rotating support seats; The host computer is used to execute the steps in the brushed motor rotor detection method described above.

[0017] Beneficial effects: The brushed motor rotor detection method and system provided in this application can effectively eliminate the influence of environmental factors and mechanical vibration on the detection results by acquiring the spatial position sampling data of the rotor shaft, the original surface contour signal and the original force detection signal in real time, and combining the determination of the instantaneous rotation axis, follow-up compensation processing and interference suppression processing. This allows for the accurate acquisition of the true geometric contour features and inherent centrifugal force feature signals of the rotor shaft, and ultimately achieves accurate determination of the centroid distribution data of the brushed motor rotor and the roundness of the rotor shaft, significantly improving the detection accuracy and reliability. Attached Figure Description

[0018] Figure 1 A flowchart of a brushed motor rotor detection method provided in this application.

[0019] Figure 2 This is a schematic diagram of a brushed motor rotor detection system provided in this application.

[0020] Labeling Explanation: 1. Host Computer; 2. Rotary Support Base; 3. Rotary Drive Device; 4. Non-contact Displacement Sensor; 5. Laser Sensor; 6. Pressure Sensor; 90. Brushed Motor Rotor; 91. Rotating Shaft. Detailed Implementation

[0021] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this application, and not all embodiments. The components of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0022] It should be noted that similar reference numerals and letters in the following figures indicate similar items; therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures. Furthermore, in the description of this application, terms such as "first," "second," etc., are used only to distinguish descriptions and should not be construed as indicating or implying relative importance.

[0023] Reference Figure 1 The present invention proposes a method for detecting the rotor of a brushed motor, the method comprising the following steps: A1. With the brushed motor rotor shaft supported by two rotating supports at both ends, acquire spatial position sampling data, original surface contour signals, and original force detection signals of the brushed motor rotor during rotation; the spatial position sampling data includes the spatial coordinates of multiple surface measuring points of the shaft collected from different axial and circumferential positions; the original surface contour signals include the spatial coordinates of circumferential measuring points of the shaft collected from preset detection positions; the original force detection signals include the pressure of the shaft on the two rotating supports; A2. Based on the spatial position sampling data, determine the instantaneous rotation axis of the rotating shaft at each sampling moment; A3. Using the instantaneous rotation axis as a reference, perform follow-up compensation processing on the original surface contour signal to obtain the true geometric contour features of the rotating shaft; A4. Extract the motion characteristics of the instantaneous rotation axis, and use the motion characteristics to perform interference suppression processing on the original force detection signal to obtain the inherent centrifugal force characteristic signal of the rotating shaft; A5. Based on the actual geometric contour features and the inherent centrifugal force feature signals, determine the centroid distribution data of the brushed motor rotor and the roundness of the shaft.

[0024] This application acquires the spatial position sampling data, original surface contour signal, and original force detection signal of the rotating shaft in real time. Combined with the determination of the instantaneous rotation axis, follow-up compensation processing, and interference suppression processing, it can effectively eliminate the influence of environmental factors and mechanical vibration on the detection results. This allows for the accurate acquisition of the true geometric contour features and inherent centrifugal force feature signals of the rotating shaft, ultimately achieving accurate determination of the centroid distribution data of the brushed motor rotor and the roundness of the rotating shaft, significantly improving detection accuracy and reliability.

[0025] Spatial position sampling data refers to the spatial coordinate data collected from different positions on the surface of the shaft during its rotation, which is used to describe the instantaneous spatial attitude of the shaft.

[0026] The original surface profile signal refers to the original profile data obtained by scanning the cross-section of the shaft with a sensor, which reflects the surface shape of the shaft.

[0027] The original force detection signal refers to the pressure signal generated by the shaft on the support during rotation, which reflects the dynamic balance characteristics of the shaft.

[0028] The instantaneous axis of rotation refers to the actual center line of rotation of the shaft at a specific moment, which may be slightly offset due to various factors.

[0029] True geometric profile features refer to profile data that, after compensation processing, can accurately reflect the actual geometric shape of the rotating shaft, eliminating errors caused by instability of the rotation axis.

[0030] The inherent centrifugal force characteristic signal refers to the pure centrifugal force signal caused by the rotor's own imbalance after eliminating external interference and mechanical vibration.

[0031] Center of mass distribution data describes the position of the rotor's center of mass, including radial offset and circumferential angle, and is a key basis for dynamic balancing correction.

[0032] Roundness is a measure of how close the cross-section of a shaft is to an ideal circle, and it is an important indicator for evaluating the geometric accuracy of a shaft.

[0033] The implementation environment of this application typically includes a detection platform equipped with a rotary drive, two rotary support seats for supporting the rotating shaft, a non-contact displacement sensor for acquiring spatial position data, a laser sensor for scanning surface contours, and a pressure sensor for detecting pressure. Data from all sensors can be received and processed by a host computer or data processing unit.

[0034] Specifically, in step A1, it is necessary to acquire the spatial position sampling data, the original surface contour signal, and the original force detection signal of the brushed motor rotor during rotation.

[0035] Spatial position sampling data can be acquired in various ways. For example, multiple contact displacement sensors can be manually placed at different axial and circumferential positions on the shaft. As the rotor rotates, these sensors periodically contact the shaft surface and record the spatial coordinates of the contact points. Another approach is to use multiple optical sensors, such as photoelectric switches or CCD cameras, to identify specific marker points on the shaft surface using image processing techniques as the shaft rotates, and calculate the spatial coordinates of these marker points.

[0036] For the original surface contour signal, a contact probe can be used for acquisition. For example, a high-precision mechanical probe is fixed at a preset detection position. When the shaft rotates, the probe will scan along the circumference of the shaft and record the spatial coordinates of the contact point between the probe tip and the surface of the shaft, thus forming the original surface contour signal.

[0037] The raw force detection signal can be obtained by installing a piezoelectric sensor under the rotating support. When the shaft applies pressure to the rotating support, the piezoelectric sensor generates an electric charge. By measuring the changes in these charges, the pressure signal of the shaft on the support can be obtained.

[0038] In step A2, it is necessary to determine the instantaneous rotation axis of the shaft at each sampling moment based on the spatial location sampling data.

[0039] One implementation involves inputting the spatial coordinates of all surface measurement points collected at each sampling time into a least-squares fitting algorithm. This algorithm calculates an optimal fitting line that minimizes the sum of the squared distances from all surface measurement points to this line. This fitted line is then determined as the instantaneous axis of rotation at that sampling time.

[0040] Another approach is to use Principal Component Analysis (PCA). For the spatial coordinates of the surface measurement points at each sampling time, PCA is performed to extract the principal direction of the data point distribution. The straight line represented by this principal direction is the instantaneous rotation axis at that sampling time.

[0041] In step A3, the instantaneous rotation axis is used as a reference to perform follow-up compensation processing on the original surface contour signal in order to obtain the true geometric contour features of the rotating shaft.

[0042] One approach is to first calculate the radial distance from each circumferential measuring point in the original surface profile signal to the instantaneous axis of rotation at the corresponding sampling time. Since the instantaneous axis of rotation is dynamically changing, this calculation method can eliminate the influence of shaft oscillation on the radial distance measurement. Then, these radial distances are used as the compensated profile data, which represents the true geometric profile features of the shaft.

[0043] Another approach is to first project all the circumferential measurement points in the original surface profile signal onto a plane perpendicular to the instantaneous rotation axis. Then, within each projection plane, the distance from each projection point to the intersection of the instantaneous rotation axis and the plane is calculated. These distances constitute the compensated profile data, which is the true geometric profile feature of the rotation axis.

[0044] In step A4, it is necessary to extract the motion characteristics of the instantaneous rotation axis and use these motion characteristics to perform interference suppression processing on the original force detection signal in order to obtain the inherent centrifugal force characteristic signal of the rotating shaft.

[0045] One approach is to first calculate the displacement of the instantaneous rotation axis relative to a pre-defined ideal rotation axis (e.g., the geometric center axis of the detection system) at each sampling moment. This displacement can be decomposed into radial offset and circumferential offset angle, which constitute the motion characteristics of the instantaneous rotation axis. Then, these motion characteristics are used as reference signals and correlated with the original force detection signal. Using a digital filter, such as a Kalman filter or Wiener filter, components related to the instantaneous rotation axis motion are filtered out from the original force detection signal based on the correlation analysis results, thereby suppressing interference and obtaining the inherent centrifugal force characteristic signal.

[0046] Another approach is to establish a mathematical model based on the instantaneous rotational axis motion characteristics. This model can predict the interference components in the force detection signal caused by the axis motion. Then, by subtracting the predicted interference components from the original force detection signal, the inherent centrifugal force characteristic signal can be obtained.

[0047] In step A5, it is necessary to determine the centroid distribution data of the brushed motor rotor and the roundness of the shaft based on the actual geometric profile features and the inherent centrifugal force feature signals.

[0048] To determine the roundness of the shaft, a geometric fit can be performed on the actual geometric contour features. For example, all radial distance data from the actual geometric contour features can be input into a least-squares circle fitting algorithm to calculate the radius and center of the best-fit circle. Then, based on the deviation between the parameters of the fitted circle and the original radial distance data, the roundness of the shaft can be calculated.

[0049] For centroid distribution data, a Fourier transform can be performed on the inherent centrifugal force characteristic signal to analyze its amplitude and phase. The amplitude of the fundamental frequency component of the inherent centrifugal force characteristic signal is proportional to the rotor's unbalance, while its phase reflects the circumferential angle of the unbalance. Through these analyses, the rotor's centroid radial offset and circumferential angle, i.e., the centroid distribution data, can be determined.

[0050] Compared with existing technologies, the core innovation of this application lies in the introduction of dynamic tracking and utilization of the instantaneous rotation axis of the shaft. Existing technologies typically measure the shaft profile based on the assumption that the shaft's rotation center remains fixed. Once this premise is broken, even if the laser sensor is functioning normally, it cannot obtain the true and stable cross-sectional profile of the shaft. This application, through steps A1 and A2, acquires the spatial position sampling data of the shaft in real time and determines its instantaneous rotation axis, thereby dynamically capturing the actual rotational attitude of the shaft.

[0051] Furthermore, in step A3, this application uses the instantaneous rotation axis as a reference to perform follow-up compensation processing on the original surface contour signal. This means that even if the shaft experiences slight wobbling or drift during rotation, this application can correct it based on its instantaneous rotation axis, thereby obtaining the true geometric contour features of the shaft. This significantly improves the accuracy of shaft roundness measurement and overcomes the problem of inaccurate geometric measurement caused by decreased support stability in the prior art.

[0052] Furthermore, in step A4 of this application, the motion characteristics of the instantaneous rotation axis are extracted, and these characteristics are used to perform interference suppression processing on the original force detection signal. In the prior art, additional high-frequency noise introduced by friction and vibration is mixed with the real signal of rotor imbalance, which significantly reduces the measurement accuracy of the radial offset of the centroid and the circumferential angle. By identifying and removing the non-periodic vibration components related to the motion of the instantaneous rotation axis, this application can more accurately extract the inherent centrifugal force characteristic signal of the shaft, thereby greatly improving the stability and repeatability of the rotor centroid distribution data measurement.

[0053] In summary, this application effectively solves the problem of decreased detection accuracy caused by environmental interference and mechanical vibration in the prior art by introducing technologies such as dynamic tracking of instantaneous rotation axis, follow-up compensation and interference suppression. It achieves high-precision and high-reliability detection of the centroid distribution data and shaft roundness of brushed motor rotors, which has significant technological progress and practical value.

[0054] In some implementations, step A1 includes: A101. By deploying at least three non-contact displacement sensors at different circumferential and axial positions of the rotating shaft, the spatial coordinates of the surface measuring points of the rotating shaft are collected in real time as the spatial position sampling data; A102. The cross-sectional profile of the rotating shaft is scanned from a preset detection position using a laser sensor to obtain the spatial coordinates of multiple circumferential measuring points, which serve as the original surface profile signal. A103. The pressure of the rotating shaft on the two rotating support seats is detected by a pressure sensor and used as the original force detection signal; A104. Perform time alignment processing on the spatial location sampling data, the original surface contour signal, and the original force detection signal.

[0055] Specifically, in step A101, at least three non-contact displacement sensors are deployed at different circumferential and axial positions of the rotating shaft. These sensors can be non-contact, such as eddy current sensors, capacitive sensors, or optical sensors, which can avoid causing wear or interference to the surface of the rotating shaft. By deploying them at different positions, the spatial attitude information of the rotating shaft during rotation can be comprehensively captured. The real-time acquired spatial coordinates of the surface measurement points constitute spatial position sampling data, providing a basis for subsequent determination of the instantaneous axis of rotation.

[0056] In step A102, a laser sensor is used to scan the cross-sectional profile of the shaft from a preset detection position. The laser sensor offers advantages such as high precision and non-contact operation, enabling it to quickly and accurately acquire the spatial coordinates of multiple circumferential measuring points on a cross-section of the shaft. These spatial coordinates of the circumferential measuring points collectively constitute the original surface profile signal, used to characterize the initial geometry of the shaft.

[0057] In practical applications, in step A103, pressure sensors are installed at two rotating support seats to detect the pressure exerted by the shaft on these seats. These pressure sensors can monitor the dynamic force generated by the shaft on the support structure during rotation in real time, thereby acquiring the raw force detection signal. This signal is a key input for analyzing the centrifugal force characteristics of the rotor.

[0058] Furthermore, in step A104, the acquired spatial location sampling data, original surface profile signals, and original force detection signals are time-aligned. Since different sensors may have different sampling frequencies and response times, time alignment ensures the consistency of all data in the time dimension, providing an accurate data foundation for subsequent synchronous analysis and compensation processing. For example, by recording a precise timestamp for each data point and using an interpolation algorithm to resample the original surface profile signals and original force detection signals with different sampling frequencies onto the same time axis as the spatial location sampling data, synchronous alignment of all data is achieved.

[0059] This application's solution ensures the accuracy and synchronization of all raw data required for the rotation of a brushed motor rotor by employing specific types of sensors and data processing steps. Specifically, the deployment of non-contact displacement sensors at different circumferential and axial positions of the shaft and their real-time acquisition capabilities enable precise capture of the shaft's instantaneous spatial attitude, providing a reliable basis for subsequent determination of the instantaneous rotation axis. The high-precision scanning of the laser sensor ensures the integrity and accuracy of the shaft's surface contour information, which is crucial for extracting true geometric contour features. Simultaneously, the real-time monitoring of the force on the shaft by the pressure sensor lays the foundation for subsequent analysis of centrifugal force characteristic signals. Furthermore, time alignment processing eliminates temporal deviations in data from different sensors, ensuring that all data is analyzed under the same time reference, thereby avoiding errors caused by data asynchrony and improving the accuracy of subsequent data processing and analysis.

[0060] In some implementations, step A2 includes: A201. For each sampling time, perform linear fitting on the spatial coordinates of the surface measuring points corresponding to that sampling time in the spatial location sampling data to construct the instantaneous rotation axis of the rotating shaft at that sampling time.

[0061] Specifically, during the rotation of the brushed motor rotor, at least three non-contact displacement sensors deployed at different circumferential and axial positions on the shaft collect the spatial coordinates of surface measuring points in real time, thus forming spatial position sampling data. These surface measuring points have unique spatial coordinates at each sampling moment. Linear fitting refers to using the least squares method or other mathematical fitting algorithms to find an optimal straight line in three-dimensional space that best represents the distribution trend of these surface measuring points. In this way, the influence of measurement errors at individual measuring points can be effectively eliminated, thereby accurately constructing the instantaneous rotation axis of the shaft at the current sampling moment. This instantaneous rotation axis reflects the dynamic centerline of the shaft during actual rotation, rather than a fixed geometric axis.

[0062] The proposed solution accurately captures the dynamic attitude of the rotating shaft during rotation by performing linear fitting on the spatial position sampling data acquired at each sampling moment. Since the shaft may experience slight yaw, jumps, or bends during high-speed rotation, its actual axis of rotation is not fixed. By performing real-time linear fitting on the spatial coordinates of multiple surface measurement points, the actual rotation centerline of the shaft at each instant can be dynamically determined. This method effectively filters out measurement fluctuations caused by sensor noise or local surface irregularities, thus obtaining a stable and accurate instantaneous axis of rotation. This instantaneous axis of rotation serves as a reference for subsequent data processing and is crucial for accurately evaluating the rotor's geometric profile and stress characteristics.

[0063] In some embodiments described above, the instantaneous rotation axis of the shaft at each sampling moment can be constructed by linearly fitting the spatial coordinates of surface measuring points corresponding to each sampling moment in the spatial position sampling data. However, in actual measurement, due to factors such as installation errors of the non-contact displacement sensor, changes in ambient temperature, or performance drift caused by long-term use, the spatial position sampling data may exhibit systematic measurement drift. If the above problems are not addressed, the determined instantaneous rotation axis may not accurately reflect the actual rotation state of the shaft, thereby affecting the accuracy of subsequent extraction of true geometric contour features and inherent centrifugal force feature signals.

[0064] Therefore, in some preferred embodiments, step A2 further includes: A202. Obtain the nominal radius of the shaft; A203. Perform inverse geometric fitting on the original surface contour signal based on the nominal radius to determine the reference axis; A204. Compare the instantaneous rotation axis with the reference axis to identify systematic measurement drift in the spatial position sampling data; A205. Dynamically compensate the instantaneous rotation axis based on the systematic measurement drift.

[0065] Specifically, in step A202, the nominal radius of the shaft refers to the theoretical radius value specified during the design or manufacturing of the shaft, which can be obtained from design drawings, product specifications, or through prior high-precision measurements. This nominal radius, as a known and stable geometric parameter, provides the basis for subsequent determination of the reference axis.

[0066] In step A203, inverse geometric fitting is performed on the original surface profile signal based on the nominal radius. The purpose is to deduce a theoretically ideal axis of rotation, i.e., a reference axis, from the actually acquired surface profile data. The original surface profile signal is obtained by scanning the cross-sectional profile of the shaft using a laser sensor, reflecting the actual geometry of the shaft. Through inverse geometric fitting, such as using the least squares method or the RANSAC algorithm, and combined with the nominal radius, the circumferential measurement points in the original surface profile signal are fitted into a circle. This allows the determination of the circle's center, thereby constructing a reference axis (i.e., a straight line perpendicularly passing through the center of the circle) representing the ideal position of the shaft. This reference axis is independent of the measurements from the non-contact displacement sensor and can therefore serve as a benchmark for evaluating the accuracy of the instantaneous axis of rotation.

[0067] In practical applications, step A204 compares the instantaneous rotation axis obtained through linear fitting with the reference axis obtained through inverse geometric fitting. This comparison may include calculating the radial offset, angular deviation, etc., between the two axes. This comparison can identify systematic measurement drift in the spatial position sampling data. For example, if the instantaneous rotation axis always has a fixed offset or tilt angle relative to the reference axis, it can be determined that systematic measurement drift exists.

[0068] Further, in step A205, dynamic compensation is performed on the instantaneous rotation axis based on the identified systematic measurement drift. Dynamic compensation can be understood as real-time correction of the instantaneous rotation axis to eliminate or reduce the influence of systematic measurement drift. For example, if a fixed offset in a certain direction is identified in the instantaneous rotation axis, an equal and opposite correction amount can be applied in that direction; if an angular deviation exists, a corresponding rotational transformation can be performed. The purpose is to ensure that the compensated instantaneous rotation axis can more accurately reflect the actual rotation center and direction of the shaft.

[0069] This application's solution constructs an independent reference axis by introducing a nominal radius and the original surface profile signal, thus providing a calibration benchmark for the accuracy of the instantaneous rotation axis. Specifically, since the original surface profile signal is acquired by a laser sensor, its measurement principle differs from that of a non-contact displacement sensor; therefore, its data is less affected by the systematic measurement drift of the non-contact displacement sensor. By comparing the instantaneous rotation axis with the reference axis determined based on the original surface profile signal and the nominal radius, the systematic measurement drift introduced by the non-contact displacement sensor can be effectively quantified and identified. Subsequently, the instantaneous rotation axis is dynamically compensated based on the identified drift, thereby correcting the calculated result of the instantaneous rotation axis and making it closer to the true rotation axis of the shaft.

[0070] In some implementations, step A3 includes: A301. Calculate the radial distance from each circumferential measuring point in the original surface profile signal to the corresponding instantaneous rotation axis; A302. Calculate the local rate of change of the radial distance along the circumferential direction of the rotating shaft; A303. A dynamic threshold is set according to the nominal geometric parameters of the rotating shaft, and the circumferential measuring points whose local rate of change exceeds the dynamic threshold are identified as abnormal reflection points; the nominal geometric parameters include the nominal radius and roundness tolerance; A304. The radial distance of the abnormal reflection point is removed or interpolated for repair, and the radial distance of each circumferential measuring point after processing is used as the true geometric contour feature.

[0071] Specifically, in step A301, the original surface profile signal typically consists of the spatial coordinates of a series of circumferential measuring points, which represent the surface shape of the shaft on a certain cross-section. The instantaneous rotation axis provides the actual center and direction of rotation of this cross-section at the current moment. Calculating the radial distance from each circumferential measuring point to the corresponding instantaneous rotation axis involves projecting the spatial coordinates of each measuring point onto a plane perpendicular to the instantaneous rotation axis and calculating its perpendicular distance to that axis. This radial distance reflects the actual size of the shaft surface relative to its instantaneous center of rotation.

[0072] In step A302, the local rate of change of radial distance along the circumference of the shaft can be understood as the degree of difference in radial distance between adjacent measuring points. This rate of change can be calculated by differentiating the radial distance sequence, calculating the gradient, or local fitting. The purpose is to identify sharp abrupt changes or anomalies that may exist in the shaft's surface profile. These abrupt changes are often not true geometric features of the shaft, but rather caused by measurement noise or local reflection anomalies.

[0073] In practical applications, in step A303, the nominal geometric parameters refer to the ideal dimensions and shape parameters specified during the design of the shaft, such as the nominal radius and roundness tolerance. Based on these parameters, a dynamic threshold can be set. This dynamic threshold is determined according to the normal geometric variation range of the shaft and the accuracy of the measurement system; for example, it can be set as a multiple of the nominal roundness tolerance. Alternatively, a preset dynamic threshold lookup table (which records dynamic thresholds corresponding to different nominal radii and roundness tolerances) can be consulted based on the nominal geometric parameters to obtain the corresponding dynamic threshold. When the local rate of change exceeds this dynamic threshold, it indicates that the radial distance at that measuring point is significantly abnormal compared to surrounding measuring points, and therefore it is identified as an abnormal reflection point.

[0074] Further, in step A304, the radial distance of the identified abnormal reflection points is processed, which can be done by removal or interpolation repair. Removal refers to directly removing the abnormal points from the dataset. Interpolation repair involves using the normal measurement point data around the abnormal point, through linear interpolation, spline interpolation, or other data fitting methods, to estimate and replace the radial distance of the abnormal point. The purpose is to eliminate the interference of abnormal points on the calculation of the true geometric contour features, thereby obtaining smoother and more accurate true geometric contour features of the rotation axis.

[0075] This application's solution effectively solves the measurement noise and outlier interference problems that may exist in traditional servo compensation methods by introducing refined processing of the original surface contour signal. Specifically, firstly, the radial distance from each circumferential measuring point to the instantaneous rotation axis is calculated, converting the original spatial coordinates into more intuitive radial deviation information. Subsequently, by calculating the local rate of change of the radial distance, local abrupt changes caused by measurement errors or surface defects can be sensitively captured. Since these local abrupt changes often correspond to non-realistic geometric features, setting a dynamic threshold based on the nominal geometric parameters of the rotation axis can distinguish these outliers from the normal contour data. Finally, by removing or interpolating these outliers, noise and interference are effectively filtered out, making the final obtained real geometric contour features more accurate and reliable, and avoiding contour distortion caused by problems with the quality of the original data.

[0076] In some implementations, step A4 includes: A401. Obtain the instantaneous displacement information of the instantaneous rotation axis relative to the preset ideal rotation axis, as the motion feature; the instantaneous displacement information includes radial offset and circumferential offset angle; A402. Identify and remove the non-periodic vibration components related to the instantaneous displacement information from the original force detection signal to extract the inherent centrifugal force characteristic signal of the shaft.

[0077] Specifically, in step A401, the motion characteristic is defined as the instantaneous displacement information of the instantaneous rotation axis relative to a preset ideal rotation axis. The preset ideal rotation axis can be understood as the theoretical rotation center line under conditions of complete rotor balance, without any external interference or internal defects; it is typically determined based on the equipment's mechanical design parameters or calibration data. The instantaneous displacement information quantifies the deviation between the actual rotation axis and this ideal axis, specifically manifested as radial offset and circumferential offset angle. The radial offset refers to the distance the instantaneous rotation axis deviates from the ideal rotation axis in the radial direction, while the circumferential offset angle describes the specific position or phase of this radial offset in the circumferential direction. This displacement information comprehensively reflects the dynamic motion characteristics of the shaft during rotation caused by factors such as imbalance, deformation, or external interference.

[0078] In step A402, by utilizing the aforementioned instantaneous displacement information, non-periodic vibration components related to this displacement information can be identified and removed from the original force detection signal. The original force detection signal typically contains periodic components caused by the rotor's inherent centrifugal force, as well as non-periodic vibration components caused by the dynamic motion of the shaft (i.e., instantaneous displacement information), such as mechanical vibration and impact. By separating the non-periodic vibration components directly related to the instantaneous displacement information from the original force detection signal, these interferences can be effectively suppressed, thereby more accurately extracting the inherent centrifugal force characteristic signal of the shaft.

[0079] This application's solution uses the instantaneous displacement information of the rotation axis as a motion feature to suppress interference in the original force detection signal, effectively solving the problem of incomplete interference suppression in traditional methods. Specifically, during the rotation of the shaft, the instantaneous displacement of its rotation axis relative to the ideal axis (i.e., radial offset and circumferential offset angle) directly causes the support to bear additional non-periodic forces closely related to the shaft's motion state. These forces manifest as non-periodic vibration components in the original force detection signal. They are not caused by the rotor's own mass imbalance (i.e., inherent centrifugal force), but rather by the dynamic motion of the shaft. By accurately acquiring this instantaneous displacement information and using it as a reference, a model can be established to predict and identify the non-periodic vibration components caused by the shaft's motion in the original force detection signal. Once these interference components are identified, they can be removed from the original force detection signal, allowing the remaining signal to more purely reflect the inherent centrifugal force characteristics of the rotor. This motion feature-based interference suppression method makes the extraction of inherent centrifugal force characteristic signals more accurate and avoids measurement errors introduced by the dynamic motion of the shaft.

[0080] In practice, non-periodic vibration components may contain complex mechanical vibration noise, the amplitude and frequency of which may dynamically change over time, making it difficult for simple identification and stripping methods to accurately separate them from the original force detection signal. If this problem is not addressed, noise may remain in the extracted inherent centrifugal force feature signal, affecting the accuracy of subsequent centrifugal center distribution data and shaft roundness determination. To address this, this application further proposes using an adaptive filtering algorithm to accurately estimate and remove these mechanical vibration noise components, thereby obtaining the inherent centrifugal force feature signal of the shaft more accurately. Specifically, step A402 includes: Using the instantaneous displacement information as a reference input and the original force detection signal as the desired output, the mechanical vibration noise component in the original force detection signal is estimated by an adaptive filtering algorithm and used as the non-periodic vibration component. The inherent centrifugal force characteristic signal is extracted by subtracting the mechanical vibration noise component from the original force detection signal.

[0081] The instantaneous displacement information refers to the instantaneous displacement information of the rotation axis relative to the preset ideal rotation axis obtained in step A401, which includes radial offset and circumferential offset angle. The original force detection signal refers to the pressure of the shaft on the two rotating supports obtained in step A1. The adaptive filtering algorithm is a digital filter that can automatically adjust its parameters according to the statistical characteristics of the input signal. Its purpose is to extract the desired signal from the observed signal in the presence of noise and interference. Here, the algorithm is used to estimate the mechanical vibration noise component related to the instantaneous displacement information. The mechanical vibration noise component can be understood as the non-periodic vibration that is highly correlated with the instantaneous displacement information, caused by factors such as imbalance, bearing clearance, and vibration of the support structure during the rotation of the shaft. By using the instantaneous displacement information as a reference input, the adaptive filtering algorithm can learn and model the dynamic relationship between instantaneous displacement and mechanical vibration noise, thereby accurately estimating the mechanical vibration noise component in the original force detection signal. Subsequently, the estimated mechanical vibration noise component is subtracted from the original force detection signal to obtain a purer inherent centrifugal force characteristic signal of the shaft.

[0082] This application's solution addresses the challenge of accurately separating non-periodic vibration components in complex dynamic environments using traditional methods by introducing an adaptive filtering algorithm. Specifically, instantaneous displacement information directly reflects the actual motion trajectory and attitude changes of the shaft during rotation, which are often the primary causes of mechanical vibration noise. Therefore, using instantaneous displacement information as the reference input to the adaptive filter provides the filter with auxiliary information highly correlated with the noise source. Simultaneously, the original force detection signal contains the shaft's inherent centrifugal force and various interference noises (including mechanical vibration noise). The adaptive filter continuously adjusts its internal parameters to optimally approximate the mechanical vibration noise component in the original force detection signal. This process essentially utilizes the correlation between instantaneous displacement information and mechanical vibration noise to separate the noise from the original force detection signal. Because the adaptive filtering algorithm can track and estimate noise components in real-time and dynamically, it effectively suppresses interference even when the shaft's rotation state or environmental conditions change, thus ensuring the accuracy of the extracted inherent centrifugal force feature signal.

[0083] In some preferred embodiments, this application is implemented as follows: Assume that during the rotation detection of a brushed motor rotor, instantaneous displacement information (including radial offset and circumferential offset angle) acquired by a non-contact displacement sensor is input to the reference input of an adaptive filter (e.g., a filter based on the Least Mean Square (LMS) algorithm). Simultaneously, the original force detection signal detected by a pressure sensor is input to the desired output of the adaptive filter. The adaptive filter adjusts its internal filter coefficients in real time according to the LMS algorithm to minimize the error between its output and the desired output (the original force detection signal). In this way, the filter can learn and estimate the mechanical vibration noise component in the original force detection signal that is related to the instantaneous displacement information. For example, when the shaft experiences radial offset at a specific circumferential position, if it is accompanied by fluctuations in the force signal at a specific frequency or amplitude, the adaptive filter will identify this correlation and estimate this fluctuation as mechanical vibration noise. Once the mechanical vibration noise component is estimated, it is subtracted from the original force detection signal, resulting in a purer intrinsic centrifugal force characteristic signal free of mechanical vibration noise. For example, if the original force detection signal is F_r at a certain moment and the estimated mechanical vibration noise is N_c, then the extracted inherent centrifugal force feature signal F_i = F_r - N_c. This processing method ensures that even in the case of slight imbalance of the shaft or small vibration of the support system, the inherent centrifugal force effect of the rotor itself can be accurately separated.

[0084] In actual testing environments, the original force detection signal, in addition to containing non-periodic vibration components related to shaft motion, may also be affected by external environmental vibrations. These external environmental vibrations may superimpose on the original force detection signal in the form of common-mode interference, thus affecting the accuracy of subsequent extraction of inherent centrifugal force characteristic signals. If the above problems are not addressed, the common-mode interference components introduced by the external environment may be misjudged as the centrifugal force characteristics of the rotor itself, leading to deviations in the detection results of the centroid distribution data and shaft roundness.

[0085] In some preferred embodiments, prior to step A401, the following steps are also included: A400a. Obtain the environmental vibration reference signal on the rotary support; A400b. Calculate the cross-power spectral density of the environmental vibration reference signal and the original force detection signal; A400c. Identify the common-mode interference component introduced by the external environment in the original force detection signal based on the cross-power spectral density; A400d. Filter out the common-mode interference component from the original force detection signal.

[0086] Specifically, in step A400a, the environmental vibration reference signal can be acquired in real time by deploying additional vibration sensors (such as accelerometers or piezoelectric sensors) on the rotating support. These sensors are configured to detect mechanical vibrations from the external environment, rather than vibrations caused by the rotor's own rotation or imbalance. The purpose is to provide a reference input that is highly correlated with the external environmental vibrations for subsequent interference identification and suppression.

[0087] In step A400b, cross-power spectral density (CPSD) is a statistical measure of the correlation between two signals in the frequency domain. By calculating the CPSD of the environmental vibration reference signal and the original force detection signal (the calculation method for CPSD is existing technology and will not be detailed here), the degree of correlation between the two at different frequencies can be quantified. The purpose is to reveal which frequency components in the original force detection signal share a common source with the external environmental vibration, thereby providing a basis for identifying common-mode interference.

[0088] In practical applications, step A400c involves identifying common-mode interference components based on the cross-power spectral density (CPD). Specifically, this means analyzing the amplitude and phase information of the CPD at different frequencies. When the CPD exhibits a high amplitude and stable phase relationship at a certain frequency, it indicates that there is a significant common-mode component between the original force detection signal and the environmental vibration reference signal at that frequency. For example, a preset CPD threshold can be set. When the CPD of a certain frequency component exceeds this threshold, it is identified as a common-mode interference component introduced by the external environment. The purpose is to accurately distinguish between the rotor's inherent signal and external interference.

[0089] Furthermore, in step A400d, common-mode interference components are filtered out from the original force detection signal. This can be achieved using signal processing techniques such as adaptive filtering, notch filtering, or Wiener filtering based on cross-power spectral density. For example, an adaptive filter can be constructed, using the environmental vibration reference signal as input and the original force detection signal as the desired output. The filter coefficients are iteratively adjusted through an algorithm to estimate and remove common-mode interference components from the original force detection signal. The aim is to purify the original force detection signal, making it more accurately reflect the force characteristics of the rotor itself.

[0090] The proposed solution effectively identifies and filters out common-mode interference components introduced by the external environment from the original force detection signal by introducing an environmental vibration reference signal and utilizing cross-power spectral density analysis. Specifically, when external vibrations occur, these vibrations simultaneously affect the rotating support and the shaft, thereby generating common-mode interference in the original force detection signal. By acquiring the environmental vibration reference signal from the rotating support, the characteristics of this external vibration can be captured. Subsequently, the cross-power spectral density of the environmental vibration reference signal and the original force detection signal is calculated, quantifying their correlation in the frequency domain, thus accurately identifying the frequency components in the original force detection signal that are highly correlated with the external environmental vibration, i.e., the common-mode interference components. Once these common-mode interference components are identified, appropriate signal processing techniques can be used to filter them out of the original force detection signal. This ensures that the signal has been free of external environmental interference when processing the original force detection signal in subsequent steps A401 and A402, resulting in purer and more accurate extracted instantaneous displacement information and inherent centrifugal force characteristic signals.

[0091] In some implementations, step A5 includes: A501. Perform statistical analysis or geometric fitting on the actual geometric contour features to determine the roundness of the rotating shaft; A502. Perform amplitude and phase analysis on the inherent centrifugal force characteristic signal to determine the centroid distribution data; the centroid distribution data includes the radial offset and circumferential angle of the centroid of the brushed motor rotor.

[0092] The statistical analysis or geometric fitting of the actual geometric contour features aims to accurately quantify the roundness of the pivot. Statistical analysis may include calculating the mean and standard deviation of radial distances to assess their dispersion; geometric fitting, using algorithms such as the least squares method, fits the measurement point data to an optimal circle or ellipse, thereby deriving the roundness parameter. For example, a least squares circle fitting algorithm can be used to fit each circumferential measurement point in the actual geometric contour features to an optimal circle, and the maximum radial deviation of each circumferential measurement point from this optimal circle can be calculated, which is then used as the roundness of the pivot.

[0093] Furthermore, amplitude and phase analysis of the inherent centrifugal force characteristic signal is performed to accurately obtain the centroid distribution data of the brushed motor rotor. Amplitude analysis reveals the magnitude of the centrifugal force, which is directly related to the radial distance of the centroid from the rotation axis; phase analysis determines the specific position of the centroid in the circumferential direction. Specifically, frequency domain analysis methods such as Fourier transform can be used to extract the amplitude and phase information corresponding to the rotor rotation frequency from the inherent centrifugal force characteristic signal. The amplitude can be used to calculate the radial offset of the center of mass (for example, a quantitative relationship between the vibration signal amplitude and the radial offset of the center of mass can be established through a pre-calibration experiment; when a vibration amplitude is actually measured, the radial offset of the center of mass can be deduced from this calibration relationship). The phase can be used to determine the circumferential angle of the center of mass (for example, a "keyphasor" or photoelectric sensor can be set up, which generates a trigger signal as a reference at a specific angle, such as 0 degrees, for each revolution of the rotor; then, the "phase difference" between the peak or zero-crossing point of the vibration signal and the trigger signal of the keyphasor is measured, and this phase difference directly corresponds to the circumferential angle of the unbalanced center of mass in the direction of rotation). Thus, the center of mass distribution data, i.e., the radial offset and circumferential angle of the center of mass of the brushed motor rotor, can be accurately determined.

[0094] The proposed solution effectively eliminates the effects of measurement noise and instantaneous rotation axis fluctuations by performing statistical analysis or geometric fitting on the true geometric profile features. This allows for accurate characterization of the shaft's true geometry and precise determination of its roundness. Simultaneously, by performing amplitude and phase analysis on the inherent centrifugal force characteristic signal, the centrifugal force component directly related to rotor imbalance can be extracted from the complex force signal, and its magnitude and direction can be quantified. This enables accurate determination of the centroid distribution data of the brushed motor rotor, including the radial offset and circumferential angle of the centroid. This method combines geometric and dynamic characteristics, providing a reliable basis for comprehensively evaluating the rotor's balance and geometric accuracy.

[0095] refer to Figure 2 This application provides a brushed motor rotor detection system, including a host computer 1, two rotating support seats 2, a rotating drive device 3, at least three non-contact displacement sensors 4, a laser sensor 5, and two pressure sensors 6; the rotating drive device 3, the non-contact displacement sensors 4, the laser sensor 5, and the pressure sensors 6 are all electrically connected to the host computer 1. The two rotating support seats 2 are respectively used to support the two ends of the rotating shaft 91 of the brushed motor rotor 90; The rotary drive device 3 is used to drive the brushed motor rotor 90 to rotate. At least three of the non-contact displacement sensors 4 are distributed along different circumferential and axial positions of the rotating shaft 91 and are used to collect the spatial coordinates of the surface measuring points of the rotating shaft 91. The laser sensor 5 is used to scan the cross-sectional profile of the shaft 91 when the brushed motor rotor 90 rotates, so as to obtain the spatial coordinates of multiple circumferential measuring points. The two pressure sensors 6 are respectively installed at the two rotating support seats 2 and are used to detect the pressure of the rotating shaft 91 on the two rotating support seats 2; The host computer 1 is used to execute the steps in the brushed motor rotor 90 detection method described above.

[0096] Specifically, the host computer 1 can be an industrial control computer, a high-performance personal computer, or an embedded processing unit. Its main function is to receive and process data from various sensors, execute complex algorithms, and output detection results.

[0097] In one specific implementation, the rotating support 2 can adopt a V-shaped block structure, with the V-groove lined with a low-friction material, such as polytetrafluoroethylene (PTFE) or a ceramic coating, to reduce friction between the rotating shaft 91 and the support. In other implementations, the rotating support 2 can also adopt a roller-type support structure, using multiple precision rollers to support the rotating shaft 91, or a magnetic levitation support method, using magnetic force to levitate the rotating shaft 91, thereby minimizing friction and vibration caused by mechanical contact.

[0098] The rotary drive device 3 is used to drive the brushed motor rotor 90 to rotate. For example, a servo motor can be used in conjunction with a synchronous belt or gear transmission mechanism to drive the rotor. The servo motor can provide precise speed and position control, ensuring that the rotor rotates at a stable speed during the detection process. As another implementation, a direct drive motor can also be used, connecting the motor rotor directly to the brushed motor rotor 90, thereby reducing errors and vibrations caused by the transmission links. A flexible coupling can be used to connect the rotary drive device 3 to the rotating shaft 91 of the brushed motor rotor 90.

[0099] The at least three non-contact displacement sensors 4 are distributed at different circumferential and axial positions along the rotating shaft 91 and are used to acquire the spatial coordinates of surface measuring points on the rotating shaft 91. These sensors can be eddy current sensors, capacitive sensors, or optical triangulation sensors. Eddy current sensors measure distance by detecting changes in eddy currents near metallic objects, offering advantages such as non-contact operation, high precision, and resistance to contamination. Capacitive sensors determine distance by measuring changes in capacitance and are suitable for non-conductive materials or coated surfaces. Optical triangulation sensors utilize laser beams and CCD arrays to measure distance, providing high spatial resolution.

[0100] The laser sensor 5 can be a laser line scanner that acquires the cross-sectional profile by projecting a laser line and using a camera to capture the reflected light. In other embodiments, a point-type laser displacement sensor can also be used, which acquires profile data point by point by performing rapid scanning as the rotating shaft 91 rotates.

[0101] The pressure sensor 6 can be a strain gauge sensor, a piezoelectric sensor, or a force-sensitive resistor sensor. Strain gauge sensors detect pressure by measuring the change in resistance caused by material deformation, exhibiting good stability and linearity. Piezoelectric sensors utilize the piezoelectric effect to convert pressure into an electrical charge signal, offering a fast response. Force-sensitive resistor sensors detect force by measuring the change in resistance with pressure, and have a simple structure.

[0102] The above descriptions are merely some embodiments of the present invention. Those skilled in the art can make various modifications and improvements without departing from the inventive concept of the present invention, and these all fall within the scope of protection of the present invention.

Claims

1. A brush motor rotor detection method, characterized by, The method includes the following steps: A1. With the brushed motor rotor shaft supported by two rotating supports at both ends, acquire spatial position sampling data, original surface contour signals, and original force detection signals of the brushed motor rotor during rotation; the spatial position sampling data includes the spatial coordinates of multiple surface measuring points of the shaft collected from different axial and circumferential positions; the original surface contour signals include the spatial coordinates of circumferential measuring points of the shaft collected from preset detection positions; the original force detection signals include the pressure of the shaft on the two rotating supports; A2. Based on the spatial position sampling data, determine the instantaneous rotation axis of the rotating shaft at each sampling moment; A3. Using the instantaneous rotation axis as a reference, perform follow-up compensation processing on the original surface contour signal to obtain the true geometric contour features of the rotating shaft; A4. Extract the motion characteristics of the instantaneous rotation axis, and use the motion characteristics to perform interference suppression processing on the original force detection signal to obtain the inherent centrifugal force characteristic signal of the rotating shaft; A5. Based on the actual geometric contour features and the inherent centrifugal force feature signals, determine the centroid distribution data of the brushed motor rotor and the roundness of the shaft.

2. The brush motor rotor detection method of claim 1, wherein, Step A1 includes: A101. By deploying at least three non-contact displacement sensors at different circumferential and axial positions of the rotating shaft, the spatial coordinates of the surface measuring points of the rotating shaft are collected in real time as the spatial position sampling data; A102. The cross-sectional profile of the rotating shaft is scanned from a preset detection position using a laser sensor to obtain the spatial coordinates of multiple circumferential measuring points, which serve as the original surface profile signal. A103. The pressure of the rotating shaft on the two rotating support seats is detected by a pressure sensor and used as the original force detection signal; A104. Perform time alignment processing on the spatial location sampling data, the original surface contour signal, and the original force detection signal.

3. The brushed motor rotor detection method according to claim 1, characterized in that, Step A2 includes: A201. For each sampling time, perform linear fitting on the spatial coordinates of the surface measuring points corresponding to that sampling time in the spatial location sampling data to construct the instantaneous rotation axis of the rotating shaft at that sampling time.

4. The brushed motor rotor detection method according to claim 3, characterized in that, Step A2 also includes: A202. Obtain the nominal radius of the shaft; A203. Perform inverse geometric fitting on the original surface contour signal based on the nominal radius to determine the reference axis; A204. Compare the instantaneous rotation axis with the reference axis to identify systematic measurement drift in the spatial position sampling data; A205. Dynamically compensate the instantaneous rotation axis based on the systematic measurement drift.

5. The brushed motor rotor detection method according to claim 1, characterized in that, Step A3 includes: A301. Calculate the radial distance from each circumferential measuring point in the original surface profile signal to the corresponding instantaneous rotation axis; A302. Calculate the local rate of change of the radial distance along the circumferential direction of the rotating shaft; A303. A dynamic threshold is set according to the nominal geometric parameters of the rotating shaft, and the circumferential measuring points whose local rate of change exceeds the dynamic threshold are identified as abnormal reflection points; the nominal geometric parameters include the nominal radius and roundness tolerance; A304. The radial distance of the abnormal reflection point is removed or interpolated for repair, and the radial distance of each circumferential measuring point after processing is used as the true geometric contour feature.

6. The brushed motor rotor detection method according to claim 1, characterized in that, Step A4 includes: A401. Obtain the instantaneous displacement information of the instantaneous rotation axis relative to the preset ideal rotation axis, as the motion feature; the instantaneous displacement information includes radial offset and circumferential offset angle; A402. Identify and remove the non-periodic vibration components related to the instantaneous displacement information from the original force detection signal to extract the inherent centrifugal force characteristic signal of the shaft.

7. The brushed motor rotor detection method according to claim 6, characterized in that, Step A402 includes: Using the instantaneous displacement information as a reference input and the original force detection signal as the desired output, the mechanical vibration noise component in the original force detection signal is estimated by an adaptive filtering algorithm and used as the non-periodic vibration component. The inherent centrifugal force characteristic signal is extracted by subtracting the mechanical vibration noise component from the original force detection signal.

8. The brushed motor rotor detection method according to claim 6, characterized in that, Before step A401, the following are also included: A400a. Obtain the environmental vibration reference signal on the rotary support; A400b. Calculate the cross-power spectral density of the environmental vibration reference signal and the original force detection signal; A400c. Identify the common-mode interference component introduced by the external environment in the original force detection signal based on the cross-power spectral density; A400d. Filter out the common-mode interference component from the original force detection signal.

9. The brushed motor rotor detection method according to claim 1, characterized in that, Step A5 includes: A501. Perform statistical analysis or geometric fitting on the actual geometric contour features to determine the roundness of the rotating shaft; A502. Perform amplitude and phase analysis on the inherent centrifugal force characteristic signal to determine the centroid distribution data; the centroid distribution data includes the radial offset and circumferential angle of the centroid of the brushed motor rotor.

10. A brushed motor rotor detection system, characterized in that, It includes a host computer, two rotating support bases, a rotating drive device, at least three non-contact displacement sensors, a laser sensor, and two pressure sensors; the rotating drive device, the non-contact displacement sensors, the laser sensor, and the pressure sensor are all electrically connected to the host computer; The two rotating support seats are respectively used to support the two ends of the rotating shaft of the brushed motor rotor; The rotary drive device is used to drive the brushed motor rotor to rotate. At least three of the non-contact displacement sensors are distributed along different circumferential and axial positions of the rotating shaft and are used to collect the spatial coordinates of surface measuring points of the rotating shaft; The laser sensor is used to scan the cross-sectional profile of the shaft when the brushed motor rotor rotates, so as to obtain the spatial coordinates of multiple circumferential measuring points. The two pressure sensors are respectively installed at the two rotating support seats and are used to detect the pressure of the rotating shaft on the two rotating support seats; The host computer is used to execute the steps in the brushed motor rotor detection method according to any one of claims 1-9.

Images