Method and system for acquiring rotating speed of steam turbine in real time

Abstract

The invention provides a steam turbine rotating speed real-time acquisition method and system, and the method comprises the steps: synchronously collecting pulse signals generated by a steam turbine rotor through a Hall effect sensor and an eddy current sensor, and obtaining two paths of original pulse sequences; a first gain is applied to a Hall effect sensor signal, a second gain is applied to an eddy current sensor signal, an independent filtering window is adopted for moving average filtering, and the initial size of the filtering window is dynamically set according to the rated rotating speed of the current working condition stage of the steam turbine; based on the filtered pulse signals, calculating to obtain a first rotating speed value and a second rotating speed value, inputting the first rotating speed value and the second rotating speed value to a pre-trained signal quality evaluation model, and outputting a fusion weight coefficient for representing the consistency of the two paths of signals; and according to the fusion weight coefficient, weighted average calculation is carried out on the first rotating speed value and the second rotating speed value, a final rotating speed value and an equipment health state identifier are output, and data are selectively transmitted to a remote cloud platform in batches for storage, so that the rotating speed measurement precision is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of steam turbine speed acquisition, and particularly relates to a steam turbine speed real-time acquisition method and system. BACKGROUND

[0002] The overspeed protection system is the key defense line for ensuring the safe operation of the steam turbine, and its reliability directly depends on the accuracy of the speed acquisition data. At present, the steam turbine speed is monitored by using a speed acquisition card-based scheme in power plants. A typical system usually relies on a single type of sensor, or although a multi-sensor redundancy design is adopted, the data processing method is relatively simple, for example, only the speed values output by multiple sensors are averaged with fixed weights or simply compared with a threshold.

[0003] Such prior art schemes mainly have the following limitations: first, the signal quality of the sensor may deteriorate due to aging, electromagnetic interference or physical damage in long-term operation, and the simple average algorithm cannot intelligently identify and weaken the influence of the faulty sensor, but may contaminate the final result, and there is a risk of false positives or false negatives. Secondly, the filtering parameters of the system are usually fixed and preset, and cannot adapt to the dynamic changes of the speed signal characteristics under different working conditions such as start-stop and load variation of the steam turbine, resulting in a decrease in measurement accuracy during the variable working condition stage.

[0004] Therefore, how to improve the measurement accuracy of the overspeed protection system has become a technical problem to be solved by those skilled in the art. SUMMARY

[0005] The present application provides a steam turbine speed real-time acquisition method and system to solve the defects of poor measurement accuracy, reliability and economy of the overspeed protection system in the prior art.

[0006] In a first aspect, the present application provides a steam turbine speed real-time acquisition method, comprising:

[0007] synchronously acquiring pulse signals generated by the steam turbine rotor by using a Hall effect sensor and an eddy current sensor to obtain two original pulse sequences;

[0008] respectively performing signal conditioning on the two original pulse sequences, wherein a first gain is applied to the Hall effect sensor signal and a second gain is applied to the eddy current sensor signal;

[0009] respectively performing sliding average filtering on the two conditioned pulse signals by using independent filtering windows; wherein the initial size of the filtering window is dynamically set according to the rated speed of the current working condition stage of the steam turbine;

[0010] based on the filtered pulse signals, respectively calculating a first speed value and a second speed value;

[0011] The first and second rotational speed values ​​are input into a pre-trained signal quality assessment model, which outputs a fusion weight coefficient to characterize the consistency of the two signals.

[0012] Based on the fusion weighting coefficient, the first speed value and the second speed value are weighted and averaged to calculate the final speed value and the equipment health status indicator.

[0013] The final rotation speed value is written to a local circular buffer, and the data is selectively transmitted in batches to a remote cloud platform for storage based on the device health status identifier.

[0014] According to the present invention, a method for real-time acquisition of turbine rotational speed, prior to the synchronous acquisition of pulse signals generated by the turbine rotor via a Hall effect sensor and an eddy current sensor, further includes:

[0015] The Hall effect sensor and the eddy current sensor are installed along the circumference of the rotor at a preset angle. The installation positions of the Hall effect sensor and the eddy current sensor are calibrated to control the pulse sequence generated when the rotor rotates to have a predictable phase difference.

[0016] Based on the phase difference, corresponding phase compensation parameters are configured in the signal processing unit.

[0017] According to the present invention, a method for real-time acquisition of turbine speed, wherein applying a first gain to the Hall effect sensor signal includes:

[0018] The original pulse voltage signal output by the Hall effect sensor is amplified by connecting it to a fixed-gain in-phase amplifier circuit.

[0019] A bidirectional Zener diode is used to clamp the amplified signal, limiting the peak voltage to within the input voltage range allowed by the subsequent analog-to-digital converter.

[0020] The first gain is a fixed gain value that is preset and remains unchanged based on the rated output level of the Hall effect sensor and the range of the analog-to-digital converter.

[0021] According to a method for real-time acquisition of turbine speed provided by the present invention, the application of a second gain to the eddy current sensor signal includes:

[0022] The original amplitude-modulated AC signal output by the eddy current sensor is connected to a variable gain amplifier circuit.

[0023] The amplitude of the amplified signal is monitored in real time by a peak detection circuit;

[0024] The amplitude is compared with a preset target amplitude range to generate a feedback control signal;

[0025] The feedback control signal is used to dynamically adjust the amplification of the variable gain amplifier, and the amplitude of the output is controlled to be within the target amplitude range.

[0026] The second gain is a dynamically adjustable gain value determined by the feedback control signal.

[0027] According to the real-time acquisition method of the steam turbine speed provided by the application, the setting process of the initial size of the filter window comprises:

[0028] The rate of change of the steam turbine speed is monitored in real time, and the current working condition stage is determined according to the threshold range in which the rate of change of the speed is located, wherein the working condition stage comprises a starting stage, a shutdown stage and a rated operation stage.

[0029] According to the starting stage, the shutdown stage and the rated operation stage, the filter window size is selected from the corresponding value range.

[0030] According to the real-time acquisition method of the steam turbine speed provided by the application, the first speed value and the second speed value are calculated based on the filtered pulse signals, comprising:

[0031] In a fixed sampling time interval T, the rising edges of the two filtered pulse signals are counted respectively to obtain a first pulse count value N1 and a second pulse count value N2.

[0032] The number of teeth Z of a speed measuring gear on a steam turbine rotor is obtained.

[0033] The first speed value is calculated according to the formula RPM1=(60*N1) / (Z*T).

[0034] The second speed value is calculated according to the formula RPM2=(60*N2) / (Z*T).

[0035] According to the real-time acquisition method of the steam turbine speed provided by the application, the first speed value and the second speed value are weighted and averaged according to the fusion weight coefficients, and the final speed value is output, comprising:

[0036] The fusion weight coefficients K1 and K2 corresponding to the first speed value and the second speed value respectively output by the signal quality evaluation model are obtained.

[0037] The first speed value is multiplied by the weight coefficient K1 to obtain a first weighted result.

[0038] The second speed value is multiplied by the weight coefficient K2 to obtain a second weighted result.

[0039] The first weighted result and the second weighted result are added to obtain a weighted sum as the final speed value.

[0040] According to the application, a real-time turbine speed acquisition method is provided, and a device health state identifier is generated, including:

[0041] When the values of K1 and K2 are both higher than a preset first threshold, and the absolute difference between them is lower than a preset second threshold, a normal state identifier representing good signal consistency is outputted;

[0042] When one of the values of K1 and K2 is lower than the preset first threshold, or the absolute difference between them is higher than the preset second threshold but lower than a preset third threshold, a warning state identifier representing signal deviation is outputted;

[0043] When the values of K1 and K2 are both lower than the preset first threshold, or the absolute difference between them is higher than the preset third threshold, a fault state identifier representing serious unreliable signal is outputted.

[0044] According to the application, a real-time turbine speed acquisition method is provided, and selectively transmitting data batches to a remote cloud platform for storage, including:

[0045] When the identifier is a normal state, a low-frequency transmission mode is adopted, and only statistical feature data is transmitted to the remote cloud platform for storage;

[0046] When the identifier is a warning state, a medium-frequency transmission mode is adopted, and statistical features and part of the original data are transmitted to the remote cloud platform for storage;

[0047] When the identifier is a fault state, a high-frequency transmission mode is adopted, and complete original data and high-precision speed values are transmitted to the remote cloud platform for storage.

[0048] In a second aspect, the application provides a real-time turbine speed acquisition system, including:

[0049] An acquisition module is configured to synchronously acquire pulse signals generated by a turbine rotor through a Hall effect sensor and an eddy current sensor, and obtain two original pulse sequences;

[0050] A conditioning module is configured to condition the two original pulse sequences respectively, wherein a first gain is applied to the Hall effect sensor signal, and a second gain is applied to the eddy current sensor signal;

[0051] A filtering module is configured to perform sliding average filtering on the two conditioned pulse signals respectively using independent filtering windows, wherein the initial size of the filtering window is dynamically set according to the rated speed of the current working condition stage of the turbine;

[0052] A calculation module is configured to calculate a first rotation speed value and a second rotation speed value based on the filtered pulse signals respectively;

[0053] An evaluation module is configured to input the first rotation speed value and the second rotation speed value into a pre-trained signal quality evaluation model, and output a fusion weight coefficient for representing consistency of the two signals;

[0054] A weighting module is configured to perform weighted average calculation on the first rotation speed value and the second rotation speed value according to the fusion weight coefficient, and output a final rotation speed value and a device health status identifier;

[0055] A transmission module is configured to write the final rotation speed value into a local ring buffer, and selectively transmit a data batch to a remote cloud platform for storage according to the device health status identifier.

[0056] In a third aspect, the present application also provides an electronic device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements the steam turbine rotation speed real-time acquisition method according to any one of the above aspects when executing the program.

[0057] In a fourth aspect, the present application also provides a non-transitory computer readable storage medium, which stores a computer program, and the computer program is executable on a processor to implement the steam turbine rotation speed real-time acquisition method according to any one of the above aspects.

[0058] In a fifth aspect, the present application also provides a computer program product, which comprises a computer program, and the computer program is executable on a processor to implement the steam turbine rotation speed real-time acquisition method according to any one of the above aspects.

[0059] The application provides a turbine rotating speed real-time acquisition method and system, comprising the following steps: synchronously collecting pulse signals generated by a turbine rotor through a Hall effect sensor and an eddy current sensor to obtain two original pulse sequences; respectively performing signal conditioning on the two original pulse sequences, wherein a first gain is applied to the Hall effect sensor signal and a second gain is applied to the eddy current sensor signal; respectively performing sliding average filtering on the two pulse signals after conditioning through independent filtering windows; wherein the initial size of the filtering window is dynamically set according to the rated rotating speed of the current working condition stage of the turbine; based on the pulse signals after filtering, a first rotating speed value and a second rotating speed value are respectively calculated; the first rotating speed value and the second rotating speed value are input into a pre-trained signal quality evaluation model to output a fusion weight coefficient for representing the consistency of the two signals; the first rotating speed value and the second rotating speed value are weighted and averaged according to the fusion weight coefficient to output a final rotating speed value and a device health state identifier; the final rotating speed value is written into a local ring buffer, and selectively batch data is transmitted to a remote cloud platform for storage according to the device health state identifier. By introducing the pre-trained signal quality evaluation model, the quality and reliability of the output data of the Hall effect sensor and the eddy current sensor can be intelligently controlled, and a dynamic fusion weight coefficient is generated for weighted averaging, effectively suppressing the measurement error caused by the performance degradation or sudden interference of a single sensor. The initial size of the filtering window is dynamically set according to the rated rotating speed of the current working condition stage of the turbine, so that the signal processing parameters can be matched with the actual operating state of the unit, and the measurement accuracy in the whole working condition range is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0060] In order to more clearly illustrate the technical solutions in the application or prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description are some embodiments of the application, and other drawings can be obtained by those skilled in the art without creative effort.

[0061] Figure 1 is a flowchart of the turbine rotating speed real-time acquisition method provided by the embodiment;

[0062] Figure 2 is a structural schematic diagram of the turbine rotating speed real-time acquisition system provided by the embodiment;

[0063] Figure 3 is a structural schematic diagram of the electronic device provided by the embodiment. DETAILED DESCRIPTION

[0064] In order to make the objects, technical solutions and advantages of the present application clearer, the technical solutions in the present application will be described clearly and completely below in conjunction with the drawings in the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative labor fall within the protection scope of the present application.

[0065] Figure 1 FIG. 1 is a flowchart of a real-time turbine speed acquisition method provided by the present embodiment.

[0066] As shown in FIG. 1, the real-time turbine speed acquisition method provided by the present embodiment mainly includes the following steps. Figure 1

[0067] 101, synchronously acquire the pulse signals generated by the turbine rotor by the Hall effect sensor and the eddy current sensor, and obtain two original pulse sequences.

[0068] Specifically, the Hall effect sensor and the eddy current sensor are fixedly installed along the circumference of the turbine rotor at a preset included angle (30-60 degrees). Taking the rotor magnetic mark (such as gear tooth top / tooth groove) as a reference, the two sensors are ensured to stably capture the mark change and output pulse signals. Due to the installation included angle, the two sensor pulse sequences form a predictable phase difference when the rotor rotates, which not only provides a basis for subsequent signal processing, but also reduces the risk of single sensor failure through double sensor redundancy, and improves the anti-interference ability of the system.

[0069] Start the rotor, acquire the initial pulse signals of the two sensors, analyze the actual phase difference and compare it with the preset target value. If the phase difference deviates from the preset range, the circumferential installation position of the sensor is adjusted, and the acquisition and comparison operations are repeated until the phase difference is stably in the predictable interval, so as to eliminate the phase deviation caused by the installation error and provide a reliable reference for subsequent phase compensation.

[0070] In the speed acquisition processing unit, the calibrated predictable phase difference data is imported, the inherent phase delay characteristics are analyzed in combination with the differences between the working principles of the two sensors, the phase compensation logic is designed and the corresponding parameters are configured. The phase of the pulse signal is corrected in real time by the parameters to eliminate the deviation, so as to realize accurate synchronization of the signal, reduce the speed measurement error, reduce the risk of misoperation of the overspeed protection system, break away from the dependence on foreign technology, help the independent controllability of key technologies, improve the reliability of the equipment and reduce the maintenance cost.

[0071] ​After completing the sensor installation and adaptation, connect the Hall effect sensor and the eddy current sensor to their respective input channels, ensuring that the wiring connections are secure and avoid areas with strong electromagnetic interference to reduce signal transmission noise. At the same time, configure the basic parameters of the acquisition module according to the working characteristics of the two sensors (Hall effect relies on magnetic field, and eddy current relies on electromagnetic induction) to adapt it to the sensor signal type. This ensures the quality of the original signal and reduces the risk of acquisition interruption due to single sensor failure through dual sensor redundancy.

[0072] Using the magnetic markings on the rotor surface as a unified reference, the sensor position is adjusted so that the detection end is directly aligned with the marking's movement trajectory and the axial detection area coincides, ensuring that both detect the marking changes during the same rotation cycle. After installation, the rotor is manually rotated or the turbine is started at low speed to fine-tune until the sensor outputs a stable and effective signal, laying the foundation for synchronous data acquisition.

[0073] After the steam turbine is running normally, a unified acquisition trigger condition and frequency are set to achieve synchronous acquisition from the time dimension; the software receives and stores the pulse signals output by the two sensors in real time to form an independent original pulse sequence.

[0074] 102. Perform signal conditioning on the two original pulse sequences respectively, wherein a first gain is applied to the Hall effect sensor signal and a second gain is applied to the eddy current sensor signal.

[0075] Specifically, the raw pulse voltage signal output from the Hall effect sensor is fed into a fixed-gain inverting amplifier circuit. The signal is amplified using a preset fixed gain (pre-set and maintained based on the sensor's rated output level and the analog-to-digital converter's range), increasing the signal amplitude to a level suitable for the analog-to-digital converter's processing. The amplified signal is then fed into a clamping circuit composed of bidirectional Zener diodes. The reverse breakdown characteristic of the diodes limits the peak signal voltage to within the allowable input voltage range of the analog-to-digital converter. During this process, the fixed gain ensures stable signal amplification, preventing signal distortion introduced by gain fluctuations; the clamping process effectively prevents overvoltage signals from damaging subsequent analog-to-digital conversion components, while ensuring the input signal amplitude remains within the effective range, improving signal stability and circuit safety, and providing a high-quality voltage signal for accurate analog-to-digital conversion.

[0076] The raw amplitude-modulated AC signal output from the eddy current sensor is fed into a variable gain amplifier circuit. A peak detection circuit monitors the amplitude of the amplified signal in real time. The detected amplitude is compared to a preset target amplitude range. If it deviates from the range, a feedback control signal is generated to dynamically adjust the amplification factor of the variable gain amplifier (i.e., the second gain changes in real time with the feedback signal), stabilizing the output signal amplitude within the target range. This dynamic gain adjustment mechanism adapts to signal amplitude fluctuations caused by rotor clearance changes and environmental interference from the eddy current sensor, ensuring that the output signal amplitude is always within the optimal processing range of the analog-to-digital converter. This avoids signal-to-noise ratio reduction due to excessively weak signals or saturation distortion due to excessively strong signals, improving the environmental adaptability and signal quality of signal conditioning, and providing a reliable AC signal foundation for subsequent data processing.

[0077] Through differentiated conditioning, both signals can adapt to analog-to-digital conversion requirements. The Hall effect sensor signal achieves stable and safe transmission with fixed gain and clamping, while the eddy current sensor signal achieves amplitude stability through dynamic gain adjustment. The two work together to improve the speed acquisition system's adaptability to different types of sensor signals and signal processing accuracy, reduce speed detection errors caused by signal quality issues, and enhance the overall reliability of the system.

[0078] 103. For the two pulse signals after conditioning, separate filtering windows are used for moving average filtering; the initial size of the filtering window is dynamically set according to the rated speed of the turbine at the current operating stage.

[0079] Specifically, independent filtering windows are configured for the two pulse signals after gain conditioning. For each signal, the system maintains a data buffer as a filtering window in real time. When a new pulse signal sample value is input, the window slides forward in chronological order, removing the earliest sample value in the window and incorporating the latest sample value. Then, the average value of all sample values ​​in the current window is calculated, and this average value is used as the filtered output value at the corresponding time.

[0080] After being filtered through its dedicated window, the Hall effect sensor signal effectively smooths out spike noise caused by electromagnetic interference while preserving the basic contour of the pulse signal, avoiding the influence of fixed noise on pulse edge detection. The eddy current sensor signal, after being processed by its independent window, can specifically suppress signal amplitude fluctuations caused by rotor vibration, reducing the interference of random noise on pulse period stability. The independent filtering design of the two signals adapts to their respective signal characteristics, avoiding the problems of over-smoothing (loss of effective pulse information) or insufficient filtering (noise residue) caused by sharing filtering parameters. The final output of the two filtered signals has a more regular waveform and lower noise, providing high-quality input for subsequent pulse counting, phase comparison, and other processing, improving the stability and accuracy of speed measurement.

[0081] The setting process of the initial size of the filter window includes: monitoring the change rate of the turbine speed in real time, and calculating the change amount of the speed in a unit time. A plurality of threshold ranges of the change rate of the speed are preset, and correspond to the start-up stage (rapid rise of the speed), the shutdown stage (rapid drop of the speed) and the rated operation stage (stable speed and small change rate) of the turbine respectively. The real-time calculated change rate of the speed is compared with the preset threshold range to determine the working condition stage of the current turbine.

[0082] For different working condition stages, the corresponding filter window size value range is set in advance: in the start-up and shutdown stages, the window size in a smaller value range is selected to reduce the filtering delay, ensure that the filtered signal can quickly respond to the dramatic change of the speed, and avoid that the signal lags behind the actual speed change due to the too large window; in the rated operation stage, the window size in a larger value range is selected to enhance the filtering effect, sufficiently smooth the signal noise caused by the slight speed fluctuation, and improve the signal stability.

[0083] Through the above process, the initial size of the filter window can be dynamically adapted according to the real-time working condition of the turbine, which not only ensures the signal response speed when the speed changes dramatically, but also ensures the signal smoothness when the turbine operates stably, avoiding the problems of insufficient filtering or excessive filtering of the fixed window size under complex working conditions, providing more accurate input for subsequent pulse signal processing, and further improving the dynamic adaptability and overall precision of the speed measurement.

[0084] 104、Based on the filtered pulse signal, the first speed value and the second speed value are calculated respectively.

[0085] Specifically, a fixed sampling time interval is set, and the rising edge detection and counting of the filtered Hall effect sensor pulse signal and the electric eddy current sensor pulse signal are performed in the interval to obtain the first pulse count value and the second pulse count value. By identifying the rising edge of the pulse signal as the counting reference, the counting error caused by the fluctuation of the signal amplitude can be reduced, and the counting result can reflect the true pulse number.

[0086] The number of teeth of the speed measurement gear on the turbine rotor is obtained (this parameter is a fixed mechanical parameter preset by the system), which is used as a basic mechanical reference for speed calculation. A unified calculation formula is used to respectively substitute the first pulse count value and the second pulse count value into the calculation: 60 multiplied by the pulse count value, and then divided by the product of the number of teeth and the sampling time interval, to obtain the first speed value and the second speed value of the corresponding sensor. For example, formulas (1) and (2):

[0087] RPM1=(60*N1) / (Z*T) (1)

[0088] RPM2=(60*N2) / (Z*T) (2)

[0089] Wherein, RPM1 represents a first rotation speed value, N1 represents a first pulse count value, RPM2 represents a second rotation speed value, N2 represents a second pulse count value, Z represents the number of teeth of a speed measuring gear, and T represents a sampling time interval.

[0090] The fixed sampling interval ensures the time reference consistency of the two rotation speed values, facilitating subsequent comparison and fusion; the rising edge counting method improves the anti-interference ability of pulse recognition; and the standardized formula based on the number of gear teeth directly links the pulse signal and the actual mechanical rotation speed, realizing accurate conversion of electrical signals to physical quantities. The two independently calculated rotation speed values not only provide redundant measurement data, but also can eliminate abnormal values through mutual verification, effectively reducing the rotation speed measurement deviation caused by single sensor failure or signal interference, and providing reliable quantitative basis for accurate monitoring and safe control of the steam turbine rotation speed.

[0091] 105. inputting the first rotation speed value and the second rotation speed value into a pre-trained signal quality evaluation model, and outputting a fusion weight coefficient for representing consistency of the two signals.

[0092] Specifically, first, a signal quality evaluation model is constructed and pre-trained. The model training data uses the first rotation speed value and the second rotation speed value samples collected by the steam turbine under various working conditions (including normal operation, noise interference, sensor slight failure, etc.), and the consistency level (such as synchronism, deviation range, etc.) of the two signals in the labeled sample is used as the training label; the model is trained through a machine learning algorithm to learn the mapping relationship between the two rotation speed values and the signal consistency, and after training, the model has the ability to evaluate the signal quality according to the input real-time rotation speed value.

[0093] In real-time operation, the calculated first rotation speed value and the second rotation speed value are input into the pre-trained signal quality evaluation model. The model analyzes the change trend, numerical deviation range and other characteristics of the two rotation speed values to evaluate the consistency degree of the two signals at the current time. If the two rotation speed values change synchronously and the deviation is within a reasonable range, the consistency is high; if there is significant asynchrony or deviation out of limit, the consistency is low.

[0094] According to the evaluation result, the model outputs the corresponding fusion weight coefficient. When the consistency is high, the weight distribution of the two signals is balanced; when the consistency of a certain signal is low (such as deviating from the trend of the other signal), its weight is reduced, and the weight of the other signal is correspondingly increased.

[0095] The dynamically output weight coefficient can quantitatively reflect the signal reliability, avoiding error accumulation caused by fixed weight fusion; by giving higher weight to high-quality signals, more reliable weighted basis is provided for subsequent rotation speed fusion results, improving the anti-interference ability and accuracy of rotation speed measurement.

[0096] 106. According to the fusion weight coefficient, the first rotation speed value and the second rotation speed value are weighted and averaged to output a final rotation speed value and a device health status identifier.

[0097] Specifically, the fusion weight coefficients K1 (corresponding to the first rotation speed value) and K2 (corresponding to the second rotation speed value) output by the signal quality evaluation model are obtained, and a weighted average algorithm is used for fusion calculation. Multiply the first rotation speed value by K1 to get the first weighted result, multiply the second rotation speed value by K2 to get the second weighted result, and the sum of the two is the final rotation speed value.

[0098] Through dynamic weight distribution, the rotation speed value with high reliability occupies a higher proportion in the fusion result, which not only retains the advantages of dual-sensor redundant measurement, but also suppresses the interference of low-quality signals through differentiated weights, improving the accuracy and anti-interference ability of the final rotation speed value.

[0099] When K1 and K2 are both higher than the first threshold and the difference between them is lower than the second threshold, a normal state identifier is output, indicating that the consistency of the two signals is good and the system is running stably. When either weight is lower than the first threshold, or the weight difference is between the second and third thresholds, a warning state identifier is output, indicating that the signal has an abnormal deviation and needs attention. When both weights are lower than the first threshold, or the weight difference is higher than the third threshold, a fault state identifier is output, indicating that the signal is severely unreliable and needs emergency treatment.

[0100] By quantifying the weight feature to evaluate the device health status, accurate mapping from signal quality to device status is achieved, providing a graded warning basis for operation and maintenance, which helps to timely discover potential faults of sensors or signal processing links, and improves the safety and maintainability of the system.

[0101] 107. The final rotation speed value is written into the local ring buffer, and selectively transmitted to the remote cloud platform for storage according to the device health status identifier.

[0102] Specifically, the final rotation speed value calculated is written into the local ring buffer in real time to realize temporary storage and cyclic coverage of data, ensuring that critical data is not lost and saving local storage resources.

[0103] At the same time, according to the device health status identifier, a differentiated strategy is used to transmit data batches to the remote cloud platform:

[0104] When the identifier is in a normal state, a low-frequency transmission mode is enabled, and only the statistical features of the rotation speed data (such as mean, peak, etc.) are sent to the cloud platform. This greatly reduces the data transmission volume, reduces the network bandwidth occupation and the storage pressure of the cloud platform while ensuring the state monitoring requirements.

[0105] When identified as a warning state, switch to a medium frequency transmission mode, in addition to statistical features, additional transmission of part of the original pulse signal data. Both retain key data for anomaly analysis, and avoid resource waste caused by full volume transmission, balance the fault diagnosis demand and transmission efficiency.

[0106] When identified as a fault state, automatically enable a high frequency transmission mode, upload complete original data and high precision speed value to the cloud platform. Ensure that the detailed data at the fault moment is recorded completely, provide comprehensive basis for subsequent fault tracing and cause analysis, and improve problem troubleshooting efficiency.

[0107] Through the state-driven hierarchical transmission mechanism, intelligent control of data storage and transmission is realized. Under normal state, optimize resource occupation, under abnormal state, guarantee data integrity, meet the basic needs of remote monitoring, provide accurate data support for fault diagnosis, and reduce the overall operation cost of the system.

[0108] Based on the same overall inventive concept, the present application also protects a steam turbine speed real-time acquisition system, which can be mutually corresponding and referred to the steam turbine speed real-time acquisition method described below.

[0109] Figure 2 It is a structural schematic diagram of the steam turbine speed real-time acquisition system provided by the embodiment.

[0110] As shown in Figure 2 , the steam turbine speed real-time acquisition system provided by the embodiment includes:

[0111] The acquisition module 201 is used for synchronously acquiring the pulse signals generated by the steam turbine rotor through the Hall effect sensor and the eddy current sensor, and obtaining two original pulse sequences;

[0112] The conditioning module 202 is used for respectively conditioning the two original pulse sequences, wherein the Hall effect sensor signal is subjected to a first gain, and the eddy current sensor signal is subjected to a second gain;

[0113] The filtering module 203 is used for respectively performing sliding average filtering on the two conditioned pulse signals by using independent filtering windows; wherein the initial size of the filtering window is dynamically set according to the rated speed of the current working condition stage of the steam turbine;

[0114] The calculation module 204 is used for respectively calculating the first speed value and the second speed value based on the filtered pulse signals;

[0115] The evaluation module 205 is used for inputting the first speed value and the second speed value into a pre-trained signal quality evaluation model, and outputting a fusion weight coefficient for representing the consistency of the two signals;

[0116] The weighting module 206 is used to calculate the weighted average of the first speed value and the second speed value according to the fusion weighting coefficient, and output the final speed value and the equipment health status indicator.

[0117] The transmission module 207 is used to write the final rotation speed value into a local circular buffer and selectively transmit data in batches to a remote cloud platform for storage based on the device health status identifier.

[0118] Figure 3 This is a schematic diagram of the structure of the electronic device provided in this embodiment.

[0119] like Figure 3 As shown, the electronic device may include a processor 301, a communication interface 302, a memory 303, and a communication bus 304. The processor 301, communication interface 302, and memory 303 communicate with each other via the communication bus 304. The processor 301 can call logical instructions from the memory 303 to execute a method for real-time acquisition of turbine speed.

[0120] Furthermore, the logical instructions in the aforementioned memory 303 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0121] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the real-time turbine speed acquisition method provided by the above methods.

[0122] In another aspect, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the real-time turbine speed acquisition method provided by the above methods.

[0123] The apparatus embodiments described above are merely illustrative, wherein the units described as separated components can or can not be physically separated, and the components displayed as units can or can not be physical units, i.e., can be located in one place, or can be distributed to multiple network units. Part or all of the modules can be selected to achieve the purpose of the embodiment scheme according to actual needs. Those skilled in the art can understand and implement without creative labor.

[0124] Through the description of the above embodiments, those skilled in the art can clearly understand that the embodiments can be realized by means of software and the necessary general hardware platform, and of course can also be realized by hardware. Based on such understanding, the above technical solutions can be embodied in the form of a software product, which can be stored in a computer readable storage medium, such as a ROM / RAM, a magnetic disk, an optical disk, etc., and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) execute the methods described in each embodiment or some parts of the embodiments.

[0125] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for real-time acquisition of the rotating speed of a steam turbine, characterized in that, The method comprises the following steps: Synchronously collecting pulse signals generated by a turbine rotor through a Hall effect sensor and an eddy current sensor to obtain two original pulse sequences; Signal conditioning is performed on the two original pulse sequences respectively, wherein a first gain is applied to the Hall effect sensor signal and a second gain is applied to the eddy current sensor signal; Sliding average filtering is performed on the two conditioned pulse signals respectively using independent filter windows; wherein the initial size of the filter window is dynamically set according to the rated speed of the current operating stage of the turbine; Based on the filtered pulse signals, a first speed value and a second speed value are calculated respectively; The first speed value and the second speed value are input into a pre-trained signal quality evaluation model to output a fusion weight coefficient for representing the consistency of the two signals; According to the fusion weight coefficient, the first speed value and the second speed value are weighted and averaged to output a final speed value and a device health status identifier; The final speed value is written to a local ring buffer, and selectively batch data is transmitted to a remote cloud platform for storage according to the device health status identifier.

2. The turbine speed real-time acquisition method of claim 1, wherein, Before synchronously collecting pulse signals generated by a turbine rotor through a Hall effect sensor and an eddy current sensor, the method further comprises the following steps: The Hall effect sensor and the eddy current sensor are installed at a preset angle along the circumference of the rotor, the installation positions of the Hall effect sensor and the eddy current sensor are calibrated, and the pulse sequence generated when the rotor rotates is controlled to have a predictable phase difference; According to the phase difference, corresponding phase compensation parameters are configured in the signal processing unit.

3. The turbine speed real-time acquisition method of claim 1, wherein, The first gain applied to the Hall effect sensor signal comprises the following steps: The original pulse voltage signal output by the Hall effect sensor is connected to a fixed-gain in-phase amplifier circuit for amplification; A bidirectional voltage stabilizing diode is used to clamp the amplified signal to limit the peak voltage within the input voltage range allowed by the subsequent analog-to-digital converter; Wherein, the first gain is a fixed gain value that is pre-set according to the rated output level of the Hall effect sensor and the range of the analog-to-digital converter and remains unchanged.

4. The turbine speed real-time acquisition method of claim 1, wherein, The second gain applied to the eddy current sensor signal comprises the following steps: The original amplitude-modulated alternating current signal output by the eddy current sensor is connected to a variable-gain amplifier circuit; The amplitude of the amplified signal is monitored in real time by a peak detection circuit; The amplitude is compared with a preset target amplitude range to generate a feedback control signal; The feedback control signal is used to dynamically adjust the amplification factor of the variable-gain amplifier to control the output amplitude to be stable within the target amplitude range; Wherein, the second gain is a dynamically adjustable gain value determined by the feedback control signal.

5. The turbine speed real-time acquisition method of claim 1, wherein, The setting process of the initial size of the filter window comprises the following steps: Real-time monitoring of the turbine speed change rate determines the current operating stage according to the threshold range of the speed change rate, wherein the operating stage includes the start-up stage, the shutdown stage and the rated running stage; According to the start-up stage, the shutdown stage and the rated running stage, the filter window size is selected from the corresponding value range.

6. The turbine speed real-time acquisition method of claim 1, wherein, The first rotation speed value and the second rotation speed value are respectively calculated based on the filtered pulse signals, and the calculation includes: In a fixed sampling time interval T, rising edge counting is respectively performed on the two filtered pulse signals to obtain a first pulse count value N1 and a second pulse count value N2; The number of teeth Z of a speed measuring gear on a steam turbine rotor is acquired; The first rotation speed value is calculated according to a formula RPM1=(60*N1) / (Z*T); The second rotation speed value is calculated according to a formula RPM2=(60*N2) / (Z*T).

7. The turbine speed real-time acquisition method of claim 1, wherein, The first rotation speed value and the second rotation speed value are weighted and averaged according to the fusion weight coefficients to output a final rotation speed value, and the calculation includes: The fusion weight coefficients K1 and K2 corresponding to the first rotation speed value and the second rotation speed value respectively output by the signal quality evaluation model are acquired; The first rotation speed value is multiplied by the weight coefficient K1 to obtain a first weighted result; The second rotation speed value is multiplied by the weight coefficient K2 to obtain a second weighted result; The first weighted result and the second weighted result are added to obtain a weighted sum as the final rotation speed value.

8. The turbine speed real-time acquisition method of claim 7, wherein, The generation process of the device health state identifier includes: When the values of K1 and K2 are both higher than a preset first threshold value, and the absolute difference between them is lower than a preset second threshold value, a normal state identifier representing good signal consistency is output; When one of the values of K1 and K2 is lower than the preset first threshold value, or the absolute difference between them is higher than the preset second threshold value but lower than a preset third threshold value, a warning state identifier representing signal deviation is output; When the values of K1 and K2 are both lower than the preset first threshold value, or the absolute difference between them is higher than the preset third threshold value, a fault state identifier representing serious unreliable signal is output.

9. The turbine speed real-time acquisition method of claim 8, wherein, The selective transmission of the data batch to the remote cloud platform for storage includes: When the identifier is in the normal state, a low-frequency transmission mode is adopted, and only statistical feature data is sent to the remote cloud platform for storage; When the identifier is in the warning state, a medium-frequency transmission mode is adopted, and statistical features and part of the original data are sent to the remote cloud platform for storage; When the identifier is in the fault state, a high-frequency transmission mode is adopted, and complete original data and high-precision rotation speed values are sent to the remote cloud platform for storage.

10. A real-time acquisition system of a steam turbine rotation speed, characterized in that, It includes: The acquisition module is configured to synchronously acquire pulse signals generated by the steam turbine rotor through the Hall effect sensor and the eddy current sensor to obtain two original pulse sequences; The conditioning module is configured to perform signal conditioning on the two original pulse sequences respectively, wherein a first gain is applied to the Hall effect sensor signal, and a second gain is applied to the eddy current sensor signal; The filtering module is configured to perform sliding average filtering on the two conditioned pulse signals respectively using independent filtering windows, wherein the initial size of the filtering window is dynamically set according to the rated rotation speed of the current working condition stage of the steam turbine; The calculation module is configured to calculate the first rotation speed value and the second rotation speed value based on the filtered pulse signals. An evaluation module is configured to input the first rotation speed value and the second rotation speed value into a pre-trained signal quality evaluation model, and output a fusion weight coefficient for representing consistency of the two signals; A weighting module is configured to perform weighted average calculation on the first rotation speed value and the second rotation speed value according to the fusion weight coefficient, and output a final rotation speed value and a device health status identifier; A transmission module is configured to write the final rotation speed value into a local ring buffer, and selectively transmit a data batch to a remote cloud platform for storage according to the device health status identifier.

Images