Fracturing truck valve body fault diagnosis method and system fusing mechanism analysis and data driving and storage medium

Abstract

The invention discloses a fracturing truck valve body fault diagnosis method and system fusing mechanism analysis and data driving and a storage medium. The method aims at solving the technical problems that in the prior art, the early-stage abrasion fault detection sensitivity of a key part valve body of a fracturing truck is low, diagnosis is not timely, and interference caused by on-site complex vibration is likely to happen. According to the scheme, the method is characterized in that firstly, vibration signals are collected, narrow-band filtering is optimized through adaptive spectral kurtosis to extract impact characteristics, then Hilbert transform is used for envelope demodulation, and periodic impact components are enhanced through exponential decay integration; fusing the average kurtosis value, the peak factor and the pulse index to generate a comprehensive characteristic index for preliminary judgment, and calculating an accumulated impact energy value and impact pulse repetition density as evaluation indexes; and finally, realizing accurate grading diagnosis and decision of the wear state by constructing a double-threshold dynamic evaluation mechanism. The method is mainly used for online state monitoring, early failure early warning and predictive maintenance support of the fracturing truck valve body.

Description

Technical Field

[0001] This invention relates to the field of valve body fault diagnosis. More specifically, this invention relates to a method, system, and storage medium for fracturing truck valve body fault diagnosis that integrates mechanism analysis and data-driven approaches. Background Technology

[0002] Accurate diagnosis of valve wear faults (a type of check valve) is crucial for ensuring the safe operation of drilling pumps and enabling predictive maintenance. Accurate diagnosis of these faults ensures safe pump operation, enables predictive maintenance, and improves the modernization of equipment management. On-site assessment of valve operation is typically done by sound, but this method lacks real-time accuracy and carries the risk of hearing damage. Vibration testing methods are suitable for drilling sites and can identify faults through changes in vibration signals. When a fault occurs in the hydraulic end (a crucial component of the fracturing pump, responsible for drawing in and discharging high-pressure fracturing fluid, including the valve body (valve assembly), plunger, valve seat, etc.), the time-domain and frequency-domain components of the vibration signal will change. For example, valve puncture and leakage faults can lead to changes in the time-domain impact amplitude and frequency-domain distribution. Although the vibration signal of the fracturing truck valve is complex and affected by multiple vibration sources, vibration analysis remains an effective method for fault diagnosis. Among these, how to process vibration analysis signals to eliminate interference and reveal the fault mechanism is a key research direction. Summary of the Invention

[0003] One object of the present invention is to solve at least the above-mentioned problems and to provide at least the advantages that will be described later.

[0004] To achieve these objectives and other advantages according to the present invention, a method for diagnosing valve failures in fracturing trucks that integrates mechanism analysis and data-driven approaches is provided, comprising the following steps: S1. Collect the vibration signal of the valve body of the fracturing truck, determine the bandpass filtering range using the adaptive spectral kurtosis analysis method, and perform narrowband filtering on the vibration signal; S2. Perform Hilbert envelope demodulation on the signal filtered in step S1 to extract the impulse envelope signal; S3. Perform exponential decay integral processing on the envelope signal obtained in step S2 to enhance the periodic impact component and suppress noise; S4. Extract the average kurtosis value from the signal processed in step S1 as the first feature, extract the peak factor and impulse index from the signal processed in step S3 as the second and third features, and weight and fuse the first feature, the second feature and the third feature to obtain the comprehensive feature index. S5. Compare the comprehensive characteristic index with a preset first threshold. If the comprehensive characteristic index is greater than or equal to the first threshold, it is determined to be an abnormal wear state; if it is less than the first threshold, it is determined to be a normal wear state and the subsequent steps are executed. S6. Based on the impact envelope signal extracted in step S2, calculate the cumulative impact energy value E per unit time. p and impact pulse repetition density L p ; S7. E calculated based on step S6 p and L p A dual-threshold dynamic evaluation mechanism is used to assess wear status and output diagnostic results and maintenance decisions.

[0005] Preferably, in step S6, the cumulative impact energy value E p and impact pulse repetition density L p The calculation method is as follows: ; ; Among them, E pi Let be the energy of the i-th impact pulse detected within time ΔT. This represents the total number of impact pulses detected within time ΔT.

[0006] Preferably, in step S7, the dual-threshold dynamic evaluation mechanism specifically includes: S701. Based on the processed data, calculate the cumulative integral of the impact amplitude A(t) and the cumulative integral of the number of impacts C(t); S702. Based on the historical full life cycle operation data of the fracturing truck valve, the cumulative impact amplitude threshold T is preliminarily determined. A and the cumulative threshold of impact number T C And dynamically adjust it in conjunction with the multidimensional normal distribution criterion; S703. If A(t) ≥ T A And C(t) ≥ T C Therefore, it can be preliminarily determined that the valve is in a worn state; S704. Using DS evidence theory, the basic probability assignments generated based on amplitude-impact joint criteria, trend evolution criteria, and pattern recognition criteria are fused to calculate the final wear probability. P wear ; S705. Based on the stated final wear probability P wear Within the preset probability range, execute the corresponding early warning, alarm, or emergency shutdown tiered operation.

[0007] Preferably, in step S704, the trend evolution criterion is obtained by predicting the development gradient of future cycles using an ARIMA time series model; the pattern recognition criterion is obtained by classifying signal features based on a deep belief network.

[0008] Preferably, in step S705, the grading operation specifically includes: When 70% ≤ P wear When the level is less than 85%, it is determined to be a warning level, triggering refined spectrum analysis and activating the backup lubrication system. When 85% ≤ P wear When the percentage is less than 95%, it is determined to be an alarm level, and a maintenance work order is automatically generated. when P wear When the failure rate is ≥ 95%, it is classified as an emergency, and shutdown protection is immediately implemented and the fault tree analysis module is activated.

[0009] Preferably, in step S3, the exponential decay kernel function used in the exponential decay integral processing is: h(τ) = e -βτ ·u(τ), where β is the attenuation coefficient, determined by the signal sampling rate and the duration of the impact, and u(τ) is the unit step function.

[0010] Preferably, in step S1, the adaptive spectral kurtosis analysis method optimizes the window function parameters to maximize the kurtosis value of the target frequency band, thereby determining the frequency band most sensitive to the impact characteristics as the bandpass filtering range.

[0011] Preferably, steps S6 and S7 further include calculating E... p and L p Based on the growth trend and cumulative value throughout the entire life cycle, analyze the trend of physical health decline and predict remaining lifespan.

[0012] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described fracturing truck valve fault diagnosis method.

[0013] This invention also provides a valve body fault diagnosis system for fracturing trucks, comprising: The signal acquisition module is used to acquire vibration signals from the valve body of the fracturing truck. The processor is configured to execute the steps of the above-described fracturing truck valve fault diagnosis method. The output module is used to output diagnostic results and maintenance decisions.

[0014] This invention offers at least the following advantages: First, it utilizes adaptive spectral kurtosis-optimized bandpass filtering to accurately extract wear impact characteristic signals, and enhances the time-domain characterization of the impact components through Hilbert envelope demodulation. Based on this, it innovatively introduces an exponential decay weighted model to construct a dual-threshold dynamic evaluation mechanism. Through the synergistic analysis of short-term decay integrals and long-term trend integrals, it achieves real-time quantitative evaluation of recent wear conditions. Finally, it verifies the algorithm using collected real-machine oil well data. Experimental results show that this algorithm combines the advantages of adaptive frequency domain feature selection and time-domain impact enhancement, significantly improving the detection sensitivity and diagnostic reliability of early weak wear signals, and providing an effective technical means for preventative maintenance of drilling pump valve bodies.

[0015] Other advantages, objectives and features of the present invention will become apparent in part from the following description, and in part from those skilled in the art through study and practice of the invention. Attached Figure Description

[0016] Figure 1 This is a flowchart of the fracturing truck valve fault diagnosis method that integrates mechanism analysis and data-driven approach as described in this invention; Figure 2 This is a flowchart of the second-level dual-threshold dynamic evaluation process of the present invention; Figure 3 This is a narrowband filtered envelope diagram for extracting impulse signals based on Hilbert envelope demodulation in an embodiment of the present invention. Figure 4 This is the narrowband filtered envelope of the impact signal after EDI optimization in this embodiment of the invention; Figure 5 This is a diagram illustrating the health degradation trend of the vulva throughout its entire life cycle in an embodiment of the present invention. Figure 6 This is a trend diagram of the impact amplitude of the valve body in an example of the present invention; Figure 7 This is a trend chart of the number of valve impacts in an example of the present invention. Detailed Implementation

[0017] The following detailed description, in conjunction with the accompanying drawings and specific embodiments, provides a method for diagnosing valve failures in fracturing trucks that integrates mechanism analysis and data-driven approaches, as described in this invention. Those skilled in the art should understand that the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without inventive effort are within the scope of protection of this invention.

[0018] In the development of unconventional resources such as oil, natural gas, and shale gas, fracturing trucks are the core equipment in hydraulic fracturing operations, and their stable operation directly affects construction efficiency and safety. The valve assembly, as a key component of the fracturing truck's high-pressure pumping system, plays a crucial role in controlling the intake and discharge of fracturing fluid. It withstands extreme conditions for extended periods, including high pressure (up to 100 MPa or more), high frequency (dozens of opening and closing times per minute), and sand-containing media, making it highly susceptible to failure due to wear, fatigue, and corrosion. Therefore, timely detection and accurate diagnosis of valve assembly failures are of great significance.

[0019] like Figure 1 As shown, this invention provides a method for diagnosing valve failures in fracturing trucks that integrates mechanism analysis and data-driven approaches, comprising the following steps: S1. Vibration signal acquisition and preprocessing A vibration acceleration sensor is installed on the rigid structure of the pump casing or valve box near the valve of the fracturing truck to collect vibration signals during the valve's operation. The collected signals are then processed by anti-aliasing filtering and analog-to-digital conversion before being input to the processing unit. In this embodiment, the sampling frequency is set to 20 kHz to meet the requirements for capturing high-frequency impact components.

[0020] S2. Narrowband Filtering Optimization Based on Adaptive Kujicism (ASK) The collected vibration signals x ( t Adaptive spectral kurtosis analysis is applied. First, the time spectrum of the signal is obtained through short-time Fourier transform (STFT). X ( t , f Subsequently, the spectral kurtosis is calculated by sliding within the time-frequency domain. SK ( t , f ): In the formula, ⟨⋅> represents the average operation within the time-frequency window. The window length and overlap rate of the STFT are optimized using a genetic algorithm to maximize the kurtosis value of the frequency band representing wear impact, thereby adaptively determining the optimal bandpass filtering range. B wear For example, in an implementation example of a sandstone oil well (the same applies to shale oil wells), the final determined sensitive frequency band was 4400 Hz to 6800 Hz. Using this frequency band, the original signal was bandpass filtered to obtain a preliminarily purified signal. x band ( t ).

[0021] S3. Envelope Demodulation Based on Hilbert Transform For the filtered signal x band ( t Perform a Hilbert transform to extract its envelope signal. ν env ( t ): ; Where H[⋅] represents the Hilbert transform. By setting a confidence threshold of 2%-3%, pulse events with significant amplitudes in the envelope signal are screened to eliminate random noise interference, and the resulting envelope waveform is as follows. Figure 3 As shown.

[0022] S4. Signal enhancement based on exponential decay integral (EDI) To enhance periodic impacts and further suppress aperiodic noise, an exponentially decaying kernel function is first constructed. h ( τ ): h(τ)=e -βτ ·u(τ) ; β The attenuation coefficient is determined by the signal sampling rate and the duration of the impact. u ( τ () is a unit step function; The envelope signal is then convolved to obtain the enhanced envelope signal. ν enh ( t ): ; This step is equivalent to a smoothing filter, which can effectively amplify the periodic impact sequence related to wear. The processed envelope waveform is as follows: Figure 4 As shown.

[0023] S5. Feature Extraction and Calculation of Comprehensive Evaluation Indicators The time-domain and frequency-domain features are extracted from the above processing and then weighted and fused.

[0024] Time-domain and frequency-domain characteristics include: ASK characteristics (frequency domain): Calculating the sensitive frequency band B wear Average spectral kurtosis value within SK avg .

[0025] EDI characteristics (time domain): from the enhanced envelope signal ν enh ( t In ), calculate the peak factor. CF and pulse indexI : ; ; Feature fusion includes: weighting and fusing the above three features to generate a comprehensive evaluation index. S Used to quantify the current wear status: S = w 1⋅ SK avg + w 2⋅ CF + w 3⋅ I ; Among them, the weighting coefficient w 1, w 2, w 3 can be determined through principal component analysis (PCA) or training optimization based on historical data. In this embodiment, we take... w 1 = 0.4, w 2 = 0.35, w 3 = 0.25, so S = 4.43.

[0026] S6. Wear condition assessment and decision-making based on dual thresholds This step employs a dual-threshold assessment process to achieve tiered diagnosis from initial screening to accurate judgment.

[0027] S601, First-level threshold determination (preliminary screening) The comprehensive evaluation index S calculated in real time is compared with the preset first fault threshold S. th Compare the threshold S. th Pre-calibration is performed using the following method: Vibration signal samples of valves in "known normal" and "known abnormal" states are collected from a historical database. Steps S1 to S5 are executed for each sample to obtain the corresponding set of S values. The optimal classification boundary is determined using a support vector machine (SVM) or by analyzing the probability distribution of the two classes (e.g., finding the minimum misclassification point), and this boundary value is set as S. th。 If S ≥ S th If S < S th If the wear is not properly assessed, it is determined to be normal or early wear, and the process proceeds to the next level of fine assessment based on cumulative wear. In this embodiment, S th =7, while S = 4.43 calculated in S5, therefore S < S th If it is determined to be normal impact wear, proceed to the following steps.

[0028] S602, Evaluation Indicator E p and L p Calculation The envelope signal obtained based on S4 ν enh ( t Identify and calculate impact events within a unit of time (1 second): Let the impact envelope waveform function of a valve body impact amplitude be: Assuming the first A shock pulse from start, End, definition Let the energy of the impact pulse be: That is E pi Corresponding to the envelope waveform function x e ( t )exist[ t 0, t The area between 1] and 2]. By definition, E pi The value of is related not only to the pulse amplitude but also to the total pulse duration. The higher the pulse amplitude, E pi The larger the pulse, the longer it lasts. E pi It is also larger. Therefore, through E pi It can accurately reflect the changes in the energy characteristics of a single impact amplitude pulse at different stages.

[0029] set up Let be the total number of impact pulses detected within time ΔT. Cumulative impact energy per unit time: ; Impact pulse repetition density per unit time: ; These two parameters form the basis for assessing wear evolution.

[0030] In this embodiment, the initial state of the lifecycle is: Cumulative total value , ; Phase 1: Normal Operation Period Daily value: E p ≈1,L p ≈5, assuming the normal operating period is n, then the cumulative total value is... , ; Phase Two: Early Wear and Tear Period Daily value: Ep ≈2.5, L p ≈8, assuming the early wear period is m, then the cumulative total value is... ; ; Phase Three: Severe Failure Period Daily value: E p ≈ 6.0, L p ≈15, assuming the severe failure period is l, then the cumulative total value is... ; ; Let the total time of the entire life cycle be t, then The growth trends of the total cumulative impact energy and the total cumulative number of impact pulses throughout the entire life cycle are as follows: Figure 5 As shown.

[0031] S603, Level 2 Dual Threshold Dynamic Evaluation This is the core assessment step, and the process is as follows: Figure 2 As shown.

[0032] First, calculate the cumulative integral of the impact amplitude A(t) and the cumulative integral of the number of impacts C(t); Accumulated integral of impact amplitude: ; Accumulated points for impact count: ; in, a i For the magnitude of the first impact, N ( t () represents the number of impacts within a given time period.

[0033] Taking a sandstone oil well as an example, the cumulative values ​​of its three life cycles are shown in Table 1.

[0034] Table 1. Cumulative number of impact amplitudes in sandstone oil wells at different cycles. type cycle Accumulated impact amplitude A(t) Cumulative number of impacts C(t) Operating characteristics Sandstone Formation Oil Wells t-2 <![CDATA[1.2×10 7 mm / s]]> <![CDATA[5.8×10 5 Next Moisture content 65%, pump pressure 12MPa Sandstone Formation Oil Wells t-1 <![CDATA[1.5×10 7 mm / s]]> <![CDATA[7.2×10 5 Next The chloride ion concentration increased to 8,500 mg / L. Sandstone Formation Oil Wells t <![CDATA[1.7×10 7 mm / s]]> <![CDATA[8.3×10 5 Next Sand particles (particle size > 50 μm) are present. Secondly, the wear state threshold of the valve is initially determined by using historical operating data of the valve throughout its entire life cycle (from initial installation to wear failure) of the fracturing truck. An exponentially weighted decay factor (e.g., λ=3) is introduced to give higher weight to recent data, so as to achieve dynamic adjustment of the threshold according to the operating conditions. ; Therefore, the cumulative integration threshold for the impact amplitude of sandstone oil wells is T. A =1.6×10 7 mm / s, impact quantity threshold T c =8×105 Second-rate.

[0035] Third, a preliminary assessment of the wear condition was made. The collected data shows that A(t) = 1.7 × 10⁻⁶. 7 mm / s, C(t) = 8.3 × 10 5 Next, determine A(t) = 1.7 × 10 7 mm / s ≥ T A =1.6×10 7 mm / s (exceeding the amplitude threshold); C(t) = 8.3 × 10 5 Times ≥ T C =8×10 5 If the number of times exceeds the threshold, the valve is preliminarily determined to be in a worn state.

[0036] Fourth, the DS evidence theory integration judgment: Evidence m1 (Amplitude-Impact Joint Criterion): The collected data satisfy A(t)≥T A And C(t)≥T c Based on the statistics of the fault case sample database, the basic probability of assignment is m1(A) = 0.85. = 0.1, = 0.05 (where A is the wear probability, This is a normal probability. (For uncertain terms) Evidence m2 (trend evolution criterion): Based on the ARIMA(2,1,2) model, the vibration amplitude for the next three cycles is predicted. Due to the gradient of amplitude changes, the value of m2(A) for sandstone oil wells is assigned as 0.7. = 0.2, = 0.1; Evidence m3 (pattern recognition criterion): SK avg , CF , I , E p , L p A deep belief network (DBN) is pre-trained with inputs of equal features. The classification confidence score of the DBN generates m3(A) = 0.78. = 0.12, = 0.1.

[0037] DBN structure: Input layer (number of features, SK avg , CF , I , E p , L p(5 features in total); 2 hidden layers; output layer (2 nodes, corresponding to wear / normal); pre-training: unsupervised pre-training of each Restricted Boltzmann Machine (RBM) using unlabeled device state feature data (learning rate 0.01, 100 iterations). Fusion using Dempster combinatorial rules m 1, m 2, m 3. Calculate the final wear probability: ; in The conflict factor is calculated by multiplying the probabilities of mutually exclusive propositions based on the evidence. m i ( X i ) No. i One piece of evidence against the proposition X i The basic probability is assigned. When K > 0.6, a manual review process is initiated; when K = 0.4, no manual review is required. Fifth, a judgment is made based on the pre-established three-level state determination system. The three-level state determination system includes: When 70% ≤ P wear When the level is less than 85%, it is determined to be a warning level, triggering refined spectrum analysis and activating the backup lubrication system. When 85% ≤ P wear When the percentage is less than 95%, it is determined to be an alarm level, and a maintenance work order is automatically generated. when P wear When the failure rate is ≥ 95%, it is classified as an emergency, and shutdown protection is immediately implemented and the fault tree analysis module is activated.

[0038] The final wear probability is obtained from the aforementioned calculations. P wear =0.7735. According to the judgment system, this state belongs to the "early warning level". The backup lubrication system was activated as recommended on site and observation was arranged. Subsequent disassembly and inspection confirmed that the valve body was in the early wear stage, verifying the effectiveness and accuracy of the method of the present invention in early fault warning.

[0039] Sixth, simultaneously generate vibration intensity reports conforming to ISO 2372 standards. Combined with current... E p and L p The value and its growth rate are used to predict the remaining useful life (RUL) of the valve based on degradation models (such as exponential models), providing a quantitative basis for preventive maintenance.

[0040] Based on the same inventive concept, the present invention also provides a fracturing truck valve fault diagnosis system that integrates mechanism analysis and data-driven approach. The system may be a personal computer, a server, or other system that implements the aforementioned fracturing truck valve fault diagnosis method.

[0041] The fracturing truck valve fault diagnosis system, which integrates mechanism analysis and data-driven methods, includes: The signal acquisition module is used to acquire vibration signals from the valve body of the fracturing truck. The processor is configured to execute the steps of the above-described fracturing truck valve fault diagnosis method. The output module is used to output diagnostic results and maintenance decisions.

[0042] All relevant content of each step involved in the aforementioned embodiments of the fracturing truck valve failure diagnosis method can be referenced in the embodiments of this application, and will not be repeated here.

[0043] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the above-described fracturing truck valve fault diagnosis method.

[0044] Through the above description of the embodiments, those skilled in the art can clearly understand that the present invention can be implemented by means of software plus necessary general-purpose hardware, or it can be implemented by special-purpose hardware including application-specific integrated circuits, dedicated CPUs, dedicated memory, dedicated components, etc. Generally, any function performed by a computer program can be easily implemented by corresponding hardware, and the specific hardware structure used to implement the same function can be diverse, such as analog circuits, digital circuits, or dedicated circuits. However, for the present invention, software program implementation is more often the preferred implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a readable storage medium, such as a computer floppy disk, USB flash drive, portable hard disk, read-only memory (ROM), random access memory (RAM), magnetic disk, or optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0045] Although embodiments of the present invention have been disclosed above, they are not limited to the applications listed in the specification and embodiments. They can be applied to various fields suitable for the present invention. For those skilled in the art, other modifications can be easily made. Therefore, without departing from the general concept defined by the claims and their equivalents, the present invention is not limited to the specific details and illustrations shown and described herein.

Claims

1. A method for diagnosing valve failures in fracturing trucks that integrates mechanism analysis and data-driven approaches, characterized in that: Includes the following steps: S1. Collect the vibration signal of the valve body of the fracturing truck, determine the bandpass filtering range using the adaptive spectral kurtosis analysis method, and perform narrowband filtering on the vibration signal; S2. Perform Hilbert envelope demodulation on the signal filtered in step S1 to extract the impulse envelope signal; S3. Perform exponential decay integral processing on the envelope signal obtained in step S2 to enhance the periodic impact component and suppress noise; S4. Extract the average kurtosis value from the signal processed in step S1 as the first feature, extract the peak factor and impulse index from the signal processed in step S3 as the second and third features, and weight and fuse the first feature, the second feature and the third feature to obtain the comprehensive feature index. S5. Compare the comprehensive characteristic index with a preset first threshold. If the comprehensive characteristic index is greater than or equal to the first threshold, it is determined to be an abnormal wear state; if it is less than the first threshold, it is determined to be a normal wear state and the subsequent steps are executed. S6. Based on the impact envelope signal extracted in step S2, calculate the cumulative impact energy value E per unit time. p and impact pulse repetition density L p ; S7. E calculated based on step S6 p and L p A dual-threshold dynamic evaluation mechanism is used to assess wear status and output diagnostic results and maintenance decisions.

2. The method for diagnosing valve failure in a fracturing truck according to claim 1, characterized in that, In step S6, the cumulative impact energy value E p and impact pulse repetition density L p The calculation method is as follows: ； ； Among them, E pi Let be the energy of the i-th impact pulse detected within time ΔT. This represents the total number of impact pulses detected within time ΔT.

3. The method for diagnosing valve failure in a fracturing truck according to claim 1, characterized in that, In step S7, the dual-threshold dynamic evaluation mechanism specifically includes: S701. Based on the processed data, calculate the cumulative integral of the impact amplitude A(t) and the cumulative integral of the number of impacts C(t); S702. Based on the historical full life cycle operation data of the fracturing truck valve, the cumulative impact amplitude threshold T is preliminarily determined. A and the cumulative threshold of impact number T C And dynamically adjust it in conjunction with the multidimensional normal distribution criterion; S703. If A(t) ≥ T A And C(t) ≥ T C Therefore, it can be preliminarily determined that the valve is in a worn state; S704. Using DS evidence theory, the basic probability assignments generated based on amplitude-impact joint criteria, trend evolution criteria, and pattern recognition criteria are fused to calculate the final wear probability. P wear ; S705. Based on the stated final wear probability P wear Within the preset probability range, execute the corresponding early warning, alarm, or emergency shutdown tiered operation.

4. The method for diagnosing valve failure in a fracturing truck according to claim 3, characterized in that, In step S704, the trend evolution criterion is obtained by predicting the development gradient of future cycles using the ARIMA time series model; the pattern recognition criterion is obtained by classifying signal features based on a deep belief network.

5. The fracturing truck valve fault diagnosis method according to claim 3, characterized in that, In step S705, the grading operation specifically includes: When 70% ≤ P wear When the level is less than 85%, it is determined to be a warning level, triggering a detailed spectrum analysis and activating the backup lubrication system; When 85% ≤ P wear When the percentage is less than 95%, it is determined to be an alarm level, and a maintenance work order is automatically generated. when P wear When the failure rate is ≥ 95%, it is classified as an emergency, and shutdown protection is immediately implemented and the fault tree analysis module is activated.

6. The method for diagnosing valve failure in a fracturing truck according to claim 1, characterized in that, In step S3, the exponential decay kernel function used in the exponential decay integral processing is: h(τ) = e -βτ ·u(τ), where β is the attenuation coefficient, determined by the signal sampling rate and the duration of the impact, and u(τ) is the unit step function.

7. The method for diagnosing valve failure in a fracturing truck according to claim 1, characterized in that, In step S1, the adaptive spectral kurtosis analysis method optimizes the window function parameters to maximize the kurtosis value of the target frequency band, thereby determining the frequency band most sensitive to the impact characteristics as the bandpass filtering range.

8. The method for diagnosing valve failures in fracturing trucks according to claim 1, characterized in that, Steps S6 and S7 further include calculating E p and L p Based on the growth trend and cumulative value throughout the entire life cycle, analyze the trend of physical health decline and predict remaining lifespan.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the fracturing truck valve fault diagnosis method as described in any one of claims 1 to 8.

10. A valve body fault diagnosis system for fracturing trucks, characterized in that, include: The signal acquisition module is used to acquire vibration signals from the valve body of the fracturing truck. The processor is configured to perform the method steps as described in any one of claims 1 to 8; The output module is used to output diagnostic results and maintenance decisions.

Images