Sea water pump fault early warning system based on monitoring data

Abstract

The invention relates to the technical field of fault early warning, in particular to a sea water pump fault early warning system based on monitoring data, which comprises a multi-source sensing module for acquiring an acoustic emission original signal, an electric corrosion current signal, a real-time turbidity value and a bearing temperature value in real time through an acoustic emission sensor, an electric corrosion current probe, a turbidity sensor and a bearing temperature sensor; the feature extraction module is used for receiving the output data, extracting an acoustic emission original signal to generate a cavitation energy migration value, analyzing an electric corrosion current signal to output electric corrosion feature intensity, associating a real-time turbidity value with the change rate of the cavitation energy migration value, and calculating a biological adhesion induction coefficient; the health assessment module is used for fusing and generating a dynamic health index; and the intelligent response module is used for correcting the action threshold value in combination with a preset environment corrosion factor, and outputting a grading protection instruction when the dynamic health index exceeds the corrected action threshold value. The method is suitable for sea water pump scenes, and has good engineering value and expandability.

Description

Technical Field

[0001] This invention relates to the field of fault early warning technology, and in particular to a seawater pump fault early warning system based on monitoring data. Background Technology

[0002] In marine engineering, seawater desalination, and ship systems, seawater pumps are key power equipment, and their stable operation is directly related to the safety and efficiency of the entire system. However, due to multiple factors such as suspended particles in seawater, electrochemical corrosion, and marine organism attachment, seawater pumps often face a complex degradation risk during operation, including cavitation erosion, localized electrolytic corrosion, and chronic biofouling. These degradation processes are characterized by their suddenness and insidiousness, as well as their complex physical mechanisms. This makes it difficult for traditional monitoring methods that rely on a single parameter (such as current, voltage, or temperature) to capture early signs of failure in a timely and accurate manner, resulting in monitoring blind spots and response delays.

[0003] Existing pump condition diagnostic technologies generally lack the ability to fuse and perceive multi-source physical signals and dynamically assess health. They are ill-suited to the real-time early warning requirements of environments with multiple coupled factors such as cavitation acoustic emission, electrical erosion current fluctuations, and biofouling evolution. This is especially true in complex marine environments where corrosive conditions often change drastically over time and across regions, and fixed action thresholds can easily lead to false alarms or missed alarms. Furthermore, existing systems lack a refined hierarchical response mechanism, making it impossible to adjust operating strategies reasonably according to the severity of the fault, which affects maintenance efficiency and equipment lifespan. Therefore, there is an urgent need for an intelligent seawater pump fault early warning system that integrates multi-source perception, feature extraction, health assessment, and adaptive response to improve the system's accuracy in identifying complex degradation risks and its emergency response capabilities. Summary of the Invention

[0004] This invention provides a seawater pump fault early warning system based on monitoring data.

[0005] A seawater pump fault early warning system based on monitoring data includes the following modules: Multi-source sensing module: Real-time acquisition of acoustic emission raw signals, electro-erosion current signals, real-time turbidity values ​​and bearing temperature values ​​through acoustic emission sensors, electro-erosion current probes, turbidity sensors and bearing temperature sensors; Feature extraction module: Receives data output by the multi-source sensing module, extracts the original acoustic emission signal to generate cavitation energy migration value, analyzes the electro-erosion current signal to output electro-erosion characteristic intensity, and correlates the change rate of the real-time turbidity value with the cavitation energy migration value to calculate the bioattachment induction coefficient; Health assessment module: Receives the cavitation energy migration value, electrical erosion characteristic intensity, and bioattachment induction coefficient, and fuses them to generate a dynamic health index; Intelligent response module: Receives the dynamic health index, combines it with preset environmental corrosion factors to correct the action threshold, and outputs a graded protection command when the dynamic health index exceeds the corrected action threshold.

[0006] Optionally, the multi-source sensing module includes: Acoustic emission and electro-erosion signal acquisition: Acquire acoustic emission signals and filter out noise, while measuring electro-erosion current signals; Turbidity and temperature data acquisition: The turbidity value at the pump outlet and the temperature value of the bearing are acquired in real time and then filtered.

[0007] Optionally, the acoustic emission and electro-erosion signal acquisition includes: Acoustic emission signal acquisition and preprocessing: Acoustic emission sensors are used to acquire the raw acoustic emission signals from the microcracks in the pump body at a set sampling frequency, and bandpass filtering is applied to eliminate low-frequency background interference and high-frequency noise. Electrolytic erosion current signal acquisition: The current fluctuation generated by the electrolysis effect inside the pump is recorded by an electrolytic erosion current probe.

[0008] Optionally, the acquisition of turbidity and temperature data includes: Real-time turbidity acquisition: Based on the principle of optical scattering, the instantaneous turbidity value of the fluid at the pump outlet is acquired through a turbidity sensor; Bearing temperature acquisition: The bearing surface temperature is measured using a temperature sensor, and steady-state compensation is performed using a first-order filtering model.

[0009] Optionally, the feature extraction module includes: Cavitation energy migration value calculation: Receive the filtered acoustic emission signal, construct its frequency power spectrum, calculate the frequency domain centroid of the energy distribution within a set frequency range, and generate cavitation energy migration value; Electro-erosion characteristic intensity extraction: Receive the electro-erosion current signal and calculate its fluctuation range within a sliding time window. Use the mean square error of the current value as the characteristic intensity index. Bioattachment induction coefficient calculation: The rate of change of the real-time collected turbidity value and the cavitation energy migration value is correlated. By constructing a ratio function that suppresses abiotic disturbances, a comprehensive index for identifying bioattachment trends, namely the bioattachment induction coefficient, is output.

[0010] Optionally, the calculation of the cavitation energy migration value includes: Acoustic emission signal power spectrum construction: Receive the original acoustic emission signal and construct the power spectral density function based on the filtered signal; Cavitation energy migration value calculation: Based on the power spectral density, the energy center position is calculated by frequency weighting to obtain the cavitation energy migration value.

[0011] Optionally, the extraction of the electro-erosion feature intensity includes: Calculate the average value of the electro-erosion current: Within a set time window, receive the electro-erosion current signal and calculate the average current within that time period. Extracting the intensity of electro-erosion characteristics: Based on the mean value of the electro-erosion current signal, statistically analyze the fluctuation amplitude range within a unit time window, and extract its effective change energy to characterize the intensity of electro-erosion characteristics.

[0012] Optionally, the calculation of the bioattachment induction coefficient includes: Calculation of cavitation energy migration rate of change: Perform time derivative calculation on the cavitation energy migration value to calculate its rate of change; Bioattachment induction coefficient generation: The rate of change of real-time turbidity value and cavitation energy migration value is jointly modeled to calculate the bioattachment induction coefficient.

[0013] Optionally, the health assessment module includes: Multi-parameter standardization processing: The cavitation energy migration value, the electrical erosion characteristic intensity and the bioattachment induction coefficient are received, and the three indicators are normalized to obtain the standardized characteristic quantities. Dynamic health index fusion calculation: Based on the sensitivity of each indicator to the health status of the equipment, weighting coefficients are set, and the dynamic health index is fused and calculated.

[0014] Optionally, the intelligent response module includes: Action threshold correction calculation: Receive the current dynamic health index and obtain the corresponding environmental corrosion factor. Correct the action threshold based on the linear decay model to obtain the corrected action threshold at the current moment. Graded protection instruction output: Compare the current dynamic health index with the corrected action threshold, and output the corresponding graded protection instruction when the conditions are met.

[0015] The beneficial effects of this invention are: This invention proposes a seawater pump fault early warning system based on monitoring data. It constructs a progressive fault diagnosis architecture of multi-source sensing, feature extraction, health assessment, and intelligent response, breaking through the limitations of existing technologies such as "single-parameter triggering," "fixed thresholds," and "coarse response." By introducing acoustic emission sensors, electro-erosion current probes, turbidity sensors, and bearing temperature sensors, it achieves real-time capture of different types of fault symptoms such as cavitation, electro-erosion, and biofouling, improving sensitivity to weak early degradation signals. Using cavitation energy migration value, electro-erosion characteristic intensity, and biofouling induction coefficient as key feature quantities, and based on spectral centroid modeling, current fluctuation energy calculation, and trend-suppressed normalized expression, it effectively enhances the model's ability to identify fault sources under multi-factor interference.

[0016] This invention proposes a dynamic health index fusion algorithm. In terms of health assessment and response strategies, it combines multi-parameter normalization and weight allocation mechanisms to achieve continuous assessment of the equipment's operational status throughout its entire lifecycle. Simultaneously, it introduces environmental corrosion factors to adaptively correct action thresholds, overcoming the problem of traditional fixed thresholds failing in complex environments. Furthermore, based on the dynamic range of the health index, it sets multi-level protection strategies, outputting graded instructions including minor intervention, moderate load reduction, and emergency shutdown, significantly improving the precision and practicality of the system's response. This algorithm is not only applicable to seawater pump scenarios but can also be extended to intelligent maintenance and early warning platforms for other multi-source interference-prone industrial equipment, demonstrating significant engineering value and scalability. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only for this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 This is a system block diagram of an embodiment of the present invention; Figure 2 This is a feature extraction diagram from an embodiment of the present invention. Detailed Implementation

[0019] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0020] It should be noted that the use of terms such as "an embodiment," "an embodiment," "an exemplary embodiment," and "some embodiments" in the specification indicates that the described embodiment may include a specific feature, structure, or characteristic, but not every embodiment necessarily includes that specific feature, structure, or characteristic. Furthermore, when a specific feature, structure, or characteristic is described in connection with an embodiment, implementing such a feature, structure, or characteristic in conjunction with other embodiments (whether explicitly described or not) should be within the knowledge of those skilled in the art.

[0021] Generally, terms can be understood at least partly from their use in context. For example, depending at least partly on the context, the term "one or more" as used herein can be used to describe any feature, structure, or characteristic in a singular sense, or a combination of features, structures, or characteristics in a plural sense. Additionally, the term "based on" can be understood not necessarily to convey an exclusive set of factors, but rather, alternatively, depending at least partly on the context, to allow for the presence of other factors that are not necessarily explicitly described.

[0022] like Figures 1-2 As shown, a seawater pump fault early warning system based on monitoring data includes the following modules: Multi-source sensing module: Real-time acquisition of acoustic emission raw signals, electro-erosion current signals, real-time turbidity values ​​and bearing temperature values ​​through acoustic emission sensors, electro-erosion current probes, turbidity sensors and bearing temperature sensors; The multi-source sensing module includes: Acoustic emission and electro-erosion signal acquisition: Acquire acoustic emission signals and filter out noise, while measuring electro-erosion current signals; Acoustic emission and electrical erosion signal acquisition includes: Acoustic emission signal acquisition and preprocessing: using an acoustic emission sensor to set the sampling frequency. Acoustic emission raw signals from microcracks within the pump body were collected. The signal is then bandpass filtered to eliminate low-frequency background interference and high-frequency noise. The filtered output signal is expressed as: ; in, It is the original acoustic emission signal, which is the time signal emitted by the acoustic emission sensor. The raw vibration electrical signals collected at all times, This is the filtered effective acoustic emission signal, the result after removing background interference, and is typically used for cavitation feature analysis. and The value range depends on the sensor's output range and the range setting of the acquisition system. This is the lower cutoff frequency of the bandpass filter, ranging from 20 to 100 Hz. It effectively avoids low-frequency mechanical resonance of the pump body and environmental background noise, ensuring that the acquired signal mainly originates from microcrack activity induced by cavitation. This is the upper limit frequency of the bandpass filter, ranging from 200 to 500 Hz. Signals above this range are often high-frequency electronic noise or invalid vibrational energy, which can easily lead to distortion in energy calculations. Setting an upper limit can prevent high-frequency interference from being mistakenly sampled. It is a bandpass filter operator used to preserve signal components and suppress other frequencies within a given frequency range. The bandwidth ranges from 100 to 300, which covers most of the acoustic emission energy distribution peaks generated during liquid cavitation, making it suitable for energy migration analysis and original acoustic emission signals. Including significant mechanical resonance, background noise, and high-frequency interference, a band-pass filter (BPF) is needed to limit the frequency domain of the signal related to cavitation in order to extract the effective frequency band. The upper and lower limits of the filter are... The frequency is usually selected based on the resonant frequency of the equipment structure and the frequency response range of the sensor, and the filtered signal... More suitable for subsequent energy transfer feature extraction; Electrolytic erosion current signal acquisition: The current fluctuations generated by the electrolytic effect inside the pump are recorded using an electrolytic erosion current probe. The acquired signal is represented as follows: ; in, It is the instantaneous electrolytic current, representing the time interval. The actual current value generated during the electrolysis process at any given time. It is the voltage across the probe, representing the electro-erosion current of the probe over time. The recorded voltage values ​​range from 0 to 5. The voltage signal originates from the electrolytic potential difference induced across the probe. The actual magnitude depends on the intensity of electrolytic activity generated by the pump's metal structure and the electrolyte concentration. This range ensures that weak electrolytic corrosion reactions can be captured while avoiding high voltage that could saturate the analog-to-digital converter (ADC). This refers to the internal resistance value of the electro-erosion current probe. The electro-erosion probe is essentially a series resistance measurement system, with a value ranging from 10 to 100 ohms. The resistance value needs to be sufficiently low to ensure that the voltage signal is not excessively attenuated, while also avoiding energy loss caused by excessive current. According to Ohm's law, the current... It can be determined by the voltage across the probe. With internal resistance Calculated, that is This relationship is valid only if the electrode electrolysis process is stable and the resistance value is constant, and it is suitable for online current monitoring. Turbidity and temperature data acquisition: Real-time acquisition of turbidity values ​​at the pump outlet and temperature values ​​of the bearings, followed by filtering. Turbidity and temperature data acquisition includes: Real-time turbidity acquisition: Based on the principle of optical scattering, the instantaneous turbidity value of the fluid at the pump outlet is acquired through a turbidity sensor, and is expressed as: ; in, It is the real-time turbidity value (NTU), which represents the degree of light scattering by suspended particles in the pump outlet fluid. This is the intensity of scattered light, ranging from 0 to 5. Higher intensity indicates more suspended matter. This value has an approximately linear relationship with the NTU (Nearest Temperature Unit). The sensor converts this value into an electrical signal for reading. It is the sensor calibration coefficient, which measures the intensity of scattered light. Convert to standard turbidity units The conversion ratio ranges from 100 to 500. Since different sensors have different optical sensitivity, structural layout and output methods, as well as the type, particle size distribution and color of the pump body liquid, etc., all affect the light scattering characteristics, in order to ensure that the measurement results are comparable and accurate in different application environments, a standard solution with known turbidity value (such as formalin solution) is usually used for multi-point calibration. The sensor's specific calibration coefficient is obtained by least squares method or linear fitting method. The range of 100-500 can meet the requirements of most industrial sites for medium and low turbidity resolution. Bearing temperature acquisition: The bearing surface temperature is measured using a temperature sensor, and steady-state compensation is performed using a first-order filtering model, expressed as: ; in, This is the current bearing temperature value (filtered). Bearing temperature reflects the lubrication status and mechanical friction; an abnormally high temperature can be a precursor to a fault. These are the sensor's raw measurement values, ranging from 30 to 120. The sensor output directly reflects temperature changes and is susceptible to environmental fluctuations and electrical noise interference, requiring filtering. It is a first-order filter factor used to control the weighting ratio between the current measurement value and the historical temperature; its value range is... , The larger the value, the faster the response to new values, but the lower the resistance to interference. The smaller the value, the slower the response but the more stable the fluctuation. Generally, an intermediate value is selected to balance responsiveness and stability.

[0023] Feature extraction module: Receives data output from multi-source sensing module, extracts raw acoustic emission signal to generate cavitation energy migration value, analyzes electro-erosion current signal to output electro-erosion characteristic intensity, correlates real-time turbidity value with cavitation energy migration value change rate, and calculates bioattachment induction coefficient; The feature extraction module includes: Cavitation energy migration value calculation: Receive the filtered acoustic emission signal, construct its frequency power spectrum, calculate the frequency domain centroid of the energy distribution within a set frequency range, and generate cavitation energy migration value, which is used to reflect the changing trend of the energy concentration position in the acoustic emission signal and to evaluate the cavitation intensity and frequency shift characteristics. The calculation of cavitation energy transfer values ​​includes: Construction of acoustic emission signal power spectrum: receiving the raw acoustic emission signal After filtering the signal Based on this, construct each moment The corresponding frequency Power spectral density function under This provides a foundation for subsequent frequency domain energy analysis; Cavitation energy migration value calculation: Based on the power spectral density, the energy center position is calculated by frequency weighting to obtain the cavitation energy migration value, which is expressed as: ; in, It is a moment The time-filtered acoustic emission signal at frequency The power spectral density, obtained by short-time Fourier transform (STFT) or power spectrum estimation methods (such as the Welch method), represents the energy density per unit frequency, and its value is affected by the intensity of the acoustic emission signal. It is the lower cutoff frequency of the bandpass filter. This is the upper limit frequency of the bandpass filter. A reasonable frequency band helps to avoid background mechanical vibration interference and high-frequency electronic noise, while focusing on frequency components related to cavitation activity. It reflects the degree of shift in the energy center of the acoustic emission spectrum, and is used to characterize the degree of shift in the energy center of the acoustic emission signal spectrum. Its value range is [value range missing]. Its changes reflect whether the spectral energy is migrating to higher frequencies (enhanced cavitation) or lower frequencies (weakened bubble disintegration), and are an important indicator for judging the cavitation evolution state. The essence of this formula is to calculate the centroid position of the frequency power spectrum. Electrolytic erosion characteristic intensity extraction: The electrolytic erosion current signal is received and its fluctuation range is calculated within a sliding time window. The mean square error of the current value is used as the characteristic intensity index to measure the intensity of the local electrolysis reaction and reflect the activity of the electrolytic erosion process. Electrolytic erosion feature intensity extraction includes: Calculate the average electrolytic erosion current: within a set time window length Inside, receive the electrolytic corrosion current signal. Calculate the average current during this time period. This serves as a reference level to reflect the current. Extraction of electro-erosion characteristic intensity: Based on the mean of the electro-erosion current signal, the fluctuation amplitude range within a unit time window is statistically analyzed to extract its effective change energy. Characteristic intensity of electrolytic erosion is expressed as: ; in, This is the length of the feature extraction time window, ranging from 0.2 to 5. It determines the time granularity of the fluctuation analysis. Too small a value will make the analysis overly sensitive to occasional noise, while too large a value will mask the characteristics of short-term discharge signals. Selecting a value of about 1 second usually balances the stability and sensitivity of the analysis, making it suitable for medium-rate electrolysis monitoring scenarios. At the current moment The average value of the electrolytic erosion current within the final time window is used to eliminate the overall bias trend and make the fluctuation analysis more focused on the abrupt change components. It is the current fluctuation intensity, which represents the fluctuation amplitude of the electrolytic current signal within a unit time window. Its essence is the standard deviation of the current within that time period. The larger the current standard deviation, the more unstable the electrolysis process is within that time period, and there may be strong local discharge, bubble-induced resistance change or micro-corrosion behavior. Bioattachment induction coefficient calculation: The rate of change of the real-time collected turbidity value and the cavitation energy migration value is correlated. By constructing a ratio function to suppress abiotic disturbances, a comprehensive index for identifying bioattachment trends, namely the bioattachment induction coefficient, is output, thereby improving the ability to identify chronic bioattachment faults. The calculation of the bioattachment induction coefficient includes: Calculation of cavitation energy migration rate: The cavitation energy migration value obtained in the previous step... Perform time derivative calculations to determine its rate of change. It is used to reflect the movement trend of the acoustic emission energy center over time; in, It is the rate of change of cavitation energy migration value, which represents the speed at which the centroid of the acoustic emission signal's spectrum shifts per unit time, and is used to determine whether the cavitation state is changing rapidly. Bioattachment induction coefficient generation: Real-time turbidity value With cavitation energy transfer value The rate of change was jointly modeled to calculate the bioattachment induction coefficient. , is represented as: ; in, This is the bioattachment induction coefficient, indicating whether there is a strong increase in turbidity caused by non-cavitation at the current moment. It reflects the potential trend of bioattachment pollution and is used to suppress interference from non-biological turbidity increases caused by rapid cavitation, enhancing the system's response to slow bioattachment behavior. Its value is determined by the relative proportion of turbidity to spectral changes. When the cavitation spectrum changes rapidly (with a large rate of change), the denominator increases. A decrease indicates that the increase in turbidity may be due to non-biological disturbances; while when the spectrum is stable but the turbidity continues to rise... The increased size suggests the possible presence of chronic pollution behaviors such as biofilms and marine organism attachment. It is the rate of change of cavitation energy transfer value, and its value range is... The larger the absolute value, the more drastic the change in the energy center of the cavitation spectrum, indicating that the system is in a state of frequent cavitation or violent cavitation collapse. When the value approaches zero, it indicates that the cavitation state is stable, which helps to distinguish whether the turbidity change is caused by biological adhesion rather than cavitation disturbance. It is a real-time turbidity value, representing the scattering intensity caused by suspended matter or contaminants in the pump outlet fluid. The value ranges from 0 to 1000, covering the standard range of industrial discharge liquid from clear water to highly turbid. This value is a directly measured physical quantity. The higher the value, the more serious the solid particles or biological pollution in the fluid. This range can identify the initial signal of light microbial attachment and also cover high-risk pollution situations.

[0024] Health assessment module: Receives cavitation energy migration value, electrical erosion characteristic intensity and bioattachment induction coefficient, and fuses them to generate a dynamic health index; The health assessment module includes: Multi-parameter normalization processing: receiving cavitation energy transfer values Electrolytic erosion characteristic intensity and bioattachment induction coefficient The three indicators were normalized to obtain the standardized feature quantities. The processing method is expressed as follows: ; in, These are the original eigenvalues. These are the lower and upper limits given by historical operating data or design benchmarks. These are the normalized index values, with a range of [value range missing]. ; Dynamic health index fusion calculation: Based on the sensitivity of each indicator to the health status of the equipment, weighting coefficients are set. ,satisfy 1. Integrated calculation of dynamic health index , is represented as: ; in, It is a dynamic health index, a quantitative health score generated by integrating three types of risk characteristics. The higher the value, the higher the risk of failure. It has a unified output range, which facilitates the setting of alarm thresholds and response levels, enabling multi-level early warning management or integration with other systems. It is the weighting coefficient of the three characteristic indicators. This is the weight of the cavitation factor, with a value ranging from 0.3 to 0.5. This represents the weight of the electrolytic corrosion factor, with a value ranging from 0.2 to 0.4. The weight of bioattachment factors ranges from 0.2 to 0.3. Cavitation is usually more destructive, so its weight is slightly higher. Although electrolytic erosion and bioattachment develop slowly, they are easily ignored and still need to account for a certain proportion.

[0025] Intelligent response module: Receives dynamic health index, combines it with preset environmental corrosion factors to correct the action threshold, and outputs graded protection command when the dynamic health index exceeds the corrected action threshold. The intelligent response module includes: Action threshold correction calculation: Receive the current dynamic health index And obtain the corresponding environmental corrosion factors. Action threshold based on linear decay model Make corrections to obtain the correction action threshold at the current moment. , is represented as: ; in, This is the action threshold at the current moment, representing the sensitivity threshold of the system's response to the health index under the current environment. This value determines whether a protection command is triggered, and it is adjusted by introducing a corrosion factor. It can automatically reduce its speed in highly corrosive environments, improving the system's response sensitivity and reducing the risk of major damage. It is the baseline action threshold, and its value range is... It falls within the high-to-medium range of the health index, balancing false alarm rate and timely response, making it suitable for most seawater pump operating scenarios. It is an environmental corrosion factor, representing the intensity of external seawater corrosion, ranging from [missing information]. It can be obtained by comprehensively modeling parameters such as salinity, pH, and temperature. 0 represents no corrosion, and 1 represents an extremely highly corrosive environment. The corrosion impact adjustment coefficient ranges from 0.1 to 0.3, which adjusts the degree of linear decay. The larger the value, the more significantly the action threshold is affected by corrosion. Setting this range can achieve sensitive but not aggressive response adjustment, effectively adapting to equipment systems with different protection levels. Graded protection instruction output: Compare the current dynamic health index. With the corrected action threshold When the conditions are met At that time, output the corresponding hierarchical protection command. The instruction level is indicated by the degree to which the health index exceeds the limit: ; in, The output protection command level indicates the control response level automatically selected by the system based on the current health status. It converts the continuous health index into control commands that can be executed by the engineering team, making it easy to embed into relay logic, control strategy library or PLC instruction set (0 indicates normal operation, 1 indicates minor intervention, 2 indicates moderate load reduction, 3 indicates emergency shutdown). This is the protection level interval factor, with a fixed value of 0.1. This value determines the response range width of each protection command. The smaller the value, the finer the system response but the more frequent the switching may be; the larger the value, the coarser the response but the stronger the anti-disturbance capability, which is suitable for systems with large fluctuations in the field environment but slow response. It is a dynamic health index. It is the corrected action threshold.

[0026] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0027] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A seawater pump fault early warning system based on monitoring data, characterized in that, Includes the following modules: Multi-source sensing module: Real-time acquisition of acoustic emission raw signals, electro-erosion current signals, real-time turbidity values ​​and bearing temperature values ​​through acoustic emission sensors, electro-erosion current probes, turbidity sensors and bearing temperature sensors; Feature extraction module: Receives data output by the multi-source sensing module, extracts the original acoustic emission signal to generate cavitation energy migration value, analyzes the electro-erosion current signal to output electro-erosion characteristic intensity, and correlates the change rate of the real-time turbidity value with the cavitation energy migration value to calculate the bioattachment induction coefficient; Health assessment module: Receives the cavitation energy migration value, electrical erosion characteristic intensity, and bioattachment induction coefficient, and fuses them to generate a dynamic health index; Intelligent response module: Receives the dynamic health index, combines it with preset environmental corrosion factors to correct the action threshold, and outputs a graded protection command when the dynamic health index exceeds the corrected action threshold.

2. The seawater pump fault early warning system based on monitoring data according to claim 1, characterized in that, The multi-source sensing module includes: Acoustic emission and electro-erosion signal acquisition: Acquire acoustic emission signals and filter out noise, while measuring electro-erosion current signals; Turbidity and temperature data acquisition: The turbidity value at the pump outlet and the temperature value of the bearing are acquired in real time and then filtered.

3. The seawater pump fault early warning system based on monitoring data according to claim 2, characterized in that, The acoustic emission and electro-erosion signal acquisition includes: Acoustic emission signal acquisition and preprocessing: Acoustic emission sensors are used to acquire the raw acoustic emission signals from the microcracks in the pump body at a set sampling frequency, and bandpass filtering is applied to eliminate low-frequency background interference and high-frequency noise. Electrolytic erosion current signal acquisition: The current fluctuation generated by the electrolysis effect inside the pump is recorded by an electrolytic erosion current probe.

4. A seawater pump fault early warning system based on monitoring data according to claim 2, characterized in that, The acquisition of turbidity and temperature data includes: Real-time turbidity acquisition: Based on the principle of optical scattering, the instantaneous turbidity value of the fluid at the pump outlet is acquired through a turbidity sensor; Bearing temperature acquisition: The bearing surface temperature is measured using a temperature sensor, and steady-state compensation is performed using a first-order filtering model.

5. A seawater pump fault early warning system based on monitoring data according to claim 4, characterized in that, The feature extraction module includes: Cavitation energy migration value calculation: Receive the filtered acoustic emission signal, construct its frequency power spectrum, calculate the frequency domain centroid of the energy distribution within a set frequency range, and generate cavitation energy migration value; Electro-erosion characteristic intensity extraction: Receive the electro-erosion current signal and calculate its fluctuation range within a sliding time window. Use the mean square error of the current value as the characteristic intensity index. Bioattachment induction coefficient calculation: The rate of change of the real-time collected turbidity value and the cavitation energy migration value is correlated. By constructing a ratio function that suppresses abiotic disturbances, a comprehensive index for identifying bioattachment trends, namely the bioattachment induction coefficient, is output.

6. A seawater pump fault early warning system based on monitoring data according to claim 5, characterized in that, The calculation of the cavitation energy migration value includes: Acoustic emission signal power spectrum construction: Receive the original acoustic emission signal and construct the power spectral density function based on the filtered signal; Cavitation energy migration value calculation: Based on the power spectral density, the energy center position is calculated by frequency weighting to obtain the cavitation energy migration value.

7. A seawater pump fault early warning system based on monitoring data according to claim 5, characterized in that, The extraction of the electrical erosion feature intensity includes: Calculate the average value of the electro-erosion current: Within a set time window, receive the electro-erosion current signal and calculate the average current within that time period. Extracting the intensity of electro-erosion characteristics: Based on the mean value of the electro-erosion current signal, statistically analyze the fluctuation amplitude range within a unit time window, and extract its effective change energy to characterize the intensity of electro-erosion characteristics.

8. A seawater pump fault early warning system based on monitoring data according to claim 5, characterized in that, The calculation of the bioattachment induction coefficient includes: Calculation of cavitation energy migration rate of change: Perform time derivative calculation on the cavitation energy migration value to calculate its rate of change; Bioattachment induction coefficient generation: The rate of change of real-time turbidity value and cavitation energy migration value is jointly modeled to calculate the bioattachment induction coefficient.

9. A seawater pump fault early warning system based on monitoring data according to claim 8, characterized in that, The health assessment module includes: Multi-parameter standardization processing: The cavitation energy migration value, the electrical erosion characteristic intensity and the bioattachment induction coefficient are received, and the three indicators are normalized to obtain the standardized characteristic quantities. Dynamic health index fusion calculation: Based on the sensitivity of each indicator to the health status of the equipment, weighting coefficients are set, and the dynamic health index is fused and calculated.

10. A seawater pump fault early warning system based on monitoring data according to claim 9, characterized in that, The intelligent response module includes: Action threshold correction calculation: Receive the current dynamic health index and obtain the corresponding environmental corrosion factor. Correct the action threshold based on the linear decay model to obtain the corrected action threshold at the current moment. Graded protection instruction output: Compare the current dynamic health index with the corrected action threshold, and output the corresponding graded protection instruction when the conditions are met.

Images