Slurry circulating pump state monitoring method and related equipment

Abstract

The invention discloses a slurry circulating pump state monitoring method and related equipment, and aims to overcome the defects of high false alarm rate and late early warning caused by multi-fault coupling effect and strong background noise interference in the prior art. The method comprises the following steps: synchronously acquiring vibration, acoustic emission, pressure pulsation and temperature signals of a pump volute, an outlet pipeline and a bearing seat; performing wavelet packet decomposition on the acoustic emission signals, extracting features and generating cavitation sensitive factors; a digital twinborn model adapted to fluid-structure coupling simulation and transfer learning is utilized to calculate a deviation value between the monitoring feature and a normal baseline; and when the deviation value exceeds a threshold value, fault identification and life prediction are carried out. Through multi-signal fusion and a dynamic threshold mechanism, a multi-fault coupling effect and strong background noise interference are effectively overcome, and the reliability of a monitoring system is improved.

Description

Technical Field

[0001] The present invention relates to the field of industrial equipment condition monitoring, and particularly to a method for monitoring the condition of a slurry circulation pump and related equipment. Background Art

[0002] Currently, the commonly used technical means for monitoring slurry circulation pumps mainly include methods based on vibration signal analysis, temperature monitoring, and pressure detection. Vibration signal analysis involves installing acceleration sensors or velocity sensors at key parts of the pump body to collect vibration data, and using signal processing techniques such as Fourier transform (FFT) to identify characteristic frequencies and determine whether there are imbalances, misalignments, bearing failures, etc. in the equipment. Temperature monitoring is to use sensors such as thermocouples or thermal resistors to measure the temperature changes of parts such as the pump body bearings and motors in real time, so as to indirectly reflect whether the operating state of the equipment is abnormal. Pressure detection mainly sets pressure sensors in the inlet and outlet pipelines of the pump to monitor the pressure fluctuation situation and provide data support for the performance evaluation of the pump.

[0003] However, these existing technologies have exposed many problems in practical applications. On the one hand, relying solely on vibration signal analysis is difficult to cope with the multi-fault coupling effect of slurry circulation pumps. During the operation of slurry circulation pumps, multiple faults such as cavitation, bearing wear, and seal leakage may occur simultaneously. These faults affect and intertwine with each other, making the vibration signal characteristics complex and not obvious. For example, the vibration characteristic frequency caused by cavitation may be close to the frequency generated by early bearing wear, resulting in misjudgment. On the other hand, the strong background noise in the industrial site seriously interferes with the accuracy of the monitoring signal. The operating environment of the slurry circulation pump is noisy, and the electromagnetic interference of surrounding equipment, the turbulent noise of the fluid in the pipeline, and the background vibration of mechanical operation will cover up the fault characteristic signal. Taking acoustic emission signal monitoring as an example, the weak acoustic emission signal generated by early cavitation is often submerged by strong noise, making the monitoring system unable to capture the early signs of the fault in time. It can only detect when the fault develops to a certain extent and the characteristic signal is strong enough, thus delaying the maintenance time and increasing the risk of sudden equipment failure, posing a threat to production continuity and safety.

[0004] In summary, in the face of the complex working conditions of multi-fault coupling and strong noise interference of slurry circulation pumps, there is an urgent need for a monitoring technical solution that can accurately identify the fault type and early warn of equipment hidden dangers. Summary of the Invention

[0005] The purpose of the present invention is to provide a method for monitoring the condition of a slurry circulation pump and related equipment to overcome the deficiencies of high false alarm rate and late warning caused by the multi-fault coupling effect and strong background noise interference of the existing technology.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions:

[0007] In a first aspect, the present invention provides a method for monitoring the state of a slurry circulation pump, including:

[0008] In a triangular monitoring topology formed by the pump body volute, the outlet pipe, and the bearing housing, synchronously collect vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals;

[0009] Perform wavelet packet decomposition on the acoustic emission signal, extract the energy ratio feature in the frequency band of 30 - 50 kHz, and fuse the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor;

[0010] Input the cavitation sensitivity factor, vibration signal features, and temperature signal features into a digital twin model established through fluid - structure coupling simulation and adapted through transfer learning to obtain the deviation amount between the real - time monitoring features and the normal operating condition baseline;

[0011] When the deviation amount exceeds the dynamic threshold, trigger fault mode recognition and remaining life prediction.

[0012] In a triangular monitoring topology formed by the pump body volute, the outlet pipe, and the bearing housing, synchronously collect vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals, where the acoustic emission sensor uses a PZT piezoelectric ceramic sensor with a resonant frequency of 80 kHz, and the vibration signal is collected through a MEMS tri - axis accelerometer.

[0013] The triangular monitoring topology satisfies:

[0014] The axial distance between the volute measuring point and the outlet pipe measuring point is 1.2 - 1.5 times the pump outlet pipe diameter;

[0015] The measuring point on the bearing housing forms an angle of 60° ± 5° with the connection line between the volute and the outlet pipe.

[0016] Perform wavelet packet decomposition on the acoustic emission signal, extract the energy ratio feature in the frequency band of 30 - 50 kHz, and fuse the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor, where the wavelet packet decomposition uses a db8 wavelet basis for 5 - layer decomposition, and the calculation formula for the cavitation sensitivity factor is:

[0017]

[0018] where α takes a value of 0.6 - 0.8.

[0019] The cavitation sensitivity factor further fuses the pressure pulsation feature:

[0020]

[0021] where β takes a value of 0.2 - 0.4, and σ(P) is the standard deviation of the pressure pulsation signal.

[0022] Input the cavitation sensitivity factor, vibration signal characteristics, and temperature signal characteristics into the digital twin model established through fluid-structure coupling simulation and adapted through transfer learning, and obtain the deviation between the real-time monitoring characteristics and the normal operating condition baseline. The transfer learning adaptation uses a domain adversarial neural network, and the digital twin model generates the normal operating condition baseline through fluid-structure coupling simulation.

[0023] When the deviation exceeds the dynamic threshold, trigger fault mode recognition and remaining life prediction. The remaining life prediction uses an attention mechanism LSTM model, and the input data includes the cavitation sensitivity factor, the kurtosis value of the shaft envelope spectrum of the vibration signal, and the third derivative of the temperature change rate.

[0024] In a second aspect, the present invention provides a slurry pump status monitoring system, including:

[0025] A data acquisition module, configured to synchronously acquire vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals in the triangular monitoring topology composed of the pump body volute, the outlet pipe, and the bearing housing;

[0026] A signal decomposition module, configured to perform wavelet packet decomposition processing on the acoustic emission signal, extract the energy ratio feature in the 30 - 50 kHz frequency band, and fuse the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor;

[0027] A deviation acquisition module, configured to input the cavitation sensitivity factor, vibration signal characteristics, and temperature signal characteristics into the digital twin model established through fluid-structure coupling simulation and adapted through transfer learning, and obtain the deviation between the real-time monitoring characteristics and the normal operating condition baseline;

[0028] A response module, configured to trigger fault mode recognition and remaining life prediction when the deviation exceeds the dynamic threshold.

[0029] In a third aspect, the present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the slurry pump status monitoring method described above are implemented.

[0030] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the steps of the slurry pump status monitoring method described above are implemented.

[0031] Compared with the prior art, the present invention has the following beneficial technical effects:

[0032] In a first aspect, the present invention provides a method for monitoring the state of a slurry circulation pump. In a triangular monitoring topology formed by the pump body volute, the outlet pipe, and the bearing housing, vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals are synchronously collected. Through this multi-dimensional signal acquisition method, various state information of the slurry circulation pump during operation can be comprehensively captured, effectively covering complex working conditions generated by the multi-fault coupling effect. At the same time, the wavelet packet decomposition processing technology is applied to the acoustic emission signal to accurately extract the energy ratio feature in the frequency band of 30 - 50 kHz, and the standard deviation of the pressure pulsation signal is skillfully fused to generate a cavitation sensitivity factor. This process greatly improves the sensitivity to cavitation phenomena, can accurately identify fault feature signals from strong background noise, and effectively reduces the false alarm rate. Further, the cavitation sensitivity factor, vibration signal features, and temperature signal features are input into a digital twin model established through fluid-structure coupling simulation and adapted by transfer learning. This model can accurately calculate the deviation amount of real-time monitoring features based on the normal working condition baseline, realizing the precise assessment of the equipment state. When the deviation amount exceeds the dynamic threshold, the fault mode recognition and remaining life prediction functions are triggered in a timely manner. Compared with the prior art, the present invention can provide earlier warning of faults, providing sufficient time for maintenance personnel to take measures, avoiding the risks of sudden equipment failures and production interruptions caused by late warnings, and significantly improving the reliability and timeliness of the state monitoring of the slurry circulation pump.

[0033] Second aspect, the present invention provides a slurry circulation pump condition monitoring system, which overcomes the problems of high false alarm rate and late warning caused by the multi-fault coupling effect and strong background noise interference in the prior art through a data acquisition module, a signal decomposition module, a deviation acquisition module and a response module. The data acquisition module synchronously acquires vibration, acoustic emission, pressure pulsation and temperature signals at key parts of the pump body, ensuring the comprehensiveness and accuracy of the monitoring data. The synchronous acquisition of such multi-source signals provides a solid data basis for subsequent signal processing and analysis, helping to identify fault-related features from complex signals. The signal decomposition module extracts the energy ratio feature in the 30 - 50 kHz frequency band of the acoustic emission signal through wavelet packet decomposition technology, which is usually closely related to the cavitation phenomenon. At the same time, this module also fuses the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor. This feature extraction and fusion method can significantly improve the recognition ability of fault phenomena such as cavitation, while reducing the interference of background noise. The deviation acquisition module uses the digital twin model established by fluid-structure coupling simulation, and adapts to the actual monitoring data through transfer learning, so as to obtain the deviation amount between the real-time monitoring feature and the normal working condition baseline. The use of the digital twin model provides a highly simulated reference benchmark for the monitoring system, making the calculation of the deviation amount more accurate, thus improving the reliability of fault detection. The response module triggers fault mode recognition and remaining life prediction when the deviation amount exceeds the dynamic threshold. The setting of the dynamic threshold takes into account the real-time change characteristics of the monitoring data, reducing false alarms caused by fixed thresholds. At the same time, timely fault recognition and life prediction help to achieve early warning and avoid equipment failures under unforeseen circumstances.

[0034] Third aspect, the present invention provides a computer device, which can efficiently implement the steps of the method of the present invention by a processor executing specific computer programs. When the computer device executes data processing tasks, it can accurately perform numerical calculations and logical judgments, avoiding errors caused by human factors; at the same time, because the computer program has high stability and reliability, the accuracy and consistency of the data processing results can be ensured.

[0035] Fourth aspect, the present invention provides a computer-readable storage medium. By programming the steps of the method of the present invention into computer programs and storing them on the computer-readable storage medium, users can easily load these programs onto any compatible computer device and execute them without having to rewrite or convert the code, greatly improving the convenience and flexibility of program execution. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a schematic diagram of a slurry circulation pump condition monitoring method in an embodiment of the present invention.

[0037] Figure 2Schematic diagram of a slurry circulation pump status monitoring system in an embodiment of the present invention. Detailed implementation

[0038] In the field of industrial equipment status monitoring, especially for the monitoring of slurry circulation pumps, the commonly used technical means mainly include methods such as vibration signal analysis, temperature monitoring, and pressure detection. Vibration signal analysis installs acceleration sensors or velocity sensors at key parts of the pump body to collect vibration data, and uses signal processing techniques such as Fourier transform (FFT) to identify characteristic frequencies to judge whether there are imbalances, misalignments, bearing failures, etc. in the equipment. Temperature monitoring is to use sensors such as thermocouples or thermal resistors to measure the temperature changes of the pump body bearings, motors, etc. in real time, so as to indirectly reflect whether the operating state of the equipment is abnormal. Pressure detection mainly sets pressure sensors in the inlet and outlet pipelines of the pump to monitor the pressure fluctuations and provide data support for the performance evaluation of the pump.

[0039] However, these existing technologies have exposed many problems in practical applications. On the one hand, relying solely on vibration signal analysis is difficult to cope with the multi-fault coupling effect of slurry circulation pumps. During the operation of the slurry circulation pump, multiple faults such as cavitation, bearing wear, and seal leakage may occur simultaneously. These faults affect and intertwine with each other, making the vibration signal characteristics complex and not obvious. For example, the vibration characteristic frequency caused by cavitation may be close to the frequency generated by early bearing wear, resulting in misjudgment. On the other hand, the strong background noise in the industrial site seriously interferes with the accuracy of the monitoring signal. The operating environment of the slurry circulation pump is noisy, and the electromagnetic interference of surrounding equipment, the turbulent noise of the fluid in the pipeline, and the background vibration of mechanical operation will cover up the fault characteristic signal. Taking acoustic emission signal monitoring as an example, the weak acoustic emission signal generated by early cavitation is often submerged by strong noise, making the monitoring system unable to capture the early signs of faults in time. It can only be detected when the fault develops to a certain extent and the characteristic signal is strong enough, thus delaying the maintenance time and increasing the risk of sudden equipment failures, posing a threat to production continuity and safety.

[0040] In summary, in the face of the complex working conditions of multiple fault couplings and strong noise interference in slurry circulation pumps, a new monitoring technology solution is urgently needed to overcome the bottleneck of the existing technology, accurately identify the fault types, and early warn of potential equipment hazards. The current market has an urgent need for efficient, reliable, and intelligent equipment condition monitoring technologies, and industry standards are also constantly improving, requiring the monitoring system to have higher sensitivity, lower false alarm rates, and stronger anti-interference capabilities. Therefore, how to break through the limitations of traditional monitoring technologies in the state monitoring of slurry circulation pumps has become a difficult problem that technicians in this field urgently need to overcome. The present invention was born precisely based on such an urgent need, aiming to provide a state monitoring method and related equipment for slurry circulation pumps that can effectively cope with multiple fault couplings and strong background noise interference, so as to ensure the stable operation of slurry circulation pumps, reduce maintenance costs, and improve the safety and efficiency of industrial production.

[0041] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0042] Refer to Figure 1 A specific implementation manner of the state monitoring method for a slurry circulation pump provided by the present invention is shown, including:

[0043] S1, synchronously collect vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals in the triangular monitoring topology composed of the pump body volute, outlet pipe, and bearing housing;

[0044] S2, perform wavelet packet decomposition processing on the acoustic emission signal, extract the energy ratio feature in the frequency band of 30 - 50 kHz, and fuse the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor;

[0045] S3, input the cavitation sensitivity factor, vibration signal features, and temperature signal features into the digital twin model established through fluid-structure coupling simulation and adapted through transfer learning to obtain the deviation amount between the real-time monitoring features and the normal working condition baseline;

[0046] S4, when the deviation amount exceeds the dynamic threshold, trigger fault mode recognition and remaining life prediction.

[0047] Specifically, in step S1, in the triangular monitoring topology formed by the pump body volute, the outlet pipe, and the bearing housing, vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals are synchronously collected. Among them, the acoustic emission sensor uses a PZT piezoelectric ceramic sensor with a resonant frequency of 80 kHz, and the vibration signals are collected by a MEMS triaxial accelerometer. The triangular monitoring topology satisfies that the axial distance between the volute measuring point and the outlet pipe measuring point is 1.2 - 1.5 times the pump outlet pipe diameter, and the bearing housing measuring point forms an angle of 60° ± 5° with the connection line of the volute - outlet pipe.

[0048] In this specific embodiment, the sensors are deployed at three key parts of the pump body volute, the outlet pipe, and the bearing housing to form a triangular monitoring layout. Through this layout, it is possible to achieve comprehensive coverage of the key parts of the pump body, ensuring the comprehensiveness and reliability of signal acquisition.

[0049] During the monitoring process, each sensor synchronously collects vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals. Among them, the acoustic emission signals are collected using a PZT piezoelectric ceramic sensor with a resonant frequency of 80 kHz. This sensor can sensitively capture the weak acoustic emission signals inside the pump, which may originate from cavitation, crack propagation, or other microscopic faults, providing an important basis for the early identification of faults.

[0050] The vibration signals are collected by a MEMS triaxial accelerometer, which can monitor the vibration conditions of the pump body in three mutually perpendicular directions in real time. This sensor has the characteristics of high precision, real - time performance, and multi - dimensional detection, providing comprehensive vibration data for analyzing the mechanical state of the pump and helping to identify common faults such as imbalance, looseness, and bearing wear.

[0051] In addition, the pressure pulsation signals are collected through a high - precision pressure sensor to monitor the pressure fluctuations in the inlet and outlet pipes of the pump. The changes in the pressure pulsation signals can reflect the hydrodynamic characteristics of the pump and help identify potential problems such as cavitation and flow rate changes.

[0052] The temperature signals are collected using sensors such as thermocouples or thermal resistors. By monitoring the temperature changes at the key parts of the pump body, problems such as abnormal lubrication, overload, or other issues that may cause the equipment to overheat can be detected in a timely manner.

[0053] By synchronously collecting the above - mentioned multiple signals, it is possible to provide multi - dimensional data support for the condition monitoring of the slurry circulation pump.

[0054] Preferably, the layout of the triangular monitoring topology needs to satisfy specific geometric relationships to ensure the accuracy and effectiveness of signal acquisition. Specifically, the axial distance between the volute measuring point and the outlet pipe measuring point should be set to 1.2 to 1.5 times the pump outlet pipe diameter. The determination of this distance range is based on a comprehensive analysis of the internal fluid dynamics characteristics and mechanical vibration characteristics of the pump. When the axial distance is too small, the collected signals may be interfered by the complex flow field inside the pump, while when the distance is too large, the signal intensity may weaken, affecting the identification of fault characteristics. Therefore, choosing 1.2 to 1.5 times the pump outlet pipe diameter as the axial distance can achieve a balance between signal intensity and interference avoidance, ensuring that the collected signals have a high signal-to-noise ratio and feature recognition degree.

[0055] Meanwhile, the position of the bearing housing measuring point needs to form an angle of 60° ± 5° with the connecting line of the volute and the outlet pipe measuring points. Such an angle setting has been verified through a large number of experiments and simulations, aiming to optimize the spatial relationship among the three measuring points and form a stable triangular monitoring structure. Through this layout, it can ensure that the vibration, acoustic emission, pressure pulsation, and temperature signals of the pump body are captured from different directions, thus more comprehensively reflecting the operating state of the pump. When the bearing housing measuring point forms an angle of 60° ± 5° with the volute-outlet pipe connecting line, it can minimize the mutual interference between signals, while increasing the sensitivity to multi-fault coupling phenomena, enabling the monitoring system to more accurately identify and locate the fault source. This carefully designed triangular monitoring topology provides a reliable basis for subsequent signal processing, fault diagnosis, and prediction, enabling the present invention to effectively overcome the deficiencies in the prior art and achieve efficient and accurate monitoring of the slurry circulation pump state.

[0056] Specifically, in S2, when performing wavelet packet decomposition processing on the acoustic emission signal, use the db8 wavelet basis for 5-layer decomposition and extract the energy ratio feature in the 30 - 50 kHz frequency band. Specifically, first calculate the energy E30 - 50kHz of the acoustic emission signal in the 30 - 50 kHz frequency band and the energy E0 - 10kHz in the 0 - 10 kHz frequency band.

[0057] Then, calculate the initial value of the cavitation sensitivity factor through the following formula:

[0058]

[0059]

[0060] ​Preferably, in a specific embodiment of the slurry circulation pump state detection method provided by the present invention, in order to improve the accuracy of the monitoring system, the cavitation sensitivity factor is fused with the standard deviation of the pressure pulsation signal. The specific fusion formula is:

[0061]

[0062] where β takes a value of 0.2 - 0.4, and σ(P) is the standard deviation of the pressure pulsation signal. This standard deviation can reflect the fluctuation characteristics of the pressure pulsation signal, which usually increases when cavitation occurs. By combining the characteristics of the pressure pulsation signal with the characteristics of the acoustic emission signal, the cavitation sensitivity factor can more comprehensively reflect the occurrence and development of the cavitation phenomenon.

[0063] This step-by-step calculation method not only improves the calculation accuracy of the cavitation sensitivity factor but also enhances its sensitivity to the cavitation phenomenon. In this way, the present invention can effectively distinguish cavitation signals from other interference signals in a complex industrial environment, providing a more accurate basis for subsequent fault diagnosis. At the same time, the introduction of this factor helps to improve the sensitivity and reliability of the monitoring system, enabling the system to promptly capture early signs of cavitation, thereby achieving efficient and accurate monitoring of the state of the slurry circulation pump.

[0064] Specifically, in S3, the cavitation sensitivity factor, the vibration signal characteristics, and the temperature signal characteristics are input into the digital twin model established through fluid-structure coupling simulation and adapted through transfer learning to obtain the deviation between the real-time monitoring characteristics and the normal operating condition baseline. The digital twin model is precisely constructed by means of fluid-structure coupling simulation technology. This simulation method synchronously simulates fluid dynamics and structural mechanics behaviors, accurately maps the operating state of the slurry circulation pump under normal operating conditions, and generates normal operating condition baseline data covering various key characteristics. These data are incorporated into the model as a reference standard for the healthy state. The model receives real-time data of the cavitation sensitivity factor, vibration signal characteristics, and temperature signal characteristics, and quickly compares and analyzes them using deep learning algorithms to calculate the deviation between the real-time monitoring characteristics and the normal operating condition baseline. The deviation intuitively reflects the degree of difference between the current state and the normal state, providing a key quantitative indicator for fault diagnosis.

[0065] In the model adaptation process, transfer learning technology is innovatively applied, using a domain adversarial neural network. This network can effectively resolve the problem of the difference in feature distributions between the source domain (simulation data) and the target domain (actual monitoring data). With the help of the domain adversarial training strategy, the model is prompted to learn the adaptive ability to actual data while using the pre-trained parameters of the simulation data. In this way, the digital twin model exhibits high accuracy and reliability in real monitoring scenarios.

[0066] The implementation of this monitoring process has advanced the status monitoring of slurry circulation pumps into a new stage of precision and intelligence. Based on the normal operating condition baseline generated by simulation, the digital twin model, combined with the enhancement adapted by transfer learning, has the ability to accurately compare and analyze real-time signals. It can quickly locate the fault source, clarify the fault type and degree, greatly shorten the fault response time, and enhance the safety and stability of industrial production.

[0067] Specifically, in S4, when the deviation exceeds the dynamic threshold, fault mode recognition and remaining life prediction are triggered. The remaining life prediction uses the attention mechanism LSTM model, and the input data includes the cavitation sensitivity factor, the kurtosis value of the shaft envelope spectrum of the vibration signal, and the third derivative of the temperature change rate. The implementation of this prediction mechanism enables the status monitoring of slurry circulation pumps to be forward-looking, capable of predicting the occurrence of equipment failures in advance, and providing sufficient time and basis for equipment maintenance and production plan adjustment.

[0068] The dynamic threshold is adaptively adjusted according to the real-time operating status and historical data of the slurry circulation pump. Compared with the fixed threshold, the dynamic threshold can better adapt to the changes in the equipment operating status, reducing false alarms and missed alarms. When the deviation between the real-time monitoring feature and the normal operating condition baseline exceeds the dynamic threshold, it indicates that the equipment may have abnormal conditions and further fault mode recognition and remaining life prediction are required.

[0069] The fault mode recognition module comprehensively analyzes the multi-source signal features collected, and quickly determines the current fault type and severity by comparing the feature database of known fault modes. This process utilizes machine learning algorithms and can identify various fault modes such as cavitation, bearing wear, and seal damage.

[0070] The remaining life prediction uses the LSTM model with the attention mechanism. This model can capture the long-term dependence relationships in time series data and is particularly suitable for processing the dynamic operation data of equipment. The input data includes the cavitation sensitivity factor, the kurtosis value of the shaft envelope spectrum of the vibration signal, and the third derivative of the temperature change rate. These data can reflect the operating status and wear degree of the equipment from different perspectives. The attention mechanism can automatically focus on key features, improving the accuracy and reliability of the prediction.

[0071] Among them, in the input features, the cavitation sensitivity factor is generated by processing the acoustic emission signal through wavelet packet decomposition and fusing the standard deviation of the pressure pulsation signal, and can sensitively reflect the occurrence and development of the cavitation phenomenon. The kurtosis value of the shaft envelope spectrum is obtained by performing shaft envelope spectrum analysis on the vibration signal, and can effectively characterize the early fault characteristics of the bearing. The third derivative of the temperature change rate is obtained by performing high-order derivative calculation on the temperature signal, and can capture the subtle characteristics of the temperature change and reflect the change of the thermal state of the equipment. The implementation of this fault mode recognition and remaining life prediction mechanism makes the condition monitoring of the slurry circulation pump more intelligent and forward-looking.

[0072] To make the slurry circulation pump condition monitoring method provided by the present invention easier to understand, a specific implementation manner combined with a use scenario is provided next to further illustrate the present solution.

[0073] In a certain thermal power plant, the slurry circulation pump is used to transport desulfurization slurry, and its stable operation is crucial for the normal operation of the environmental protection facilities of the power plant. The slurry circulation pump condition monitoring method provided by the present invention is used for it.

[0074] In the slurry circulation pump monitoring system of the thermal power plant, sensors are installed at three key parts: the pump body volute, the outlet pipe, and the bearing housing, forming a triangular monitoring topology structure. Specifically, the axial distance between the volute measurement point and the outlet pipe measurement point is set to 1.2 times the pump outlet pipe diameter, and the bearing housing measurement point forms an angle of 60° with the volute-outlet pipe connection line. Such a geometric layout aims to optimize the accuracy and reliability of signal acquisition. The installed sensors include a PZT piezoelectric ceramic sensor with a resonant frequency of 80 kHz for collecting acoustic emission signals; a MEMS triaxial accelerometer for collecting vibration signals; a high-precision pressure sensor for collecting pressure pulsation signals; and a thermocouple temperature sensor for collecting temperature signals. After the monitoring system is started, the acoustic emission signal, vibration signal, pressure pulsation signal, and temperature signal are collected synchronously. The acoustic emission signal is processed through wavelet packet decomposition, using the db8 wavelet basis for 5-layer decomposition, and the energy characteristics in the 30-50 kHz frequency band and the energy characteristics in the 0-10 kHz frequency band are extracted. Based on these characteristics, the initial value of the cavitation sensitivity factor is calculated where α is taken as 0.7. Further, the standard deviation σ(P) of the pressure pulsation signal is fused, and the cavitation sensitivity factor is updated to F cav_updated =F cav +β·σ(P).

[0075] Subsequently, the updated cavitation sensitivity factor F cav_updatedThe kurtosis value of the shaft envelope spectrum of the vibration signal and the third derivative of the temperature change rate are used as input features and input into the digital twin model adapted through transfer learning. This model generates a normal operating condition baseline through fluid-structure coupling simulation and uses a domain adversarial neural network for transfer learning adaptation to adapt to actual monitoring data. The system calculates the deviation between the current monitoring features and the normal operating condition baseline in real time and compares it with a dynamic threshold. The dynamic threshold is adaptively adjusted based on historical data and real-time operating status to reduce false alarms and missed detections. When the deviation exceeds the dynamic threshold, the fault mode recognition module is triggered. This module quickly determines the current fault type and severity by comparing with the feature database of known fault modes. For example, a significant increase in the cavitation sensitivity factor may indicate the occurrence of cavitation, while an abnormal increase in the kurtosis value of the shaft envelope spectrum of the vibration signal may indicate bearing wear. At the same time, the remaining life prediction module is triggered and an LSTM model with an attention mechanism is used for prediction. The input data of the model includes the cavitation sensitivity factor, the kurtosis value of the shaft envelope spectrum of the vibration signal, and the third derivative of the temperature change rate, and the output is the remaining life prediction result of the slurry circulation pump, providing a basis for maintenance personnel to formulate maintenance plans in advance.

[0076] Through the slurry circulation pump status detection method provided by the present invention, the thermal power plant can achieve real-time status monitoring of the slurry circulation pump, timely discover potential faults, and avoid production interruptions and equipment damage caused by sudden faults. This monitoring system not only improves the reliability and operating efficiency of the equipment, but also reduces maintenance costs, extends the service life of the equipment, and provides a strong guarantee for the stable production and environmental protection compliance of the power plant.

[0077] Through innovative multi-source signal fusion and feature extraction methods, the present invention effectively solves the problems of high false alarm rate and late warning in the prior art under multi-fault coupling and strong background noise interference. At the key parts of the slurry circulation pump, namely the pump body volute, outlet pipe, and bearing housing, a triangular monitoring topology is formed and acoustic emission signals, vibration signals, pressure pulsation signals, and temperature signals are synchronously collected. Such a layout can comprehensively capture the operating status information of the pump. For the acoustic emission signal, wavelet packet decomposition is used for processing, and the energy ratio feature in the 30 - 50 kHz frequency band is extracted, and the standard deviation of the pressure pulsation signal is fused to generate the cavitation sensitivity factor, so as to accurately reflect the cavitation phenomenon.

[0078] Meanwhile, the optimized design of the triangular monitoring topology, with reasonable settings of the measuring point spacing and angles, further improves the accuracy and reliability of signal acquisition, reduces signal interference, and enhances the sensitivity to fault characteristics. On this basis, the present invention establishes a digital twin model by fluid-structure coupling simulation and adapts it through the domain adversarial neural network in transfer learning, enabling it to better adapt to actual monitoring data. This model can generate a normal operating condition baseline, compare it with the real-time monitoring characteristics, and calculate the deviation amount. The introduction of a dynamic threshold enables the system to trigger fault mode recognition and remaining life prediction based on the deviation amount, effectively reducing false alarms and missed detections and achieving early fault warning.

[0079] In addition, the present invention uses an advanced attention mechanism LSTM model for remaining life prediction, comprehensively considering the cavitation sensitivity factor, vibration signal characteristics, and temperature signal characteristics, further improving the accuracy and reliability of fault diagnosis. In summary, through the synergistic effect of the above multiple innovative points, the present invention significantly improves the accuracy, reliability, and forward-looking of the slurry pump status monitoring, effectively overcoming the deficiencies in the prior art.

[0080] Refer to Figure 2 As shown, a slurry pump status monitoring system provided in another specific embodiment of the present invention includes:

[0081] A data acquisition module for synchronously acquiring vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals in the triangular monitoring topology formed by the pump volute, outlet pipe, and bearing housing;

[0082] A signal decomposition module for performing wavelet packet decomposition on the acoustic emission signal, extracting the energy ratio feature in the 30 - 50 kHz frequency band, and fusing the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor;

[0083] A deviation acquisition module for inputting the cavitation sensitivity factor, vibration signal characteristics, and temperature signal characteristics into a digital twin model established by fluid-structure coupling simulation and adapted through transfer learning to obtain the deviation amount between the real-time monitoring characteristics and the normal operating condition baseline;

[0084] A response module for triggering fault mode recognition and remaining life prediction when the deviation amount exceeds the dynamic threshold.

[0085] In a specific embodiment of the present invention, a computer device is further provided. Specifically, the computer device includes a processor and a memory. The memory is used to store a computer program, and the computer program includes program instructions. The processor is used to execute the program instructions stored in the computer storage medium. The processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions. Specifically, it is suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function. The processor described in the embodiment of the present invention can be used to synchronously collect vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals in a triangular monitoring topology composed of a pump body volute, an outlet pipe, and a bearing housing; perform wavelet packet decomposition processing on the acoustic emission signals, extract the energy ratio feature in the frequency band of 30 - 50 kHz, and fuse the standard deviation of the pressure pulsation signals to generate a cavitation sensitivity factor; input the cavitation sensitivity factor, vibration signal features, and temperature signal features into a digital twin model established through fluid-structure coupling simulation and adapted through transfer learning to obtain the deviation amount between the real-time monitoring features and the normal operating condition baseline; when the deviation amount exceeds the dynamic threshold, trigger fault mode recognition and remaining life prediction.

[0086] In a specific embodiment of the present invention, a storage medium is further provided. Specifically, it is a computer-readable storage medium, which is a memory device in a terminal device and is used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and, of course, the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space, and the operating system of the terminal is stored in this storage space. Moreover, one or more instructions suitable for being loaded and executed by a processor are stored in this storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the method in the above embodiments; one or more instructions in the computer-readable storage medium are loaded and executed by the processor to perform the following steps: synchronously collect vibration signals, acoustic emission signals, pressure pulsation signals, and temperature signals in the triangular monitoring topology composed of the pump body volute, outlet pipeline, and bearing housing; perform wavelet packet decomposition processing on the acoustic emission signals, extract the energy ratio feature in the frequency band of 30 - 50 kHz, and fuse the standard deviation of the pressure pulsation signals to generate a cavitation sensitivity factor; input the cavitation sensitivity factor, vibration signal features, and temperature signal features into a digital twin model established through fluid-structure coupling simulation and adapted through transfer learning to obtain the deviation amount between the real-time monitoring features and the normal operating condition baseline; when the deviation amount exceeds the dynamic threshold, trigger fault mode recognition and remaining life prediction.

[0087] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0088] This application is described with reference to the flowcharts and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing device produce a means for implementing the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more of the blocks.

[0089] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to operate in a specific manner, such that the instructions stored in the computer-readable memory produce a manufacture including an instruction means that implements the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more of the blocks.

[0090] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operational steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in one or more of the flows Figure 1 one or more of the flows and / or blocks Figure 1 or a means for implementing the functions specified in one or more of the blocks.

[0091] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above-described exemplary embodiments, and without departing from the spirit or basic characteristics of the present invention, the present invention can be implemented in other specific forms. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-limiting. The scope of the present invention is defined by the appended claims rather than the above description. Therefore, it is intended to embrace all changes that fall within the meaning and scope of the equivalent elements of the claims in the present invention. Any reference signs in the claims should not be construed as limiting the claimed rights.

[0092] In addition, it should be understood that although this specification is described according to embodiments, not every embodiment only contains an independent technical solution. This narrative way of the specification is only for clarity. Those skilled in the art should regard the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art. The above content is only to illustrate the technical idea of the present invention, and the protection scope of the present invention cannot be limited thereby. Any modification made on the basis of the technical solution according to the technical idea proposed by the present invention falls within the protection scope of the claims of the present invention.

Claims

1. A method for monitoring the state of a slurry circulation pump, characterized in that Including: In the triangular monitoring topology formed by the pump body volute, outlet pipe and bearing housing, synchronously collect vibration signals, acoustic emission signals, pressure pulsation signals and temperature signals; Perform wavelet packet decomposition on the acoustic emission signal, extract the energy ratio feature in the 30 - 50 kHz frequency band, and fuse the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor; Input the cavitation sensitivity factor, vibration signal features and temperature signal features into a digital twin model established through fluid - structure coupling simulation and adapted through transfer learning, and obtain the deviation between the real - time monitoring features and the normal operating condition baseline; When the deviation exceeds the dynamic threshold, trigger fault mode recognition and remaining life prediction.

2. The state monitoring method of a slurry circulation pump according to claim 1, characterized in that In the triangular monitoring topology formed by the pump body volute, outlet pipe and bearing housing, synchronously collect vibration signals, acoustic emission signals, pressure pulsation signals and temperature signals, where the acoustic emission sensor uses a PZT piezoelectric ceramic sensor with a resonant frequency of 80 kHz, and the vibration signal is collected by a MEMS tri - axis accelerometer.

3. The method for monitoring the state of a slurry circulation pump according to claim 1, characterized in that, The triangular monitoring topology satisfies: The axial distance between the volute measurement point and the outlet pipe measurement point is 1.2 - 1.5 times the pump outlet pipe diameter; The measurement point on the bearing housing forms an angle of 60° ± 5° with the connection line of the volute - outlet pipe.

4. A method for monitoring the state of a slurry circulation pump according to claim 1, characterized in that, Perform wavelet packet decomposition on the acoustic emission signal, extract the energy ratio feature in the 30 - 50 kHz frequency band, and fuse the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor, where the wavelet packet decomposition is performed with a db8 wavelet basis for 5 - layer decomposition, and the calculation formula for the cavitation sensitivity factor is: where α takes a value of 0.6 - 0.

8.

5. A method for monitoring the state of a slurry circulation pump according to claim 4, characterized in that, The cavitation sensitivity factor further fuses the pressure pulsation feature: where β takes a value of 0.2 - 0.4, and σ(P) is the standard deviation of the pressure pulsation signal.

6. The state monitoring method of a slurry circulation pump according to claim 1, characterized in that, Input the cavitation sensitivity factor, vibration signal features and temperature signal features into a digital twin model established through fluid - structure coupling simulation and adapted through transfer learning, and obtain the deviation between the real - time monitoring features and the normal operating condition baseline, where the transfer learning adaptation uses a domain - adversarial neural network, and the digital twin model generates the normal operating condition baseline through fluid - structure coupling simulation.

7. A method for monitoring the state of a slurry circulation pump according to claim 1, characterized in that, When the deviation exceeds the dynamic threshold, trigger fault mode recognition and remaining life prediction, where the remaining life prediction uses an attention - mechanism LSTM model, and the input data includes the cavitation sensitivity factor, the kurtosis value of the shaft envelope spectrum of the vibration signal, and the third - order derivative of the temperature change rate.

8. A state monitoring system for a slurry circulation pump, characterized in that, Including: A data acquisition module for synchronously collecting vibration signals, acoustic emission signals, pressure pulsation signals and temperature signals in the triangular monitoring topology formed by the pump body volute, outlet pipe and bearing housing; A signal decomposition module for performing wavelet packet decomposition on the acoustic emission signal, extracting the energy ratio feature in the 30 - 50 kHz frequency band, and fusing the standard deviation of the pressure pulsation signal to generate a cavitation sensitivity factor; A deviation acquisition module for inputting the cavitation sensitivity factor, vibration signal features and temperature signal features into a digital twin model established through fluid - structure coupling simulation and adapted through transfer learning, and obtaining the deviation between the real - time monitoring features and the normal operating condition baseline; A response module, configured to trigger fault mode recognition and remaining life prediction when a deviation amount exceeds a dynamic threshold.

9. A computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the steps of the slurry circulation pump status monitoring method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, the steps of the slurry circulation pump status monitoring method according to any one of claims 1 to 7 are implemented.

Images