Method and system for on-line measurement of generator set rotor eccentricity

Abstract

The present application relates to the technical field of generator set rotor eccentricity measurement, in particular to a kind of generator set rotor eccentricity online measurement method and system, two displacement sensors are respectively arranged in the cross section of the outside of rotor bearing and the cross section of across, to form double cross section measurement array;Synchronously collect the gap signal of each sensor and convert into effective physical gap value;Respectively solve to obtain the common mode motion coordinates of bearing and the original comprehensive motion coordinates of across;Using the shafting transfer function model established by system identification in advance, the common mode motion coordinates are corrected, and the transfer component in the cross section of across is predicted;Subtract the transfer component from the original comprehensive motion coordinates, obtain the pure rotor dynamic eccentricity coordinates;Finally, the total eccentricity is synthesized and intelligent early warning is carried out.The present application realizes the high-precision online measurement of rotor eccentricity, effectively eliminates the interference of shaft neck wobble and other common mode motion, and the measurement result is accurate and reliable.

Description

Technical Field

[0001] This invention relates to the field of generator set rotor eccentricity measurement technology, and in particular to an online measurement method and system for generator set rotor eccentricity. Background Technology

[0002] During generator operation, the rotor may deviate from its theoretical center of rotation due to various factors such as thermal stress, gravity, and electromagnetic forces; this is known as rotor eccentricity. Rotor eccentricity leads to increased unit vibration, bearing wear, and uneven air gap. In severe cases, it can cause rotor-stator friction, resulting in major accidents. Currently, rotor eccentricity measurement mainly relies on offline manual measurement after shutdown or obtaining approximate signals during operation using simple slip ring devices. Offline measurements cannot reflect the true eccentricity under operating conditions, while online measurement methods generally suffer from low accuracy, poor stability, susceptibility to electromagnetic interference, complex installation, and inability to effectively separate common-mode movements such as rotor bending and journal sway. Therefore, there is an urgent need for an online measurement method that can achieve high accuracy, high stability, and ease of implementation under complex operating conditions. Summary of the Invention

[0003] The purpose of this invention is to provide a method and system for online measurement of rotor eccentricity in generator sets, which achieves high-precision online measurement of rotor eccentricity, effectively eliminates interference from common-mode motions such as journal sway, and provides accurate and reliable measurement results.

[0004] To achieve the above objectives, in a first aspect, the present invention provides a method for online measurement of rotor eccentricity of a generator set, comprising the following steps:

[0005] S1: A first vertical sensor and a first horizontal sensor are installed on the outer section of the generator rotor bearing, and a second vertical sensor and a second horizontal sensor are installed on the mid-span monitoring section of the rotor, forming a dual-section sensor array. The outer section of the bearing refers to the area located on the generator rotor journal, outside the outer end face of the main bearing. On the stator housing or a rigid support fixed to the foundation at the outer section of the bearing, the first vertical sensor and the first horizontal sensor are installed in both the vertical and horizontal directions. The probes of the two sensors face the rotor surface, and their axes should intersect near the theoretical geometric center line of the rotor. The mid-span monitoring section refers to the area on the rotor shaft located precisely between the two main bearings. Sensors are installed on the stator housing at this location, and the second vertical sensor and the second horizontal sensor are installed in both the vertical and horizontal directions, respectively.

[0006] S2: Simultaneously acquire the rotor surface gap voltage signals sensed by four sensors, and convert each gap voltage signal into the corresponding effective physical gap value;

[0007] S3: Based on the effective physical gap value of the first vertical sensor and the first horizontal sensor, the journal center coordinates of the outer section of the bearing are calculated and used as common mode motion coordinates;

[0008] S4: Based on the effective physical gap values ​​of the second vertical sensor and the second horizontal sensor, the original coordinates of the rotor center of the mid-span monitoring section are calculated and used as the original comprehensive motion coordinates;

[0009] S5: Using a pre-established shaft system transfer function model that reflects the motion transmission relationship from the outer section of the bearing to the mid-span monitoring section, the common mode motion coordinates are processed to predict the transmission influence component at the mid-span monitoring section;

[0010] S6: Subtract the transmitted influence component from the original comprehensive motion coordinates to obtain the rotor dynamic eccentricity coordinates of the mid-span monitoring section;

[0011] S7: Perform vector synthesis and feature extraction on the rotor dynamic eccentricity coordinates to obtain eccentricity feature values ​​for state assessment.

[0012] Specifically, converting each gap voltage signal into a corresponding effective physical gap value includes:

[0013] Based on the pre-stored calibration models of each sensor, the gap voltage signal is converted into the original physical gap value;

[0014] The average value of the gap readings of each sensor is obtained under stable and unbiased conditions and used as the dynamic baseline reference value.

[0015] The effective physical gap value is obtained by subtracting the corresponding dynamic baseline reference value from the original physical gap value of each sensor.

[0016] The methods for establishing the shaft system transfer function model include:

[0017] During unit operation, the common mode motion coordinates and the original composite motion coordinates are synchronously collected within a set time period;

[0018] Spectral analysis was performed on the X and Y components of the two coordinates respectively, and the frequency response function in the corresponding direction from the outer section of the bearing to the mid-span monitoring section was calculated.

[0019] Extract the amplitude-frequency and phase-frequency characteristics of the frequency response function within the main operating frequency range of the unit, and construct and store them as the shaft transfer function model.

[0020] Specifically, step S5 includes:

[0021] The common-mode motion coordinates acquired in real time are decomposed in the frequency domain;

[0022] Using the aforementioned shaft transfer function model, amplitude and phase corrections are performed on each frequency component;

[0023] The corrected frequency components are synthesized into a time-domain signal to obtain the transmitted influence component.

[0024] Vector synthesis and feature extraction include:

[0025] Calculate the real-time total eccentricity based on the dynamic eccentricity coordinates;

[0026] Statistical analysis is performed on the real-time total eccentricity, and at least the average value of the frequency band eccentricity and the peak-to-peak value of the frequency band eccentricity are extracted as the characteristic values ​​of the eccentricity.

[0027] The method further includes

[0028] Set attention thresholds and danger thresholds for the eccentricity characteristic values;

[0029] The eccentricity feature value is compared with the attention threshold and the danger threshold in real time. If the attention threshold is exceeded, a warning is issued. If the danger threshold is reached or exceeded, an alarm is issued.

[0030] The methods for setting the attention threshold and the danger threshold are as follows:

[0031] During the healthy operation phase of the unit, the characteristic values ​​of the eccentricity are statistically analyzed to establish a health baseline;

[0032] The attention threshold and danger threshold are set based on the statistical mean and standard deviation of the health baseline.

[0033] The method further includes a threshold adaptive update step:

[0034] Periodically reassess the long-term statistical distribution of the eccentricity eigenvalues;

[0035] If the statistical distribution undergoes a stable, non-abrupt shift, the attention threshold and danger threshold are gradually adjusted based on the new statistical distribution.

[0036] The installation of the four sensors includes:

[0037] By rotating the rotor in a static state and adjusting each sensor individually to minimize the fluctuation of its output reading during one revolution of the rotor, a uniform initial target spacing can be set.

[0038] Adjust the sensor orientation so that the static circumferential runout error of each sensor in its measurement direction is lower than a preset threshold.

[0039] Secondly, the present invention provides an online measurement system for generator rotor eccentricity, applied to the online measurement method for generator rotor eccentricity as provided in the first aspect. The online measurement system for generator rotor eccentricity includes a dual-section sensor array, a synchronous data acquisition unit, a signal processing and calculation unit, and an early warning output unit.

[0040] The dual-section sensor array includes a first vertical sensor and a first horizontal sensor installed on the outer section of the bearing, and a second vertical sensor and a second horizontal sensor installed on the mid-span monitoring section.

[0041] The synchronous data acquisition unit is used to synchronously acquire signals from the four sensors;

[0042] The signal processing and resolution unit is used to execute steps S2 to S7 of the method;

[0043] The warning output unit is used to output warning or alarm information based on the eccentricity characteristic value.

[0044] This invention discloses an online measurement method and system for generator rotor eccentricity. Two displacement sensors are arranged at the outer section of the rotor bearing and the mid-span monitoring section, forming a dual-section measurement array. The gap signals of each sensor are synchronously acquired and converted into effective physical gap values. The common-mode motion coordinates at the bearing and the original comprehensive motion coordinates at the mid-span are calculated separately. Using a shaft transfer function model pre-established through system identification, the common-mode motion coordinates are corrected, and their transmission component at the mid-span section is predicted. This transmission component is subtracted from the original comprehensive motion coordinates to obtain the pure rotor dynamic eccentricity coordinates. Finally, the total eccentricity is synthesized, and intelligent early warning is provided. This invention achieves high-precision online measurement of rotor eccentricity, effectively eliminating interference from common-mode motions such as journal sway, and the measurement results are accurate and reliable. Attached Figure Description

[0045] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below.

[0046] Figure 1 This is a schematic diagram of the steps of an online measurement method for rotor eccentricity of a generator set according to the first embodiment of the present invention.

[0047] Figure 2 This is a flowchart illustrating an online method for measuring rotor eccentricity of a generator set provided by the present invention.

[0048] Figure 3 This is a schematic diagram showing the positions of the outer cross section and the mid-span monitoring cross section of the bearing provided by the present invention.

[0049] Figure 4This is a schematic diagram of the electronic device of the present invention. Detailed Implementation

[0050] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application.

[0051] The terminology used in this application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application. The singular forms “a,” “the,” and “the” used in this application and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used herein refers to and includes any or all possible combinations of one or more of the associated listed items.

[0052] It should be understood that although the terms first, second, third, etc., may be used in this application to describe various information, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, without departing from the scope of this application, first information may also be referred to as second information, and similarly, second information may also be referred to as first information. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to determination."

[0053] The first embodiment of this application is as follows:

[0054] Please see Figures 1-3 This invention provides an online method for measuring the rotor eccentricity of a generator set, comprising the following steps:

[0055] S1: A first vertical sensor and a first horizontal sensor are installed on the outer side of the bearing section of the generator rotor, and a second vertical sensor and a second horizontal sensor are installed on the mid-span monitoring section of the rotor to form a dual-section sensor array.

[0056] Specifically, the installation accuracy of the sensors directly determines the reliability of the measurement reference. This scheme adopts a step-by-step fine-tuning installation method to ensure that the sensor sensing surface axis is perpendicular to the rotor surface, and that their spatial position strictly conforms to the measurement model. The outer bearing section refers to the main bearing located on the generator rotor journal, adjacent to its driving or non-driving end, and outside the outer end face of that bearing. At this location, on the stator housing or a rigid support fixed to the foundation, two non-contact displacement sensors (referred to as the first vertical sensor and the first horizontal sensor, respectively) are installed in the vertical direction (Y-axis) and horizontal direction (X-axis). The probes of the two sensors face the rotor surface, and their axes should intersect near the rotor's theoretical geometric center line. The core purpose is to monitor the relative movement of the rotor journal within the bearing (i.e., the shaft center trajectory). The mid-span monitoring section refers to the area on the rotor shaft located precisely between the two main bearings. This is the location where the rotor is most prone to static bending and dynamic deflection under the influence of gravity and electromagnetic forces, and is also the most sensitive location for monitoring rotor thermal bending and other faults. The sensor is mounted on the stator housing at this location, with one non-contact displacement sensor (referred to as the second vertical sensor and the second horizontal sensor, respectively) mounted in both the vertical (Y-axis) and horizontal (X-axis) directions. The installation requirements are the same as for the first measurement location. The purpose is to monitor the overall motion of the rotor body relative to the stator. These four sensors together form a dual-measurement-point, orthogonally arranged sensor array.

[0057] Each sensor is fixed to the generator stator housing or a separate, stable base via a rigid mounting bracket with three-dimensional adjustment functions (including radial advance / retreat, yaw angle, and pitch angle). Before installation, a high-precision dial indicator or laser alignment instrument is used to initially adjust the mechanical center of the four sensor mounting brackets, using the bearing housing bore or a calibrated rotor dummy shaft (in the stopped state) as a reference, so that they are aligned as closely as possible with the theoretical rotation center line of the rotor.

[0058] A suitable initial spacing refers to the static installation distance between the sensor probe end face and the rotor surface being measured. This distance must meet two core requirements: first, it must be located in the central region of the sensor's linear measurement range (for example, for a sensor with a range of 2mm, the linear region is typically 0.5mm to 1.8mm, so 1.0mm to 1.2mm is preferred as the initial spacing); second, it must allow sufficient positive and negative space for the rotor's potential maximum dynamic eccentricity, shaft drift, and the unit's thermal expansion. The specific setting steps are as follows:

[0059] Reference setting: With the unit in a cold, stationary state, manually rotate the rotor slowly to one revolution. Using the precision radial advance / retreat adjustment mechanism of the mounting bracket, combined with the gap voltage indication on the sensor, adjust the first vertical sensor, the first horizontal sensor, the second vertical sensor, and the second horizontal sensor respectively, so that the gap voltage reading output by the sensor fluctuates minimally near its linear midpoint value when the rotor rotates one revolution. The gap at this point is the set target initial gap (e.g., corresponding to 1.1 mm).

[0060] Dynamic Verification and Unification: After initial setup, rotate the rotor for another full revolution, continuously recording the gap values ​​of the four sensors at multiple equally spaced angles. The range of each sensor reading within one revolution is calculated using the host computer software. The range of all sensor readings must be less than a preset strict threshold (e.g., 0.02 mm). If the range of a sensor exceeds the limit, its pitch or yaw angle is fine-tuned to eliminate errors caused by the probe not being parallel to the measured surface. This step ensures that, under static conditions, the circumferential runout of the rotor surface relative to each sensor is minimized, meaning each sensor uses the same "electrical zero point" as a reference.

[0061] Calibration information entry: The theoretical installation angle (e.g., 90° for the first vertical sensor), the measurement section (bearing reference section or mid-span monitoring section), and the measured static initial spacing value (e.g., 1.12mm) of each sensor are entered into the host computer configuration database as inherent attribute parameters of the sensor to achieve a strong correlation between physical location and data channel.

[0062] S2: Synchronously acquire the rotor surface gap voltage signals sensed by four sensors, and convert each gap voltage signal into the corresponding effective physical gap value.

[0063] Specifically, during unit operation, four sensors work simultaneously, generating continuous intermittent analog signals. Correctly distinguishing, conditioning, and synchronously acquiring these signals is fundamental to subsequent algorithmic processing. Each sensor's signal cable connects to an independent channel with a unique number in the signal conditioning module. This channel number is permanently bound in the host computer software to the sensor attributes entered during installation (e.g., "First Vertical Sensor"). Therefore, data streams from any sensor carry their physical identifier (e.g., "CH1: Bearing Section - Vertical") from the start of acquisition, achieving natural data differentiation and traceability. The software interface and data storage structure are organized using this identifier, eliminating the possibility of channel confusion.

[0064] The signal conditioning module is responsible for signal amplification and filtering. It is optimized for the signal characteristics of eddy current sensors: due to slight differences in the initial spacing of sensors at different locations and varying vibration amplitudes at different rotor positions, each channel is equipped with an independent programmable gain amplifier to ensure optimal signal-to-noise ratio and resolution for the signal fed into the analog-to-digital converter (ADC). The system automatically calculates and sets a base gain based on the input initial spacing value, ensuring the quiescent operating point is in the middle of the ADC's range. After amplification, each channel signal passes through a low-pass filter with a configurable cutoff frequency. This cutoff frequency is set based on at least 2.5 times the unit's highest operating speed (corresponding to the highest eccentric vibration frequency) (following the Nyquist sampling theorem) and considering potential high-frequency electromagnetic interference characteristics (e.g., for a 3000 rpm unit, with a fundamental frequency of 50Hz, focusing primarily on harmonics, the cutoff frequency can be set from 500Hz to 1kHz). This filter effectively removes unwanted signals such as high-frequency switching noise and radio frequency interference, preventing spectral aliasing during ADC conversion.

[0065] The synchronous data acquisition unit internally includes a highly stable master clock oscillator and a multi-channel synchronous sample-and-hold circuit. All analog input channels share the same sampling clock signal. Upon receiving the "start acquisition" command, the sample-and-hold circuits of all channels simultaneously freeze the analog voltage values ​​of their respective channels at the same rising edge of the sampling clock. Subsequently, these simultaneously sampled voltage values ​​are sequentially sent to a high-speed analog-to-digital converter for digitization (for multi-channel ADC systems, it is necessary to ensure that the clocks of each ADC are synchronized), or polled to a single high-speed ADC via a multiplexer. Regardless of the circuit structure used, the core is to ensure that the sampling timestamps of all channels are strictly consistent at the microsecond or even nanosecond level. The digitized data is packaged into a synchronous data frame, which contains a unified timestamp and the gap digital values ​​of all channels (first vertical sensor, first horizontal sensor, second vertical sensor, second horizontal sensor) at that moment. This data frame is sent to the host computer in real time.

[0066] For each eddy current displacement sensor (first vertical sensor, first horizontal sensor, etc.), high-precision calibration must be performed in the laboratory before installation. The calibration process is as follows: Under standard temperature, using a precision displacement stage, the distance between the sensor probe and the standard target surface is changed in steps much smaller than the sensor's measurement range (e.g., 0.05 mm), while simultaneously recording the corresponding output voltage values. This yields a set of discrete "distance-voltage" calibration points covering the entire linear measurement range. Based on these calibration points, a unique "voltage-gap" inverse function model G(V) is established for each sensor using piecewise linear interpolation or high-order polynomial fitting. This model is stored in the host computer database and bound to the sensor's identifier. In subsequent measurements, the real-time acquired voltage value V is directly converted into the original physical gap value d_raw by querying its unique model G(V).

[0067] Dynamic baseline compensation:

[0068] Considering the potential zero-point drift of sensors after long-term operation, and the impact of ambient temperature changes on sensors and cables during unit operation, directly using a static calibration model would introduce significant system errors. Therefore, this solution introduces a dynamic baseline compensation mechanism.

[0069] Establishment of baseline reference values: During the low-speed warm-up and stabilization phase after each unit startup and before grid connection and load bearing, the system automatically records the long-term average value of the gap readings of each sensor during this period. Since the rotor eccentricity is small and changes slowly during this phase, this average value is regarded as the gap reference value corresponding to the new "electrical zero point" of each sensor under the current operating conditions, denoted as d_ref.

[0070] Compensation Calculation: In subsequent real-time measurements, the original gap value d_raw calculated by the model is subtracted from its corresponding value d_ref to obtain the effective physical gap value d_comp after zero-point drift compensation. That is: d_comp = d_raw - d_ref. This operation dynamically "zeros" the measurement baseline, effectively eliminating the slow drift error of the sensor channel and ensuring the long-term stability of the measurement reference.

[0071] S3: Based on the effective physical gap values ​​of the first vertical sensor and the first horizontal sensor, the journal center coordinates of the outer section of the bearing are calculated and used as common mode motion coordinates.

[0072] Specifically, assume the rotor is an ideal circle at this cross-section with a known nominal radius R. Let the effective clearance in the Y direction measured by the first vertical sensor be d_Y1, and the effective clearance in the X direction measured by the first horizontal sensor be d_X1. Define the initial target spacing set during sensor installation as S (1.1 mm as mentioned above). When the rotor center shifts in the XY plane, the clearances in each direction will change.

[0073] Based on geometric relationships, the instantaneous coordinates (X_bearing, Y_bearing) of the rotor center at this cross-section can be calculated using the following formula:

[0074] X_bearing = (d_X1 - d_X1_static) ≈ (S - d_X1) [because d_X1_static ≈ S]

[0075] Y_bearing = (d_Y1 - d_Y1_static) ≈ (S - d_Y1)

[0076] More precisely, if the calibration data for static circumferential runout is considered, the average value of one cycle of static turning can be used as d_X1_static and d_Y1_static. The calculated (X_bearing, Y_bearing) represents the real-time offset of the journal center within the bearing. This motion is mainly caused by oil film fluctuations, bearing housing vibrations, etc., and is defined as the common-mode motion to be subtracted.

[0077] S4: Based on the effective physical gap values ​​of the second vertical sensor and the second horizontal sensor, the original coordinates of the rotor center of the mid-span monitoring section are calculated and used as the original comprehensive motion coordinates.

[0078] Specifically, based on the effective gap values ​​d_Y2 and d_X2 measured by the second vertical sensor and the second horizontal sensor, and using the same geometric model, the instantaneous coordinates (X_mid_raw, Y_mid_raw) of the rotor center at the mid-span section relative to the sensor mounting base at that location can be calculated. This coordinate system contains two motion components: one is the actual rotor eccentricity / bending deformation that we want to obtain; the other is the common-mode motion transmitted from the bearing through the rotor body.

[0079] S5: Using a pre-established shaft system transfer function model that reflects the motion transmission relationship from the outer section of the bearing to the mid-span monitoring section, the common mode motion coordinates are processed to predict the transmission influence component at the mid-span monitoring section.

[0080] Specifically, traditional methods might simply subtract the proportionally scaled bearing coordinates from the mid-span coordinates, based on the assumption of a rigid rotor. However, in reality, a rotor is an elastic body, and the amplitude and phase of common-mode motion may change as it is transmitted from the bearings to the mid-span. This solution employs a more accurate transfer function correction method.

[0081] During the initial startup after unit installation, commissioning, or overhaul, a stable operating speed range (such as rated speed) is selected. During this period, the common-mode motion coordinates (X_bearing, Y_bearing) of the journal at the bearing naturally contain rich frequency components (such as rotational speed frequency and oil film whirl frequency), which can be used as the system's input excitation signal. The system synchronously acquires this input signal and the original composite motion coordinates (X_mid_raw, Y_mid_raw) of the mid-span section as the output response signal, continuously acquiring data for a sufficiently long time (e.g., 10 minutes) to ensure that the data covers the main vibration modes. Spectral analysis is performed on the synchronously acquired input and output time series data. For the X direction, the frequency response function H_x(f) between the input sequence X_bearing(t) and the output sequence X_mid_raw(t) is calculated. Specifically, this can be obtained by calculating the ratio of their cross-power spectral density to the input self-power spectral density (e.g., using the H1 estimation method). Similarly, the frequency response function H_y(f) in the Y direction is calculated. This frequency response function is the transfer model we need. It contains complete information on the amplitude amplification factor |H(f)| and phase delay angle ∠H(f) for vibration components at different frequencies f, from the bearing to the mid-span. In practice, to reduce online computation, the identified frequency response function can be rationally simplified. Key features within the main operating frequency range of the unit (e.g., 0.5 to 5th harmonics) are extracted, such as the amplitude ratio and phase difference at the rotational frequency (1X) and its second harmonic (2X), to construct a sparse frequency transfer parameter table. This parameter table is stored in the host computer as the shaft transfer function model.

[0082] In real-time online measurement, the bearing section coordinate sequence (X_bearing(t), Y_bearing(t)) is processed by fast Fourier transform or digital filtering. Then, the transfer functions H_x(f) and H_y(f) are used to calculate the transmission influence components (X_coupled(t), Y_coupled(t)) that should be presented when the common mode motion is transmitted to the mid-span section.

[0083] S6: Subtract the transmitted influence component from the original comprehensive motion coordinates to obtain the rotor dynamic eccentricity coordinates of the mid-span monitoring section.

[0084] Specifically, during the real-time online monitoring phase, for each synchronized data frame (including X_bearing, Y_bearing, X_mid_raw, and Y_mid_raw at the current moment), the actual eccentricity is extracted according to the following steps:

[0085] The bearing coordinate sequence (X_bearing, Y_bearing) within the current time window and the previous time window is decomposed into different frequency components using a Fast Fourier Transform (FFT). Then, each frequency component is corrected using the stored shaft transfer function model (frequency response function or parameter table): the amplitude of the frequency component at the bearing is multiplied by the corresponding amplitude amplification factor |H(f)| of the model, and the phase is added to the corresponding phase delay angle ∠H(f) of the model. Finally, all corrected frequency components are synthesized back into the time domain signal using an Inverse Fourier Transform (IFT), yielding the predicted common-mode motion components (X_coupled_pred, Y_coupled_pred) transmitted to the mid-span section. This process is equivalent to performing a "filter" on the bearing motion using the transfer function.

[0086] Subtracting the predicted common-mode motion transmission components from the original composite motion coordinates measured at mid-span completes the stripping process.

[0087] X_eccentric = X_mid_raw - X_coupled_pred

[0088] Y_eccentric = Y_mid_raw - Y_coupled_pred

[0089] The obtained (X_eccentric, Y_eccentric) is the pure rotor dynamic eccentricity coordinate after removing the bearing wobble interference.

[0090] S7: Perform vector synthesis and feature extraction on the rotor dynamic eccentricity coordinates to obtain eccentricity feature values ​​for state assessment.

[0091] Specifically, after obtaining the real-time dynamic eccentricity coordinates, they need to be converted into more intuitive monitoring feature quantities.

[0092] Real-time total eccentricity and phase angle calculation:

[0093] Total eccentricity (trajectory radius): The X and Y eccentricities at each moment are vectorized and their instantaneous amplitude is calculated, i.e., the real-time total eccentricity R_inst.

[0094]

[0095] This value directly reflects the radial distance between the rotor's geometric center and the ideal rotation center at that moment.

[0096] Eccentric phase angle: Calculates the angle between the instantaneous eccentric vector and the reference reference (usually defined as 0° phase in the positive X-axis direction), i.e., the real-time eccentric phase angle θ_inst.

[0097] θ_inst(t) = arctan2[ Y_eccentric(t), X_eccentric(t) ] (using the four-quadrant arctangent function)

[0098] This angle indicates the direction of eccentricity, which helps determine the directionality of thermal bending or the effect of gravity.

[0099] Statistical characteristic generation (for trend analysis and early warning):

[0100] The system simultaneously calculates the real-time total eccentricity R_inst(t) within one rotation cycle or a fixed time window (e.g., 1 second), extracting two key statistical features:

[0101] Average frequency eccentricity R_avg: The average value of R_inst over a period of time, reflecting the static (or slowly varying) component of the eccentricity, which is related to thermal bending and installation alignment.

[0102] Peak-to-peak value of eccentricity R_pp: The difference between the maximum and minimum values ​​of R_inst over a period of time, reflecting the dynamic fluctuation component of eccentricity, which is related to factors such as rotor imbalance and looseness.

[0103] The warning threshold is not a fixed value, but is adaptively established based on the unit's own historical operating status, reflecting the intelligent warning concept of "combining relative change rate and absolute limit" from "absolute value alarm".

[0104] Warning threshold setting (baseline learning and dynamic update):

[0105] Health baseline establishment: During the initial break-in period after a new unit is put into operation or a major overhaul (e.g., 200 hours of stable operation), the system automatically records the long-term statistical values ​​of the average frequency eccentricity R_avg and the peak-to-peak frequency eccentricity R_pp during this period, and calculates their mean and standard deviation. Based on this, an initial warning threshold is set:

[0106] Note the threshold (yellow alert): set as the healthy baseline mean + 3 standard deviations. This threshold is used to detect early abnormal trends.

[0107] Danger threshold (red alert): Set as the average of the healthy baseline plus 5 standard deviations, while also considering the percentage of the rotor's design air gap (e.g., specified to be no more than 20% of the minimum air gap) as an absolute upper limit protection value. The smaller of the two values ​​is taken to ensure safety.

[0108] Threshold dynamic fine-tuning: The system runs continuously and periodically (e.g., monthly) reassesses the eccentric statistical characteristics over a stable period. If the new statistical distribution shows a stable, non-abrupt shift compared to the original healthy baseline, the threshold is allowed to be slowly adjusted in tracking. This avoids long-term false alarms due to slow degradation of normal unit performance (e.g., minor bearing wear), but prevents missed alarms of sudden failures by limiting the adjustment rate.

[0109] Warning triggering and output:

[0110] The system compares the average value of frequency eccentricity R_avg and the peak-to-peak value of frequency eccentricity R_pp with the currently set attention threshold and danger threshold in real time.

[0111] When any feature quantity exceeds the attention threshold for the first time, the system triggers a yellow warning, highlights it on the host computer interface, and generates a warning log to remind maintenance personnel to pay attention to the trend of eccentricity changes.

[0112] When any characteristic reaches or exceeds the danger threshold, the system immediately triggers a red alarm. In addition to strong on-screen warnings and log recordings, it can also notify key personnel through audible and visual alarms, SMS or network message pushes, and provide shutdown suggestion signals to the interlocking protection system.

[0113] In addition to amplitude warnings, the system also includes a rate of change warning. Specifically, it calculates the slope of change of R_avg and R_pp in real time over a short period. If this slope exceeds an empirically set value (indicating that eccentricity is rapidly worsening), an orange warning is triggered even if the absolute value does not exceed a threshold, enabling earlier detection of fault symptoms.

[0114] The second embodiment of this application is as follows:

[0115] This invention provides an online measurement system for generator rotor eccentricity, applied to an online measurement method for generator rotor eccentricity as provided in the first embodiment. The online measurement system includes a dual-section sensor array, a synchronous data acquisition unit, a signal processing and calculation unit, and an early warning output unit.

[0116] The dual-section sensor array includes a first vertical sensor and a first horizontal sensor installed on the outer section of the bearing, and a second vertical sensor and a second horizontal sensor installed on the mid-span monitoring section.

[0117] The synchronous data acquisition unit is used to synchronously acquire signals from the four sensors;

[0118] The signal processing and resolution unit is used to execute steps S2 to S7 of the method;

[0119] The warning output unit is used to output warning or alarm information based on the eccentricity characteristic value.

[0120] Regarding the system in the above embodiments, the specific manner in which each module performs its operations has been described in detail in the embodiments related to the method, and will not be elaborated upon here.

[0121] For the system embodiments, since they basically correspond to the method embodiments, the relevant parts can be referred to in the description of the method embodiments. The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this application according to actual needs. Those skilled in the art can understand and implement this without creative effort.

[0122] Accordingly, this application also provides an electronic device, comprising: one or more processors; a memory for storing one or more programs; and, when the one or more programs are executed by the one or more processors, causing the one or more processors to implement the online measurement method for generator rotor eccentricity as described above. Figure 4 The diagram shown is a hardware structure diagram of any device with data processing capabilities, used in an online measurement system for generator rotor eccentricity provided by an embodiment of the present invention. (Except for...) Figure 4 In addition to the processor, memory, and network interface shown, any data processing device in the embodiment may also include other hardware depending on the actual function of the data processing device, which will not be described in detail here.

[0123] Accordingly, this application also provides a computer-readable storage medium storing computer instructions, which, when executed by a processor, implement the online measurement method for generator rotor eccentricity as described above. The computer-readable storage medium can be an internal storage unit of any data processing device as described in any of the foregoing embodiments, such as a hard disk or memory. The computer-readable storage medium can also be an external storage device, such as a plug-in hard disk, smart media card (SMC), SD card, flash card, etc., equipped on the device. Furthermore, the computer-readable storage medium can include both internal storage units of any data processing device and external storage devices. The computer-readable storage medium is used to store the computer program and other programs and data required by the data processing device, and can also be used to temporarily store data that has been output or will be output.

[0124] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the disclosure herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

Claims

1. A method for online measurement of rotor eccentricity of a generator set, characterized in that, Includes the following steps: S1: A first vertical sensor and a first horizontal sensor are installed on the outer section of the generator rotor bearing, and a second vertical sensor and a second horizontal sensor are installed on the mid-span monitoring section of the rotor, forming a dual-section sensor array; wherein, the outer section of the bearing refers to the area located on the generator rotor journal and outside the outer end face of the main bearing, on the stator housing of the outer section of the bearing or on a rigid bracket fixed to the foundation, the first vertical sensor and the first horizontal sensor are installed in the vertical and horizontal directions respectively, the probes of the two sensors face the rotor surface, and their axes should intersect near the theoretical geometric center line of the rotor; the mid-span monitoring section refers to the area on the rotor shaft located in the middle of the two main bearings, the sensor is installed on the stator housing in the middle area of ​​the two main bearings, the second vertical sensor and the second horizontal sensor are installed in the vertical and horizontal directions respectively; S2: Simultaneously acquire the rotor surface gap voltage signals sensed by four sensors, and convert each gap voltage signal into the corresponding effective physical gap value; S3: Based on the effective physical gap value of the first vertical sensor and the first horizontal sensor, the journal center coordinates of the outer section of the bearing are calculated and used as common mode motion coordinates; S4: Based on the effective physical gap values ​​of the second vertical sensor and the second horizontal sensor, the original coordinates of the rotor center of the mid-span monitoring section are calculated and used as the original comprehensive motion coordinates; S5: Using a pre-established shaft system transfer function model that reflects the motion transmission relationship from the outer section of the bearing to the mid-span monitoring section, the common-mode motion coordinates are processed to predict the transmission influence components at the mid-span monitoring section, specifically: For the X direction, the frequency response function H_x(f) between the input sequence X_bearing(t) and the output sequence X_mid_raw(t) is calculated, specifically by calculating the ratio of their cross power spectral density to the input self power spectral density. Similarly, the frequency response function H_y(f) in the Y direction is calculated, along with the amplitude ratio and phase difference at the rotational frequency (1X) and its second harmonic (2X), to construct a sparse frequency transfer parameter table. This sparse frequency transfer parameter table is stored in the host computer as a shaft system transfer function model. In real-time online measurement, the bearing section coordinate sequence (X_bearing(t), Y_bearing(t)) is subjected to fast Fourier transform or digital filtering. Then, the transfer functions H_x(f) and H_y(f) are used to calculate the transfer influence components (X_coupled(t), Y_coupled(t)) that should be presented when the common mode motion is transferred to the mid-span section. S6: Subtract the transmitted influence component from the original comprehensive motion coordinates to obtain the rotor dynamic eccentricity coordinates of the mid-span monitoring section; S7: Perform vector synthesis and feature extraction on the rotor dynamic eccentricity coordinates to obtain eccentricity feature values ​​for state assessment.

2. The online measurement method for generator rotor eccentricity as described in claim 1, characterized in that, Convert each gap voltage signal into a corresponding effective physical gap value, specifically including: Based on the pre-stored calibration models of each sensor, the gap voltage signal is converted into the original physical gap value; The average value of the gap readings of each sensor is obtained under stable and unbiased conditions and used as the dynamic baseline reference value. The effective physical gap value is obtained by subtracting the corresponding dynamic baseline reference value from the original physical gap value of each sensor.

3. The online measurement method for generator rotor eccentricity as described in claim 1, characterized in that, Methods for establishing shaft system transfer function models include: During unit operation, the common mode motion coordinates and the original composite motion coordinates are synchronously collected within a set time period; Spectral analysis was performed on the X and Y components of the two coordinates respectively, and the frequency response function in the corresponding direction from the outer section of the bearing to the mid-span monitoring section was calculated. Extract the amplitude-frequency and phase-frequency characteristics of the frequency response function within the main operating frequency range of the unit, and construct and store them as the shaft transfer function model.

4. The online measurement method for generator rotor eccentricity as described in claim 1, characterized in that, Step S5 specifically includes: The common-mode motion coordinates acquired in real time are decomposed in the frequency domain; Using the aforementioned shaft transfer function model, amplitude and phase corrections are performed on each frequency component; The corrected frequency components are synthesized into a time-domain signal to obtain the transmitted influence component.

5. The online measurement method for generator rotor eccentricity as described in claim 1, characterized in that, Vector synthesis and feature extraction include: Calculate the real-time total eccentricity based on the dynamic eccentricity coordinates; Statistical analysis is performed on the real-time total eccentricity, and at least the average value of the frequency band eccentricity and the peak-to-peak value of the frequency band eccentricity are extracted as the characteristic values ​​of the eccentricity.

6. The online measurement method for rotor eccentricity of a generator set as described in claim 1, characterized in that, The method also includes Set attention thresholds and danger thresholds for the eccentricity characteristic values; The eccentricity feature value is compared with the attention threshold and the danger threshold in real time. If the attention threshold is exceeded, a warning is issued. If the danger threshold is reached or exceeded, an alarm is issued.

7. The online measurement method for generator rotor eccentricity as described in claim 6, characterized in that, The method for setting the threshold and danger threshold is as follows: During the healthy operation phase of the unit, the characteristic values ​​of the eccentricity are statistically analyzed to establish a health baseline; The attention threshold and danger threshold are set based on the statistical mean and standard deviation of the health baseline.

8. The online measurement method for generator rotor eccentricity as described in claim 7, characterized in that, The method further includes a threshold adaptive update step: Periodically reassess the long-term statistical distribution of the eccentricity eigenvalues; If the statistical distribution undergoes a stable, non-abrupt shift, the attention threshold and danger threshold are gradually adjusted based on the new statistical distribution.

9. The online measurement method for rotor eccentricity of a generator set as described in claim 1, characterized in that, The installation of the four sensors includes: By rotating the rotor in a static state and adjusting each sensor individually to minimize the fluctuation of its output reading during one revolution of the rotor, a uniform initial target spacing can be set. Adjust the sensor orientation so that the static circumferential runout error of each sensor in its measurement direction is lower than a preset threshold.

10. An online measurement system for generator rotor eccentricity, applied to the online measurement method for generator rotor eccentricity as described in claim 1, characterized in that, The online measurement system for generator rotor eccentricity includes a dual-section sensor array, a synchronous data acquisition unit, a signal processing and calculation unit, and an early warning output unit. The dual-section sensor array includes a first vertical sensor and a first horizontal sensor installed on the outer section of the bearing, and a second vertical sensor and a second horizontal sensor installed on the mid-span monitoring section. The synchronous data acquisition unit is used to synchronously acquire signals from the four sensors; The signal processing and resolution unit is used to execute steps S2 to S7 of the method; The warning output unit is used to output warning or alarm information based on the eccentricity characteristic value.

Images