Fracturing pump multi-parameter cooperative sensing control system based on fault self-diagnosis
Through a multi-parameter collaborative sensing and control system, the system can accurately monitor and provide early warnings of filter blockage and valve component failures in fracturing pumps, thus overcoming the shortcomings of traditional fracturing pump status monitoring and improving equipment operational stability and efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-11
- Publication Date
- 2026-04-03
AI Technical Summary
Traditional fracturing pump condition monitoring suffers from problems such as filter blockage failing to provide early warnings, inaccurate valve component fault diagnosis, and susceptibility to environmental interference, leading to unstable equipment operation and lost work time.
A multi-parameter collaborative sensing control system based on fault self-diagnosis is adopted. Through high-frequency dynamic pressure sensors, crankshaft encoders and online oil sensors, it can accurately monitor and warn of filter blockage, valve assembly lag and lubrication status. A standard waveform library is constructed for dynamic time warping. Combined with dielectric constant and moisture analysis, multi-dimensional fault diagnosis is achieved.
It enables early warning and accurate fault diagnosis of filter clogging, reduces the risk of equipment failure, improves equipment operation stability and efficiency, and reduces maintenance costs.
Smart Images
Figure CN121782152A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of fracturing pump collaborative control technology, specifically to a multi-parameter collaborative sensing and control system for fracturing pumps based on fault self-diagnosis. Background Technology
[0002] A fracturing pump is a mechanical engineering device that provides a high-pressure fluid power source for oil well fracturing operations. It injects fluid into the oil well at high speed through the pressure of the pump. With the help of the high pressure generated at the bottom of the well, the rock in the oil layer is fractured to create cracks. At the same time, sand with a density greater than that of the formation is mixed into the pumped fluid, allowing it to enter the cracks along with the fluid, keeping the cracks open. This increases the permeability of the formation near the bottom of the well and improves the efficiency of oil and gas extraction.
[0003] As the core power equipment for fracturing operations in oil and gas fields, the operational stability of fracturing pumps directly determines operational efficiency and construction safety. As oil and gas exploration expands to deeper and unconventional areas, fracturing pumps face high-pressure, high-frequency, and long-cycle operating conditions, making the requirements for equipment condition monitoring and fault diagnosis increasingly stringent. In the traditional industry technology system, there are still many technical limitations in the condition monitoring of fracturing pumps. On the one hand, filter clogging monitoring mostly uses traditional differential pressure gauges, which only trigger alarms when the degree of clogging is high, and cannot achieve early warning. There is a lack of dynamic tracking of the clogging process, and the sudden clogging of the filter often leads to pump suction failures, cavitation and other faults. On the other hand, valve component fault diagnosis often relies on vibration monitoring or manual inspection. Vibration signals are easily affected by the on-site environment, making it impossible to accurately distinguish between the specific types of valve opening lag and closing lag, and it is also difficult to correspond to specific fault sources such as spring fatigue and valve seat wear. Often, it is necessary to stop the machine and disassemble it to locate the problem, resulting in a lot of lost operation time.
[0004] To address the aforementioned technical shortcomings, a solution is proposed. Summary of the Invention
[0005] The purpose of this invention is to solve the problems mentioned above by proposing a multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis.
[0006] The objective of this invention can be achieved through the following technical solution: a multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis, comprising a collaborative sensing control platform, wherein the collaborative sensing control platform is communicatively connected to: The progressive clogging warning unit for the filter screen provides early warning of filter screen clogging in the fracturing pump; it determines the suction stroke waveform and provides clogging warning based on the analysis. The valve assembly hysteresis fault diagnosis unit diagnoses hysteresis faults in the valve assembly; based on monitoring of the discharge valve chamber and the suction valve chamber, it identifies hysteresis in the discharge valve and the suction valve. The power-end lubrication condition detection unit detects the condition of the lubricating oil in the power-end gears and bearings.
[0007] Furthermore, the process of the filter progressive clogging warning unit is as follows: Set up a dynamic pressure sensor; acquire continuous voltage / current signals from the sensor; simultaneously connect the crankshaft angle encoder signal from the pump power end; use the rising edge of the encoder pulse as a trigger signal to accurately segment each independent suction stroke waveform segment from the continuous dynamic pressure signal. After determining the corresponding intake stroke waveform segment, the intake pressure waveform segment of each stroke is analyzed; the absolute pressure value of the lowest point in the waveform corresponding to the intake stroke waveform segment is obtained and marked as the trough pressure value; at the same time, the trough point of the intake stroke waveform segment is determined, and the waveform is screened to restore the steady-state pressure point according to the intake pressure waveform segment and marked as the steady-state point; the slope of the line segment obtained by connecting the trough point and the steady-state point is collected and marked as the average slope of the pressure rise segment.
[0008] Furthermore, when the filter is brand new or thoroughly cleaned, hundreds of waveforms under normal operating conditions are collected to construct a standard waveform library; the real-time collected suction stroke waveform is dynamically time-normalized and matched with the standard waveform library; the cumulative distance of the waveform is obtained from the two output waveform sequences, and after normalization, it is marked as the waveform distortion index, ranging from 0 to 1. Meanwhile, during the period when the filter input time continues to increase, a characteristic parameter rule is set, namely: if the short-term trend of the trough pressure value is continuously decreasing, it indicates that the inlet negative pressure is intensifying; if the short-term trend of the average slope of the pressure rise segment is continuously decreasing, it indicates that the fluid replenishment capacity is slowing down.
[0009] Furthermore, the suction stroke waveform was analyzed based on the real-time operation of the fracturing pump: If the waveform distortion index in the real-time suction stroke waveform continuously exceeds the set index threshold A, or if any of the characteristic parameter rules is triggered, an early degradation signal is generated and sent to the collaborative sensing and control platform. After receiving the early degradation signal, the collaborative sensing and control platform continuously monitors the pump suction filter and records all waveform characteristics and waveform distortion index in the background to provide a long-term basis for maintenance decisions. If the waveform distortion index in the real-time suction stroke waveform continuously increases and exceeds the set index threshold B, or if any of the characteristic parameter rules is triggered, a progressive blockage signal is generated and sent to the collaborative sensing control platform. After receiving the progressive blockage signal, the collaborative sensing control platform performs blockage treatment on the pump suction filter.
[0010] Furthermore, the process of the valve assembly hysteresis fault diagnosis unit is as follows: A dynamic pressure sensor is installed in both the discharge valve chamber and the suction valve chamber at each hydraulic end to synchronously collect the pressure signals of both; the pressure signals of the discharge and suction valves at the same hydraulic end are compared and analyzed. Identify the points in each working cycle where the discharge valve pressure exceeds the suction valve pressure, and the points where the suction valve pressure reverses the discharge valve pressure; calculate the phase difference between the actual time and the theoretical piston position.
[0011] Furthermore, if the phase difference of the discharge valve opening is positive and continues to increase, it is inferred that the discharge valve opening is lagging, and a discharge opening lag signal is generated and sent to the collaborative sensing and control platform; if the phase difference of the discharge valve closing is negative and its absolute value increases, it is inferred that the discharge valve closing is lagging, and a discharge closing lag signal is generated and sent to the collaborative sensing and control platform. If the phase difference of the inhalation valve opening is positive and continues to increase, it is inferred that the inhalation valve opening is lagging, and an inhalation opening lag signal is generated and sent to the collaborative sensing and control platform; if the phase difference of the inhalation valve closing is negative and its absolute value increases, it is inferred that the inhalation valve closing is lagging, and an inhalation closing lag signal is generated and sent to the collaborative sensing and control platform.
[0012] Furthermore, the process of the power-end lubrication condition detection unit is as follows: An online oil sensor array is installed on the lubricating oil circulation loop; the dielectric constant and moisture content of the lubricating oil are read in real time; the rate of change of the dielectric constant is calculated, and a time series correlation analysis is performed with the changes in moisture content and particle number; a floating curve of dielectric constant-moisture-contamination degree is established. Based on the floating curve, the stage of moisture content increase was obtained, and it was divided into a rapid stage and a gradual stage according to the rate of increase; the range of synchronous increase of dielectric constant value in the rapid stage was obtained, and the synchronous duration of the increasing trend of dielectric constant value in the gradual stage was also obtained. The synchronous increase span of the dielectric constant value during the rapid phase and the synchronous duration of the dielectric constant value showing an increasing trend during the gradual phase are compared with the threshold for the increase span and the synchronous duration threshold, respectively.
[0013] Furthermore, if the synchronous increase in dielectric constant value exceeds the threshold for the increase in value during the rapid phase, or if the synchronous duration of the increasing dielectric constant value exceeds the threshold for the synchronous duration during the gradual phase, it is inferred that severe moisture intrusion has occurred, generating an abnormal moisture signal and sending it to the collaborative sensing and control platform. If the synchronous increase in dielectric constant value during the rapid phase does not exceed the threshold for the increase in value, and the synchronous duration of the increasing trend in dielectric constant value during the smooth phase does not exceed the threshold for the synchronous duration, then a normal moisture signal is generated and sent to the collaborative sensing and control platform.
[0014] Furthermore, if the moisture value does not fluctuate during the plateau phase, but the dielectric constant continues to drift slowly in one direction and the particle number increases normally, an abnormal state signal is generated and sent to the collaborative sensing and control platform. If the moisture value does not fluctuate during the plateau phase, and the dielectric constant fluctuation is strongly correlated with the sudden change in the particle number, it is judged as a surge in solid pollutants, and an abnormal pollutant signal is generated and sent to the collaborative sensing and control platform. Strong correlation means that the interval between the fluctuation of the dielectric constant and the corresponding particle number values is lower than the set interval threshold, and the fluctuation frequency deviation is lower than the set frequency deviation threshold.
[0015] Compared with the prior art, the beneficial effects of the present invention are: 1. High-frequency, high-precision dynamic pressure sensors are installed at specific locations on the suction pipeline of each pump. These sensors accurately capture pressure pulsations, providing a highly reliable data source for subsequent waveform analysis and avoiding signal distortion caused by improper sensor placement or insufficient accuracy. The crankshaft angle encoder signal is integrated, and its pulse rising edge is used as a trigger to segment the suction stroke waveform, achieving a precise correspondence between the waveform and the piston stroke. This ensures the spatiotemporal consistency of each analysis unit and eliminates interference from irrelevant signals. Key characteristic parameters such as trough pressure values and the average slope of the pressure rise segment are extracted, allowing for a characterization of the fluid state during the suction process from both the magnitude and rate of pressure change, providing a basis for blockage detection. Core quantitative indicators: A standard waveform library is constructed, and the Dynamic Time Warping (DTW) algorithm is used for waveform matching. This effectively solves the problem of time axis scaling caused by fluctuations in operating conditions in real-time waveforms, improving the accuracy of waveform comparison. The normalization of the waveform distortion index provides an intuitive and quantifiable standard for the degree of waveform deviation. A dual-dimensional judgment rule (waveform distortion index threshold + characteristic parameter trend) is set to achieve early identification of filter clogging and accurate triggering of progressive clogging warnings. At the same time, the waveform data recorded in the background provides detailed historical evidence for long-term maintenance decisions, avoiding false alarms and missed alarms due to single-indicator judgment, and realizing the transformation from passive maintenance to preventive monitoring.
[0016] 2. By synchronously installing dynamic pressure sensors in the hydraulic end suction and discharge valve chambers, pressure change signals of the suction and discharge valves within the same working cycle can be acquired, providing a basis for comparison in determining the valve opening and closing times. By identifying the theoretical moment of pressure crossover between the suction and discharge valves and calculating the phase difference based on the piston position calculated from the crankshaft encoder signal, the valve opening and closing states are precisely correlated with the piston mechanical movement, achieving a quantitative characterization of valve action lag. Separate judgment logics for opening and closing phase differences are applied to the discharge and suction valves, enabling precise differentiation of different types of valve component faults. For example, lag in discharge valve opening corresponds to spring fatigue or valve disc jamming, while lag in closing corresponds to valve seat wear or foreign objects. This precise fault attribution can directly guide targeted on-site repairs, avoiding the blindness of traditional fault diagnosis. Real-time feedback of various lag signals to the collaborative sensing and control platform triggers corresponding control actions, allowing for intervention in the early stages of a fault to prevent further deterioration and significantly reducing the risk of hydraulic end efficiency reduction or shutdown of the fracturing pump due to valve component failure.
[0017] 3. An online oil sensor group consisting of a capacitive dielectric constant sensor, a moisture sensor, and a particle counter is deployed in the lubricating oil circulation loop to achieve real-time monitoring of multi-dimensional indicators of lubrication status, breaking the timeliness limitations of traditional offline sampling and testing. Time series correlation analysis of dielectric constant, moisture content, and particle number is performed to establish floating curves, which can uncover the correlation patterns between various indicators and avoid the one-sidedness of single indicator monitoring. The rapid and gradual stages are divided according to the rate of increase of moisture content, and the judgment threshold of dielectric constant change is set accordingly, which can accurately determine the severity of moisture intrusion and quickly locate the root cause of moisture abnormalities such as cooling water leakage. Furthermore, the system can distinguish between lubricating oil oxidation / additive decay problems corresponding to unidirectional drift of dielectric constant (no moisture floating) and the surge of solid contaminants that are strongly correlated with the change of dielectric constant fluctuation and particle number, so as to accurately identify the causes of lubrication deterioration and change lubricating oil maintenance from "periodic replacement" to "on-demand treatment", which reduces maintenance costs and avoids wear of key components such as gears and bearings due to lubrication failure. Attached Figure Description
[0018] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings.
[0019] Figure 1 This is a system principle block diagram of the present invention; Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation
[0020] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0021] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of the invention. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.
[0022] Please see Figure 1 As shown, the multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis includes a collaborative sensing control platform, wherein the collaborative sensing control platform is communicatively connected to a filter progressive blockage early warning unit, a valve assembly hysteresis fault diagnosis unit, and a power end lubrication status detection unit. The collaborative sensing and control platform generates a progressive clogging warning signal for the filter screen and sends it to the progressive clogging warning unit for the filter screen. After receiving the progressive clogging warning signal, the progressive clogging warning unit issues a filter screen clogging warning for the fracturing pump. Please see Figure 2 As shown, a dynamic pressure sensor is installed, that is, a high-frequency, high-precision dynamic pressure sensor (such as piezoelectric or capacitive type) is installed on the suction line of each pump, after the filter and before the pump inlet valve, to ensure that it can accurately capture pressure pulsations; at the same time, a crankshaft position encoder is installed. The sensor continuously acquires voltage / current signals; simultaneously, it receives crankshaft angle encoder signals (or piston displacement signals) from the pump power end; using the rising edge of the encoder pulse (corresponding to the start of the piston suction stroke) as a trigger signal, it accurately segments each independent suction stroke waveform from the continuous dynamic pressure signal. After determining the corresponding intake stroke waveform segment, the intake pressure waveform segment of each stroke is analyzed; the absolute pressure value of the lowest point in the waveform corresponding to the intake stroke waveform segment is obtained and marked as the trough pressure value; at the same time, the trough point of the intake stroke waveform segment is determined, and the waveform is screened to restore the steady-state pressure point according to the intake pressure waveform segment, and marked as the steady-state point; the slope of the line segment obtained by connecting the trough point and the steady-state point is collected and marked as the average slope of the pressure rise segment; When the filter is brand new or thoroughly cleaned, hundreds of waveforms under normal operating conditions are collected to build a standard waveform library. The real-time acquired inhalation stroke waveform is matched with a standard waveform library using dynamic time warping (DTW). The DTW algorithm can effectively align and compare two waveform sequences that have slight stretching on the time axis. The cumulative distance of the waveforms is obtained from the two output waveform sequences. After normalization, it is marked as the waveform distortion index, ranging from 0 to 1. This index directly reflects the overall deviation between the real-time waveform and the healthy waveform.
[0023] Meanwhile, during the period when the filter screen input time continues to increase, characteristic parameter rules are set, namely: if the short-term trend of the trough pressure value is continuously decreasing, it indicates that the inlet negative pressure is intensifying; if the short-term trend of the average slope of the pressure rise segment is continuously decreasing, it indicates that the fluid replenishment capacity is slowing down. The suction stroke waveform was analyzed based on the real-time operation of the fracturing pump. If the waveform distortion index in the real-time suction stroke waveform continuously exceeds the set index threshold A, or if any of the characteristic parameter rules is triggered, an early degradation signal is generated and sent to the collaborative sensing and control platform. After receiving the early degradation signal, the collaborative sensing and control platform continuously monitors the pump suction filter and records all waveform characteristics and waveform distortion index in the background to provide a long-term basis for maintenance decisions. If the waveform distortion index in the real-time suction stroke waveform increases continuously and exceeds the set index threshold B, or if any of the characteristic parameter rules is triggered, a progressive blockage signal is generated and sent to the collaborative sensing control platform. After receiving the progressive blockage signal, the collaborative sensing control platform performs blockage treatment on the pump suction filter. The collaborative sensing and control platform generates a valve assembly hysteresis fault diagnosis signal and sends it to the valve assembly hysteresis fault diagnosis unit. After receiving the valve assembly hysteresis fault diagnosis signal, the valve assembly hysteresis fault diagnosis unit performs valve assembly hysteresis fault diagnosis. A dynamic pressure sensor is installed in both the discharge valve chamber and the suction valve chamber at each hydraulic end to synchronously collect the pressure signals of both; the pressure signals of the discharge and suction valves at the same hydraulic end are compared and analyzed. Identify the point in each work cycle where the discharge valve pressure exceeds the suction valve pressure (theoretical opening point) and the point in each work cycle where the suction valve pressure exceeds the discharge valve pressure (theoretical closing point). Calculate the phase difference (time difference) between the actual time and the theoretical piston position (calculated from the crankshaft encoder signal). If the phase difference of the discharge valve opening is positive and continues to increase, it is inferred that the discharge valve opening is lagging, a discharge opening lag signal is generated and sent to the collaborative sensing and control platform. After receiving the signal, the collaborative sensing and control platform performs discharge valve opening control. In actual scenarios, the phenomenon is spring fatigue or valve disc jamming. If the phase difference of the discharge valve closing is negative and the absolute value increases, it is inferred that the discharge valve closing is delayed, a discharge closing delay signal is generated and sent to the collaborative sensing and control platform. The collaborative sensing and control platform receives the signal and performs discharge valve closing control. In actual scenarios, the phenomenon is valve seat wear or foreign objects. If the phase difference of the inhalation valve opening is positive and continues to increase, it is inferred that the inhalation valve opening is lagging, an inhalation opening lag signal is generated and sent to the collaborative sensing and control platform, and the collaborative sensing and control platform receives it and performs inhalation valve opening control. If the phase difference between the closing of the inhalation valve is negative and its absolute value increases, it is inferred that the closing of the inhalation valve is delayed, an inhalation closing delay signal is generated and sent to the collaborative sensing and control platform, and the collaborative sensing and control platform receives the signal and performs inhalation valve closing control. The collaborative sensing and control platform generates a power-end lubrication status detection signal and sends it to the power-end lubrication status detection unit; after receiving the power-end lubrication status detection signal, the power-end lubrication status detection unit performs status detection on the lubricating oil of the power-end gears and bearings; An online oil sensor group, including a capacitive dielectric constant sensor, a moisture (ppm) sensor, and a particle counter, is installed on the lubricating oil circulation loop; the dielectric constant and moisture content of the lubricating oil are read in real time; the rate of change of the dielectric constant is calculated, and a time-series correlation analysis is performed with the changes in moisture content and particle number; a floating curve of dielectric constant-moisture-contamination degree is established. Based on the floating curve, the stage of moisture content increase was obtained, and it was divided into a rapid stage and a gradual stage according to the rate of increase; the range of synchronous increase of dielectric constant value in the rapid stage was obtained, and the synchronous duration of the increasing trend of dielectric constant value in the gradual stage was also obtained. The synchronous duration of the synchronous increase in dielectric constant value during the rapid phase and the synchronous duration of the increasing trend in dielectric constant value during the gradual phase are compared with the threshold for the increase in value and the threshold for the synchronous duration, respectively: If the dielectric constant value increases synchronously by more than the threshold during the rapid phase, or if the synchronous duration of the increasing dielectric constant value exceeds the threshold during the gradual phase, it is inferred that moisture has seriously intruded, generating a moisture anomaly signal and sending it to the collaborative sensing and control platform. After receiving the moisture anomaly signal, the collaborative sensing and control platform issues a moisture content warning and implements control measures. In actual scenarios, this manifests as cooling water leakage. If the synchronous increase in dielectric constant value during the rapid phase does not exceed the threshold for the increase in value, and the synchronous duration of the increasing trend in dielectric constant value during the smooth phase does not exceed the threshold for the synchronous duration, then a normal moisture signal is generated and sent to the collaborative sensing and control platform. If the moisture content does not fluctuate during the plateau phase, but the dielectric constant continues to drift slowly in one direction (beyond the baseline band) and the particle count increases normally, it is determined that the lubricating oil base oil is oxidized / additives degrade, resulting in a qualitative change in performance. This generates an abnormal status signal and sends it to the collaborative sensing and control platform. After receiving the abnormal status signal, the collaborative sensing and control platform replaces the lubricating oil. If the moisture content does not fluctuate during the plateau phase, and the dielectric constant fluctuation is strongly correlated with the sudden change in particle number, it is judged as a surge in solid contaminants. An abnormal contaminant signal is generated and sent to the collaborative sensing and control platform. After receiving the abnormal contaminant signal, the collaborative sensing and control platform reduces the contaminant content inside the lubricating oil. A strong correlation is indicated by the interval between the fluctuation of the dielectric constant and the corresponding particle number values being lower than the set interval threshold, and the fluctuation frequency deviation being lower than the set frequency deviation threshold.
[0024] Thresholds, preset values, or preset ranges are set for result comparison and analysis to determine whether they are good or bad. The value of these thresholds is determined by a combination of large-scale model analysis of sample data and human experience. They can also be adjusted appropriately based on seasonal or rational factors.
[0025] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.
Claims
1. A multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis, characterized in that, This includes a collaborative sensing and control platform, whose communication connections include: The progressive clogging warning unit for the filter screen provides early warning of filter screen clogging in the fracturing pump; it determines the suction stroke waveform and provides clogging warning based on the analysis. The valve assembly hysteresis fault diagnosis unit diagnoses hysteresis faults in the valve assembly; based on monitoring of the discharge valve chamber and the suction valve chamber, it identifies hysteresis in the discharge valve and the suction valve. The power-end lubrication condition detection unit detects the condition of the lubricating oil in the power-end gears and bearings.
2. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 1, characterized in that, The process of the filter progressive clogging warning unit is as follows: Set up a dynamic pressure sensor; acquire continuous voltage / current signals from the sensor; simultaneously connect the crankshaft angle encoder signal from the pump power end; use the rising edge of the encoder pulse as a trigger signal to accurately segment each independent suction stroke waveform segment from the continuous dynamic pressure signal. After determining the corresponding intake stroke waveform segment, the intake pressure waveform segment of each stroke is analyzed; the absolute pressure value of the lowest point in the waveform corresponding to the intake stroke waveform segment is obtained and marked as the trough pressure value; at the same time, the trough point of the intake stroke waveform segment is determined, and the waveform is screened to restore the steady-state pressure point according to the intake pressure waveform segment and marked as the steady-state point; the slope of the line segment obtained by connecting the trough point and the steady-state point is collected and marked as the average slope of the pressure rise segment.
3. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 2, characterized in that, When the filter is brand new or thoroughly cleaned, hundreds of waveforms under normal operating conditions are collected to build a standard waveform library; the real-time collected suction stroke waveform is dynamically time-normalized and matched with the standard waveform library; the cumulative distance of the waveform is obtained from the two output waveform sequences, and after normalization, it is marked as the waveform distortion index, ranging from 0 to 1; Meanwhile, during the period when the filter input time continues to increase, a characteristic parameter rule is set, namely: if the short-term trend of the trough pressure value is continuously decreasing, it indicates that the inlet negative pressure is intensifying; if the short-term trend of the average slope of the pressure rise segment is continuously decreasing, it indicates that the fluid replenishment capacity is slowing down.
4. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 3, characterized in that, Analysis of the suction stroke waveform based on the real-time operation of the fracturing pump: If the waveform distortion index in the real-time suction stroke waveform continuously exceeds the set index threshold A, or if any of the characteristic parameter rules is triggered, an early degradation signal is generated and sent to the collaborative sensing and control platform. After receiving the early degradation signal, the collaborative sensing and control platform continuously monitors the pump suction filter and records all waveform characteristics and waveform distortion index in the background to provide a long-term basis for maintenance decisions. If the waveform distortion index in the real-time suction stroke waveform continuously increases and exceeds the set index threshold B, or if any of the characteristic parameter rules is triggered, a progressive blockage signal is generated and sent to the collaborative sensing control platform. After receiving the progressive blockage signal, the collaborative sensing control platform performs blockage treatment on the pump suction filter.
5. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 1, characterized in that, The process of the valve assembly hysteresis fault diagnosis unit is as follows: A dynamic pressure sensor is installed in the discharge valve chamber and the suction valve chamber of each hydraulic end to synchronously collect the pressure signals of both; the pressure signals of the discharge and suction valves at the same hydraulic end are compared and analyzed. Identify the points in each working cycle where the discharge valve pressure exceeds the suction valve pressure, and the points where the suction valve pressure reverses the discharge valve pressure; calculate the phase difference between the actual time and the theoretical piston position.
6. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 5, characterized in that, If the phase difference of the discharge valve opening is positive and continues to increase, it is inferred that the discharge valve opening is lagging, and a discharge opening lag signal is generated and sent to the collaborative sensing and control platform; if the phase difference of the discharge valve closing is negative and the absolute value increases, it is inferred that the discharge valve closing is lagging, and a discharge closing lag signal is generated and sent to the collaborative sensing and control platform. If the phase difference of the inhalation valve opening is positive and continues to increase, it is inferred that the inhalation valve opening is lagging, and an inhalation opening lag signal is generated and sent to the collaborative sensing and control platform; if the phase difference of the inhalation valve closing is negative and its absolute value increases, it is inferred that the inhalation valve closing is lagging, and an inhalation closing lag signal is generated and sent to the collaborative sensing and control platform.
7. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 1, characterized in that, The process of the power-end lubrication condition detection unit is as follows: An online oil sensor array is installed on the lubricating oil circulation loop; the dielectric constant and moisture content of the lubricating oil are read in real time; the rate of change of the dielectric constant is calculated, and a time series correlation analysis is performed with the changes in moisture content and particle number; a floating curve of dielectric constant-moisture-contamination degree is established. Based on the floating curve, the stage of moisture content increase was obtained, and it was divided into a rapid stage and a gradual stage according to the rate of increase; the range of synchronous increase of dielectric constant value in the rapid stage was obtained, and the synchronous duration of the increasing trend of dielectric constant value in the gradual stage was also obtained. The synchronous duration of the synchronous increase in dielectric constant during the rapid phase and the synchronous duration of the increasing trend in dielectric constant during the gradual phase are compared with the threshold for the increase in dielectric constant and the threshold for the synchronous duration, respectively.
8. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 7, characterized in that, If the synchronous increase in dielectric constant value exceeds the threshold for the increase in value during the rapid phase, or if the synchronous duration of the increasing dielectric constant value exceeds the threshold for the synchronous duration during the gradual phase, it is inferred that severe moisture intrusion has occurred, and an abnormal moisture signal is generated and sent to the collaborative sensing and control platform. If the synchronous increase in dielectric constant value during the rapid phase does not exceed the threshold for the increase in value, and the synchronous duration of the increasing trend in dielectric constant value during the smooth phase does not exceed the threshold for the synchronous duration, then a normal moisture signal is generated and sent to the collaborative sensing and control platform.
9. The multi-parameter collaborative sensing control system for fracturing pumps based on fault self-diagnosis according to claim 8, characterized in that, If the moisture value does not fluctuate during the plateau phase, but the dielectric constant continues to drift slowly in one direction and the particle number increases normally, an abnormal state signal is generated and sent to the collaborative sensing and control platform. If the moisture value does not fluctuate during the plateau phase, but the dielectric constant fluctuation is strongly correlated with the sudden change in the particle number, it is judged as a surge in solid pollutants, and an abnormal pollutant signal is generated and sent to the collaborative sensing and control platform. Strong correlation means that the interval between the fluctuation of the dielectric constant and the corresponding particle number values is lower than the set interval threshold, and the fluctuation frequency deviation is lower than the set frequency deviation threshold.