Intelligent sensing modular ball valve and control method

Abstract

The present invention relates to the field of ball valve technology, provides an intelligent sensing modular ball valve and control method, and designs a valve body module and a control module. The valve body module integrates multi-modal sensors, including a spirally distributed acoustic sensor array, a physical quantity sensor group, and a valve core status detection sensor, to monitor the fluid acoustic characteristics, pressure, temperature, flow, liquid level, and valve core displacement and torque in real time. The control module analyzes the sensor data through a data fusion algorithm to achieve early identification of cavitation faults and microcrack fault detection. The system uses an 8-bit dynamic coding mechanism to accurately locate the fault type and adjust the actuator opening or torque in a linked manner. The present invention has the capabilities of multi-parameter collaborative perception, early fault warning, and adaptive control, which significantly improves the reliability and safety of the ball valve under complex working conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of ball valves, and in particular relates to an intelligent sensing modular ball valve and a control method thereof. Background Art

[0002] As a key actuator in the industrial fluid control field, smart ball valves achieve real-time monitoring and precise adjustment of medium parameters such as pressure, flow, and temperature through integrated sensing, driving, and intelligent control units. Compared to traditional ball valves, smart ball valves can autonomously perform fault diagnosis, status warnings, and adaptive adjustment functions through embedded sensor networks and control systems, significantly improving the automation level of pipeline systems. Existing smart ball valves mostly use a structure that combines single physical quantity sensors such as pressure and flow with electric actuators, and achieve basic closed-loop control through PLCs or industrial controllers. However, technical bottlenecks still exist in multi-dimensional status perception and early fault warning under complex working conditions.

[0003] Traditional solutions rely on a single type of sensor for condition monitoring, lacking a coordinated detection mechanism for multiple failure modes, such as cavitation and mechanical fatigue. The weak acoustic signature of cavitation in its infancy is easily drowned out by fluid noise, while the torque fluctuations and structural resonance frequency shifts associated with microcrack initiation exhibit nonlinear coupling. Existing single-parameter detection methods struggle to achieve accurate early identification, leading to delayed maintenance and increased component damage.

[0004] Detection of valve body microcracks often relies on regular manual inspections or vibration spectrum analysis, which cannot monitor the cumulative effects of fatigue under alternating stress in real time. Existing technologies lack a coupling model between the standard deviation of torque fluctuations and the resonant frequency shift, resulting in insufficient accuracy in crack initiation probability calculations. This makes it difficult to trigger early warnings at the microscopic damage stage, increasing the risk of sudden structural failure.

[0005] Traditional fault coding systems often use independent alarm identifiers and lack a mechanism for integrating multidimensional data. When cavitation and mechanical anomalies occur simultaneously, existing coding rules cannot accurately characterize the combined fault type and priority, complicating host computer analysis and delaying emergency response decisions.

[0006] Existing smart ball valves often use a single-circuit design for their power supply and alarm modules, which can easily lead to critical data loss during power fluctuations or communication interruptions. Especially in high-risk operating conditions, the lack of an independent power supply alarm circuit can render early warning signals ineffective, potentially causing serious safety incidents.

[0007] In summary, existing intelligent ball valve technology has significant limitations in multimodal perception, early fault diagnosis, multi-source data fusion, and system reliability. There is an urgent need for an intelligent perception solution that integrates high-sensitivity acoustic detection, structural damage probability modeling, and dynamic coding decision-making to comprehensively improve valve status monitoring and fault response capabilities in complex industrial environments. Summary of the Invention

[0008] To solve the problems existing in the background technology, the present invention provides an intelligent sensing modular ball valve, including a valve body module and a control module, wherein:

[0009] The valve body module includes a valve core, a valve body, a drive mechanism and a multimodal sensor. The drive mechanism includes an electric actuator. The multimodal sensor includes an acoustic sensor array, a physical quantity sensor group and a valve core state detection sensor.

[0010] The acoustic sensor array consists of three MEMS acoustic sensors, which are embedded in the inner wall of the valve body in a spiral distribution. They collect fluid acoustic pressure signals and connect to the control module through the SPI interface.

[0011] The physical quantity sensor group is arranged at the inlet or outlet of the valve body, including a pressure sensor, a temperature sensor, a flow sensor and a liquid level sensor, and is connected to the PLC processor through a standard industrial interface;

[0012] The valve core status detection sensor integrates a displacement sensor and a Hall sensor to monitor the valve core opening and closing degree and the electric actuator torque respectively, and is connected to the main controller through the CAN bus;

[0013] The control module includes a main controller and a PLC processor, both of which are powered by a 24V power line and interconnected via a data transmission line. The main controller is integrated with a data fusion module and an acoustic feature library, and the control module is connected to the host computer for communication.

[0014] In a preferred embodiment, the control module further includes an alarm and a backup power supply. The alarm is connected to the GPIO port of the main controller through an independent line, and the backup power supply is connected to the 24V power line and the alarm power supply circuit respectively through a dual redundant circuit.

[0015] In a preferred embodiment, the three MEMS acoustic sensors of the acoustic sensor array are distributed at a spiral angle of 15°-30°, the axial spacing between adjacent sensors is 0.3-0.5 times the valve body flow channel diameter, the installation depth of each sensor is 1 / 3 of the valve body wall thickness and is evenly distributed at a circumferential interval of 120°.

[0016] The present invention provides a control method for an intelligent sensing modular ball valve, comprising the following steps:

[0017] S1. Multimodal data acquisition: The acoustic sensor array collects the real-time acoustic pressure signal of the fluid in the valve body. The physical quantity sensor group obtains pressure, temperature, flow, and liquid level data. The valve core state detection sensor monitors the valve core opening and closing degree and the electric actuator torque.

[0018] S2. Early Identification of Cavitation Faults: Based on the spectral characteristics of the acoustic pressure signal from the acoustic sensor array, the critical energy mutation threshold at the initial stage of cavitation is calculated to identify the initial cavitation signal.

[0019] S3. Microcrack Fault Detection: A crack initiation probability model is constructed by combining displacement data from the valve core status detection sensor with the resonance frequency shift from the acoustic sensor array.

[0020] S4 multi-sensor data fusion: The abnormal signal of steps S1-S3 is input into the data fusion module of the main controller to generate a fault type code;

[0021] S5. Fault and alarm: Trigger the alarm and send an alarm signal to the host computer based on abnormal conditions, including cavitation detection abnormalities and microcrack detection abnormalities.

[0022] In a preferred solution, the specific steps of step S2 include:

[0023] S21, the acoustic sensor array is embedded in the valve body wall in a spiral distribution, with a spiral angle range of 15°-30°; the spacing between adjacent sensors is 0.3-0.5 times the valve body flow diameter; the three sensors are evenly spaced 120° apart, and the installation depth is 1 / 3 of the valve body wall thickness; the sound pressure signal is transmitted to the main controller and wavelet packet decomposition is performed using the Daubechies4 wavelet basis function;

[0024] S22. High-frequency energy calculation: Extract the energy value of the 8-20kHz frequency band. The calculation formula is:

[0025]

[0026] in: : kth wavelet packet decomposition coefficient; N: total number of high-frequency sub-bands, default N=5; when the flow sensor detects that the flow rate is greater than 80% of the rated value, N is adjusted to 8;

[0027] S23. Energy mutation rate determination: Calculate the energy mutation rate of adjacent sampling periods:

[0028]

[0029] in: : High-frequency energy value at the current moment; : High-frequency energy value at the last sampling moment; : Sampling interval, the default is 0.1 seconds, synchronized with the PLC processor clock;

[0030] S24. Cavitation initiation judgment rules: When If the sampling period exceeds the limit for three consecutive times, the initial alarm will be triggered;

[0031] Under complex working conditions, that is, when the pressure sensor detects a pressure fluctuation of ±10%, or the flow sensor detects a sudden flow change of >30%, the threshold is adjusted to 20%;

[0032] S25, frequency domain feature matching: The main controller calls the cavitation noise spectrum template pre-stored in the acoustic feature library with a frequency resolution of 1kHz and calculates the current spectrum matching degree:

[0033] Match

[0034] Where: S(f): current spectrum amplitude; T(f): template spectrum amplitude;

[0035] When the matching degree is >90%, cavitation initiation is confirmed and a fault code is generated.

[0036] In a preferred solution, the specific steps of step S3 include:

[0037] S31, sensor data acquisition: The valve core status detection sensor monitors the valve core displacement and electric actuator torque in real time, including:

[0038] The displacement sensor has an accuracy of ±0.1mm and a measuring range of 0-100mm;

[0039] The Hall sensor torque range is 0-500N·m, and the sampling frequency is 1kHz;

[0040] Acoustic sensor array detects valve body resonant frequency offset , frequency resolution 0.5Hz;

[0041] S32, Torque Fluctuation Analysis: Torque Sampling Value Based on Hall Sensor , calculate the standard deviation of torque fluctuation within the window period:

[0042]

[0043] in: : torque value of the i-th sampling; : Average torque value within the window period; M: Sampling window length, the default value is 50, which corresponds to a 50ms time window;

[0044] S33, frequency shift detection: The acoustic sensor array calculates the current resonant frequency through the FFT analysis module of the main controller With reference frequency Offset:

[0045]

[0046] in: : The resonant frequency of the valve body without cracks is provided by the factory calibration value;

[0047] S34, crack probability calculation: combined and Construct a crack initiation probability model:

[0048]

[0049] in: : material fatigue coefficient; : Coupling parameter of torque fluctuation and frequency shift;

[0050] S35, fault judgment and output: When When the fault occurs, it is determined to be micro-crack initiation; the main controller sends the fault code 0x1F to the alarm through the CAN bus and adjusts the driving torque of the electric actuator to a safety threshold, which is reduced to 70% of the rated value by default.

[0051] In a preferred solution, the generation logic of the fault type code in step S4 includes the following steps:

[0052] S41. Coding structure definition: Define 8-bit binary fault code, format: 0bXXXXXXXX, the meaning of each bit is as follows:

[0053] Bit0-1, indicates the cavitation level;

[0054] 00: normal state;

[0055] 01: Cavitation is born, indicating the corresponding ;

[0056] 10: Severe cavitation, indicating the corresponding ;

[0057] Bit2-4, indicates abnormal physical quantity;

[0058] 001: Abnormal pressure, indicating that the pressure sensor exceeds the limit by ±15%;

[0059] 010: Temperature abnormality, indicating that the temperature sensor is > set value ±10℃;

[0060] 100: Flow abnormality, indicating that the flow sensor deviates from the set value by ±20%;

[0061] Bit5, indicating abnormal liquid level;

[0062] 0: normal;

[0063] 1: The detection value of the liquid level sensor is lower than the safety threshold;

[0064] Bit6-7, indicating the valve core status;

[0065] 00: Normal;

[0066] 01: The valve core opening and closing degree is out of tolerance, indicating that the displacement sensor detection error is >±5%;

[0067] 10: The electric actuator torque exceeds the limit, indicating that it is greater than the rated value;

[0068] S42. Formulate dynamic code generation rules: The main controller receives sensor data in real time through the data fusion module and updates the code according to the following logic:

[0069] When the pressure sensor detection value exceeds , default 0.2MPa; or , by default, when the pressure is 10MPa, Bit2 is forcibly set; when the liquid level sensor data fails, that is, when the signal is lost for more than 3 seconds, Bit5 is set to 1 and the backup power supply is activated; when the displacement data of the valve core status detection sensor deviates from the set value by more than ±5%, Bit6-7 are set to 01.

[0070] In a preferred solution, in step S5, the specific steps of processing the abnormal cavitation detection and the abnormal microcrack detection include:

[0071] S54. Cavitation detection abnormality handling:

[0072] S541, trigger condition: the acoustic sensor array detects the energy mutation rate of the sound pressure signal , and the acoustic feature library matching degree> 90%;

[0073] S542, alarm execution: the main controller sends the cavitation level code through the data transmission line; the alarm triggers a high-frequency buzzer, which continues until the cavitation disappears;

[0074] S543, linkage control: automatically reduce the opening of the electric actuator to 70% of the current value;

[0075] S55. Microcrack detection abnormality handling:

[0076] S551, Trigger condition: Crack initiation probability model output ;

[0077] S552, alarm execution: The main controller sends the fault code 0x1F, and the alarm starts the sound and light alarm; the torque of the electric actuator is limited to 50% of the rated value; the host computer locates and records the crack position through the valve core displacement sensor data.

[0078] The beneficial effects achieved by the present invention are:

[0079] First, the present invention integrates an acoustic sensor array, a physical quantity sensor group, and a valve core status detection sensor into the valve body module, creating an independent control module. This simplifies system maintenance and upgrades. The collaborative operation of these multiple sensors comprehensively senses fluid, mechanical, and acoustic conditions, significantly improving fault detection coverage. Data from different sensors efficiently interacts via a standard interface, enhancing the system's adaptability to diverse operating conditions and avoiding the risk of missed detections by a single sensor.

[0080] Second, the present invention utilizes a MEMS acoustic sensor array with a spiral pitch angle layout. Combined with wavelet packet decomposition and an acoustic feature library matching algorithm, the invention designs a spiral distribution to optimize signal spatial coverage, dynamically adjust the high-frequency energy mutation threshold, and implement a spectrum template matching mechanism. The spiral layout enhances the spatial resolution of acoustic signals, effectively capturing the local sound pressure characteristics of cavitation incipient events. The combination of wavelet packet decomposition and dynamic thresholding improves the sensitivity of weak cavitation signals. Frequency-domain feature matching significantly reduces the false alarm rate under complex noise interference, providing early warning of cavitation failures.

[0081] Third, the present invention constructs a crack initiation probability model based on the coupling parameters of valve core torque fluctuation and resonant frequency shift. This model incorporates dynamic calculation of torque fluctuation standard deviation, real-time resonant frequency shift detection, and exponential probability function modeling. This overcomes the limitations of traditional single-parameter detection and enables early prediction of microcrack initiation through correlation analysis between mechanical stress and structural dynamic response. A probabilistic threshold trigger mechanism adapts to different material properties, avoiding over-maintenance or missed inspections and extending the service life of key valve components.

[0082] Fourth, the present invention designs an 8-bit binary dynamic coding rule, integrates multi-dimensional information such as cavitation level, physical quantity anomaly and valve core status, and designs a specific scheme for real-time coding bit dynamic update, priority logic coverage and fusion decision-making, which realizes the precise positioning and priority sorting of multiple fault types and simplifies the analysis process of complex faults; the data fusion mechanism enhances the reliability of the diagnostic results through weighted integration of multi-source evidence, which is particularly suitable for multiple concurrent fault scenarios and provides a clear basis for operation and maintenance decisions.

[0083] Fifth, the present invention adopts a dual-channel redundant power supply and independent GPIO alarm circuit design, combined with torque limiting and opening adaptive adjustment strategies, and designs specific solutions for seamless switching of supercapacitors, independent triggering of sound and light alarms, and dynamic optimization of drive parameters under fault conditions, ensuring the continuous operation of key sensors and controllers under power outages and ensuring the reliable transmission of alarm signals under extreme working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0084] Figure 1 This is a schematic diagram of the intelligent sensing modular ball valve structure module of the present invention;

[0085] Figure 2 This is a structural module framework diagram of the intelligent sensing modular ball valve of the present invention;

[0086] Figure 3 Schematic diagram of the layout structure of the acoustic sensor array of the present invention;

[0087] Figure 4 It is a flow chart of the intelligent perception modular ball valve control method of the present invention.

[0088] Numbers in the figure:

[0089] 1. Valve core; 2. Electric actuator; 3. Pressure sensor; 4. Temperature sensor; 5. Flow sensor; 6. Liquid level sensor; 7. Valve core status detection sensor; 8. Acoustic sensor array; 9. 24V power cord; 10. Data transmission line; 11. PLC processor; 12. Main controller; 13. Alarm; 14. Backup power supply. DETAILED DESCRIPTION

[0090] The technical solutions of the present invention will be described clearly and completely below in conjunction with the drawings in the present invention. In addition, the forms of the various structures described in the following embodiments are merely examples. The present invention is not limited to the various structures described in the following embodiments. All other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0091] Reference Figure 1-Figure 4 The present invention provides an intelligent sensing modular ball valve, including a valve body module and a control module, wherein the valve body module includes a valve core 1, a valve body, a driving mechanism and a multimodal sensor, the driving mechanism includes an electric actuator 2, and the multimodal sensor includes an acoustic sensor array 8, a physical quantity sensor group and a valve core state detection sensor 7; the acoustic sensor array 8 is composed of three MEMS acoustic sensors, which are embedded in the inner wall of the valve body according to a spiral distribution, collect fluid sound pressure signals and are connected to the control module through an SPI interface; the physical quantity sensor group is arranged at the inlet or outlet of the valve body, including a pressure sensor Sensor 3, temperature sensor 4, flow sensor 5 and liquid level sensor 6 are connected to the PLC processor 11 through a standard industrial interface; the valve core state detection sensor 7 integrates a displacement sensor and a Hall sensor, which respectively monitor the opening and closing degree of the valve core 1 and the torque of the electric actuator 2, and is connected to the main controller 12 through the CAN bus; the control module includes the main controller 12 and the PLC processor 11, both of which are powered by a 24V power line 9 and interconnected via a data transmission line 10. The main controller 12 is integrated with a data fusion module and an acoustic feature library, and the control module is connected to the host computer for communication.

[0092] The control module further includes an alarm 13 and a backup power supply 14. The alarm 13 is connected to the GPIO port of the main controller 12 through an independent line, and the backup power supply 14 is connected to the 24V power line 9 and the power supply circuit of the alarm 13 through a dual redundant circuit.

[0093] The three MEMS acoustic sensors of the acoustic sensor array 8 are distributed at a helix angle of 15°-30°, the axial spacing between adjacent sensors is 0.3-0.5 times the valve body flow channel diameter, the installation depth of each sensor is 1 / 3 of the valve body wall thickness and is evenly distributed at 120° circumferential intervals.

[0094] The present invention designs a control method for an intelligent sensing modular ball valve, comprising the following steps:

[0095] S1. Multimodal data acquisition: Acoustic sensor array 8 collects real-time acoustic pressure signals of the fluid in the valve body, and physical quantity sensor group obtains pressure, temperature, flow, and liquid level data. The valve core state detection sensor 7 monitors the opening and closing degree of the valve core 1 and the torque of the electric actuator 2;

[0096] S2. Early identification of cavitation faults: Based on the spectral characteristics of the sound pressure signal from the acoustic sensor array 8, the critical energy mutation threshold at the initial stage of cavitation is calculated to identify the initial cavitation signal;

[0097] S3. Microcrack fault detection: Combining the displacement data of the valve core state detection sensor 7 and the resonance frequency shift of the acoustic sensor array 8, a crack initiation probability model is constructed;

[0098] S4 multi-sensor data fusion: The abnormal signal of steps S1-S3 is input into the data fusion module of the main controller 12 to generate a fault type code;

[0099] S5. Fault and alarm: trigger the alarm 13 according to the fault code and send an alarm signal to the host computer, including abnormal fluid parameters, abnormal valve core movement, abnormal control module power failure, abnormal cavitation detection and abnormal microcrack detection.

[0100] The specific steps of step S2 include:

[0101] S21, the acoustic sensor array 8 is embedded in the valve body wall in a spiral distribution, with a spiral angle range of 15°-30°; the spacing between adjacent sensors is 0.3-0.5 times the valve body flow channel diameter; the three sensors are evenly spaced 120° apart, and the installation depth is 1 / 3 of the valve body wall thickness; the sound pressure signal is transmitted to the main controller 12, and wavelet packet decomposition is performed using the Daubechies4 wavelet basis function;

[0102] S22. High-frequency energy calculation: Extract the energy value of the 8-20kHz frequency band. The calculation formula is:

[0103]

[0104] in: : kth wavelet packet decomposition coefficient; : Total number of high-frequency sub-bands, the default N=5; when flow sensor 5 detects a flow rate > 80% of the rated value, N is adjusted to 8;

[0105] S23. Energy mutation rate determination: Calculate the energy mutation rate of adjacent sampling periods:

[0106]

[0107] in: : High-frequency energy value at the current moment; : The last sampling moment ( Front) high frequency energy value;

[0108] : Sampling interval, default is 0.1 seconds, synchronized with the clock of the PLC processor (11);

[0109] S24. Cavitation initiation judgment rules: When If the sampling period exceeds the limit for three consecutive times, the initial alarm will be triggered;

[0110] Under complex working conditions, i.e., when the pressure sensor (3) detects a pressure fluctuation of ±10%, or the flow sensor (5) detects a flow mutation of >30%, the threshold value is adjusted to 20%;

[0111] S25, frequency domain feature matching: the main controller (12) calls the cavitation noise spectrum template pre-stored in the acoustic feature library with a frequency resolution of 1kHz and calculates the current spectrum matching degree:

[0112]

[0113] in: : current spectrum amplitude; : template spectrum amplitude;

[0114] When the matching degree is >90%, cavitation initiation is confirmed and fault code Bit0-1=01 is generated.

[0115] The specific steps of step S3 include:

[0116] S31, sensor data acquisition: the valve core state detection sensor 7 monitors the displacement of the valve core 1 and the torque of the electric actuator 2 in real time, wherein:

[0117] The displacement sensor has an accuracy of ±0.1mm and a measuring range of 0-100mm;

[0118] The Hall sensor torque range is 0-500N·m, and the sampling frequency is 1kHz;

[0119] Acoustic sensor array 8 detects the valve body resonance frequency offset , frequency resolution 0.5Hz;

[0120] S32, Torque Fluctuation Analysis: Torque Sampling Value Based on Hall Sensor (Unit: N·m), calculate the standard deviation of torque fluctuation within the window period:

[0121]

[0122] in: : torque value of the i-th sampling (i=1,2,...,M); : average torque value within the window period (unit: N·m); : Sampling window length, the default value is 50 (corresponding to a 50ms time window);

[0123] S33, frequency shift detection: The acoustic sensor array 8 calculates the current resonance frequency through the FFT analysis module of the main controller 12 With reference frequency Offset:

[0124]

[0125] in: : The resonant frequency of the valve body without cracks is provided by the factory calibration value;

[0126] S34, crack probability calculation: combined and Construct a crack initiation probability model:

[0127]

[0128] in: : Material fatigue coefficient (experimental calibration value, Q235 steel takes 0.0023, 304 stainless steel takes 0.0018); : coupling parameter of torque fluctuation and frequency shift (unit: N·m·Hz);

[0129] S35, fault judgment and output: When When the fault occurs, it is determined that micro cracks have occurred; the main controller (12) sends a fault code 0x1F to the alarm (13) via the CAN bus, and adjusts the driving torque of the electric actuator (2) to a safety threshold, which is reduced to 70% of the rated value by default.

[0130] The generation logic of the fault type code in step S4 includes the following steps:

[0131] S41. Coding structure definition: Define 8-bit binary fault code, format: 0bXXXXXXXX, the meaning of each bit is as follows:

[0132] Bits 0-1 indicate the cavitation level; 00: Normal; 01: Initial cavitation, corresponding to ΔE > 15%; 10: Severe cavitation, corresponding to ΔE > 30%; Bits 2-4 indicate physical quantity abnormalities; 001: Pressure abnormality, indicating that pressure sensor 3 exceeds the limit by ±15%; 010: Temperature abnormality, indicating that temperature sensor 4 exceeds the set value by ±10°C; 100: Flow abnormality, indicating that flow sensor 5 deviates from the set value by ±20%; Bit 5 indicates liquid level abnormality; 0: Normal; 1: The detection value of liquid level sensor 6 is lower than the safety threshold; Bits 6-7 indicate the valve core status; 00: Normal; 01: Valve core 1 opening and closing degree exceeds the tolerance, indicating that the displacement sensor detection error is > ±5%; 10: Electric actuator 2 torque exceeds the limit, indicating that it exceeds the rated value of 500N·m;

[0133] S42, formulating dynamic code generation rules: The main controller 12 receives sensor data in real time through the data fusion module and updates the code according to the following logic:

[0134] When the pressure sensor 3 detects a value exceeding (Default 0.2MPa) or (default 10MPa), Bit2 is forcibly set; when the data of liquid level sensor 6 fails, Bit5 is set to 1 and the backup power supply 14 is activated; when the displacement data of valve core status detection sensor 7 deviates from the set value by more than ±5%, Bit6-7 are set to 01.

[0135] The specific steps of step S5 include: S51. Fluid parameter abnormal alarm determination:

[0136] S511, pressure abnormality determination:

[0137]

[0138] in: : Real-time measurement value of pressure sensor 3 (unit: MPa); : preset pressure threshold (unit: MPa);

[0139] S512, temperature abnormality determination:

[0140] According to the resistance value of temperature sensor 4 (Unit: Ω) Calculate real-time temperature:

[0141]

[0142] Abnormal conditions: When ℃ or ℃ The alarm is triggered when is the preset temperature threshold;

[0143] S513, flow abnormality determination:

[0144] The pulse frequency of the flow sensor (5) (Unit: Hz) Calculate flow rate:

[0145]

[0146] Abnormal conditions: When When the alarm is triggered, To set the flow rate;

[0147] S514, liquid level abnormality determination:

[0148] Liquid level sensor 6 outputs analog voltage signal (Unit: V), converted to liquid level height:

[0149]

[0150] Abnormal conditions: When The low level alarm is triggered when is the full scale of the sensor);

[0151] S52. Abnormal valve core action processing: S521, abnormal detection: valve core state detection sensor 7 real-time monitoring: displacement accuracy: ± 0.1mm (range 0-100mm); torque range: 0-500N m (Hall sensor linearity ± 0.5%); when the displacement deviation or torque When , an exception is triggered;

[0152] S522, closed-loop control intervention: the main controller 12 adjusts the drive current of the electric actuator 2:

[0153]

[0154] in: : Reference current (4-20mA); ; (torque deviation);

[0155] S53. Control module power failure exception handling: S531, power switching logic: backup power supply 14 monitors the voltage of 24V power line 9 ,when When: switch to super capacitor group (capacity 10F, power supply ≥ 30 seconds); trigger the power failure alarm signal of the independent line; S532, calculate the pulse width of the alarm (13):

[0156]

[0157] in: (Alarm equivalent resistance); (Energy storage capacitor);

[0158] Output pulse width , drive the buzzer to alarm intermittently;

[0159] S533, forced code upload: the main controller (12) forces the fault code to be 0xFF and sends it cyclically via the CAN bus;

[0160] S54. Cavitation detection abnormality handling:

[0161] S541, trigger condition: the acoustic sensor array 8 detects the energy mutation rate of the sound pressure signal , and the acoustic feature library matching degree is greater than 90%; S542, alarm execution: the main controller 12 sends the cavitation level code through the data transmission line 10; the alarm 13 triggers a high-frequency buzzer, which continues until the cavitation disappears; S543, linkage control: automatically reduce the opening of the electric actuator 2 to 70% of the current value, according to the formula:

[0162]

[0163] in The current opening and closing angle of valve core 1;

[0164] S55. Microcrack detection abnormality processing; S551. Trigger condition: crack initiation probability model output ; S552, alarm execution: the main controller 12 sends the fault code 0x1F, the alarm 13 starts the sound and light alarm; the torque of the electric actuator 2 is limited to 50% of the rated value:

[0165]

[0166] in is the rated torque; the host computer locates and records the crack position through the valve core displacement sensor data.

[0167] The design background of the technical solution of the present invention is described below. Construction and matching method of acoustic feature library: The acoustic feature library is used to store predefined cavitation noise spectrum templates and is constructed by the following steps: Simulate working conditions of different cavitation levels in the laboratory and use a high-frequency hydrophone to collect the sound pressure signal of the fluid in the valve body. The frequency range of the hydrophone is 0.1 Hz to 100 kHz, the sampling rate is set to 256 kHz, and the acquisition time for each working condition is not less than 60 seconds. Perform one-third octave analysis on the signal to extract the spectrum characteristics of the frequency band from 8 kHz to 20 kHz, including the main peak frequency, the amplitude ratio of the harmonic components and the spectrum centroid. The spectrum centroid is calculated by weighted average frequency. The characteristic parameters are stored in the FLASH memory of the main controller in the form of a table corresponding to frequency and amplitude, with a frequency resolution of 1 kHz, an amplitude point is stored in each interval, and the data format is 32-bit floating point number. The matching degree between the real-time spectrum and the template is calculated by the normalized correlation coefficient. When the matching degree exceeds 90%, it is determined to be a cavitation incipient state.

[0168] Daubechies4 wavelet basis function and wavelet packet decomposition implementation method: Wavelet basis selection: Daubechies4 wavelet has a fourth-order vanishing moment, and its low-pass filter coefficients include values ​​such as 0.1629, 0.5055, and 0.4461, and the high-pass filter coefficients are corresponding mirror-symmetric values. The sound pressure signal is decomposed into five layers of wavelet packets to generate 32 sub-bands, and the high-frequency sub-bands corresponding to 8 kHz to 20 kHz in the fifth layer are extracted. Five sub-bands are selected by default to calculate the energy value, and when the flow exceeds 80% of the rated value, it is expanded to eight sub-bands. Through simulation tests, the energy reconstruction error of the wavelet basis in the frequency band of 8 kHz to 20 kHz is less than plus or minus 2%.

[0169] Data Fusion Module Algorithm Implementation: Data fusion utilizes the DS evidence theory, with the following specific rules: Evidence Source Weighting: The cavitation energy mutation rate is weighted at 0.6, and the spectrum matching degree is weighted at 0.4. Probability Assignment: The basic probability of cavitation initiation is calculated based on the ratio of the energy mutation rate to the threshold, and the spectrum matching probability is calculated based on the ratio of the real-time matching degree to the threshold. Fusion Decision: A cavitation alarm is triggered when the joint probability value is greater than or equal to 0.8.

[0170] Calibration of the material fatigue coefficient: Standard Q235 steel specimens were prepared and subjected to alternating torque in a fatigue testing machine with an amplitude range of 0 to 500 N·m and a frequency of 5 Hz for at least one million cycles. The standard deviation of the torque fluctuation and the resonant frequency shift at crack initiation were recorded. Parameter fitting: Using Weibull distribution curve fitting, the fatigue coefficients of Q235 steel were determined to be 0.0023, and those of 304 stainless steel to be 0.0018. Scanning electron microscopy revealed that the product of the torque fluctuation and the frequency shift was linearly positively correlated with the crack length.

[0171] The communication protocol for the standard industrial interface defines a physical quantity sensor group connected to the PLC processor via the Modbus RTU protocol. Communication parameters: RS-485 differential transmission is used, with a baud rate of 9600 bits per second. The data frame format includes the address code, function code, starting register address, number of registers, and CRC checksum. Troubleshooting: The PLC sends a heartbeat packet every 200 milliseconds. If there is no response three times in a row, the main controller switches to the backup CAN bus channel.

[0172] GPIO Port Connection and Alarm Triggering Logic: Hardware Design: The alarm is connected to the main controller's GPIO pin via an optocoupler isolation circuit. A 1 kilohm current-limiting resistor is connected in series with the optocoupler input, and the output drives a 5V buzzer. Trigger condition: When the GPIO output is high at 3.3V, the optocoupler turns on, and the buzzer sounds at a 2 kHz frequency with a 50% duty cycle.

[0173] Dual redundant power supply switching mechanism: Circuit topology: The primary power supply and backup supercapacitor bank are connected in parallel via MOSFETs, with a voltage comparator monitoring the primary power supply voltage. Switchover condition: When the primary power supply voltage falls below 18V, the supercapacitor bank automatically switches in. The capacitor bank consists of six Maxwell 350F capacitors connected in series, with a total capacity of 58.3F, and can maintain power for 30 seconds at full load.

[0174] Crack initiation probability model: Based on fracture mechanics theory, the crack growth rate is related to the magnitude of the stress intensity factor. In this model, the product of the standard deviation of torque fluctuation and the resonant frequency shift reflects the change in the stress intensity factor. Therefore, an exponential function is used to simulate the cumulative effect of crack initiation probability.

[0175] Experimental basis for setting dynamic coding thresholds: Pressure threshold: According to the American Petroleum Institute (API) 6D standard, ball valve sealing fails when pressure exceeds the set value by ±15%. Burst tests show that at this point, the helium leakage rate exceeds 5 liters per minute. Flow rate mutation threshold: According to the ISO 5208 standard, a flow rate mutation exceeding 30% will trigger a water hammer effect. Simulation analysis determined that the energy mutation rate threshold should be raised to 20%.

[0176] Calibration and compensation of the resonant frequency reference value: Calibration method: A hammer impact modal test is used. After striking the valve body with a hammer, the fundamental frequency is extracted through a data acquisition system. The fundamental frequency of a typical Q235 steel valve body is 1.2 kHz, with an allowable error of plus or minus 5%. Temperature compensation: A stored temperature compensation coefficient of minus 0.05 Hz per degree Celsius is used. During real-time calculations, the reference frequency is corrected based on the temperature sensor's measurement value. Cavitation is caused by abnormal fluid pressure, while microcracks are caused by mechanical fatigue or stress concentration. Both are diagnosed early through acoustic characteristics and structural dynamic response, respectively.

[0177] Cavitation failure: Cavitation occurs when fluid flows through a valve, resulting in a sudden drop in local pressure below the liquid's saturated vapor pressure, forming cavitation bubbles. These bubbles then collapse in high-pressure areas, generating shock waves that can cause erosion damage to the valve's inner walls or component surfaces. This occurs due to improper valve opening or excessive flow rates, which cause the pressure in constricted passages to fall below the liquid's vapor pressure. Sudden changes in flow or increased turbulence can create localized low-pressure areas. The valve body material cannot withstand the high-frequency impact of the cavitation bubble collapse, accelerating surface erosion.

[0178] Microcrack failure: Microcrack failure occurs when critical valve components, including the valve body and valve core, are subjected to long-term alternating stress or fatigue loads, leading to the initiation and propagation of tiny cracks within the material, ultimately causing structural failure. This occurs due to frequent opening and closing operations or torque fluctuations, which cause cyclical stress on the components exceeding the material fatigue limit. Structural design flaws or manufacturing defects can lead to localized stress increases. Electric actuators can also experience excessive torque, or external shock loads can exceed the component's strength threshold.

[0179] In this invention, cavitation faults are detected by capturing 8-20kHz high-frequency sound pressure signals through an acoustic sensor array, analyzing the energy mutation rate and spectrum matching to identify cavitation collapse characteristics. Microcrack faults are detected by combining valve core displacement deviation, torque fluctuation standard deviation, and resonant frequency offset to calculate the crack initiation probability and determine microcrack formation.

[0180] Example 1: This example implements early warning of cavitation failure based on acoustic feature analysis. Figure 3 The spirally distributed acoustic sensor array shown in the figure has the following specific implementation steps:

[0181] Three MEMS acoustic sensors are embedded in the inner wall of the valve body at a 25° helix angle, with an axial spacing of 0.4 times the flow channel diameter and an installation depth of 1 / 3 the wall thickness. The sensors acquire sound pressure signals via an SPI interface at a 1MHz sampling rate. The signals are decomposed into five layers using the Daubechies4 wavelet basis to extract energy in the 8-20kHz frequency band:

[0182]

[0183] When the flow rate is greater than 80% of the rated value, it is expanded to 8 sub-bands (N=8). Calculate the energy mutation rate:

[0184]

[0185] The default threshold is 15%. In complex working conditions (pressure fluctuation ±10% or flow rate mutation >30%), the threshold is adjusted to 20%. The 32 sets of cavitation spectrum templates pre-stored in the acoustic feature library are called to calculate the normalized correlation coefficient:

[0186]

[0187] A match degree >90% confirms the initiation of cavitation and generates a fault code Bit0-1 = 01. Alarm linkage control: After the alarm is triggered, the main controller automatically reduces the opening of the electric actuator to 70% of the current value, and the alarm starts a 2kHz high-frequency buzzer until the cavitation disappears.

[0188] Example 2: This example achieves early diagnosis of microcracks through dynamic resonance frequency monitoring and torque fluctuation analysis. The implementation process is as follows: the Hall sensor samples the torque value at 1kHz; the acoustic array detects the valve body resonance frequency; characteristic parameter calculation: torque fluctuation standard deviation:

[0189]

[0190] Resonance frequency shift:

[0191]

[0192] Using exponential probability model:

[0193]

[0194] This embodiment uses Q235 steel. crack When the value is greater than 0.7, a crack is detected. The main controller sends the fault code 0x1F to the alarm (Bit 6-7 = 10). The electric actuator torque is limited to 50% of the rated value. The host computer locates the crack based on the valve core displacement data.

[0195] Example 3: This example integrates four types of sensors: pressure, temperature, flow, and liquid level to achieve full parameter monitoring. Abnormal processing logic: Pressure abnormality: When P curren t∉[0.8P set ,1.2P set ], trigger Bit2=001 alarm. Temperature abnormality is calculated by PT100 resistance value:

[0196]

[0197] When the setting value is exceeded by ±10℃, Bit3=010 alarm is triggered. Flow abnormality is calculated based on the pulse frequency:

[0198]

[0199] When the deviation is ±20% from the set value, Bit4=100 alarm is triggered. Abnormal liquid level: Liquid level height conversion formula:

[0200]

[0201] When H<1m (20% of the range), Bit5=1 alarm is triggered.

[0202] Example 4: This example designs multiple protection mechanisms for abnormal valve core movement and system power failure. When the valve core opens and closes abnormally and the displacement sensor detection error is greater than ±5%, the Bit6-7=01 alarm is triggered. The closed-loop control adjustment formula is:

[0203]

[0204] When the torque exceeds 500 N·m, the main controller forcibly disconnects the drive circuit, triggering an alarm at Bit 6-7 = 10. The fault code is cyclically transmitted via the CAN bus.

[0205] Example 5: This example uses 8-bit dynamic coding to achieve accurate fault location. The coding rules of this example are shown in Table 1:

[0206] Table 1 Coding rules table

[0207] Bit definition Trigger Conditions Bit0-1 Cavitation level 01:ΔE>15%10:ΔE>30% Bit2-4 Abnormal physical quantity 001: Pressure 010: Temperature 100: Flow Bit5 Abnormal liquid level 1: Liquid level <20% Bit6-7 Spool status 01: Displacement out of tolerance 10: Torque out of limit

[0208] The coding example of this embodiment is as follows:

[0209] Code 0b01001101: Cavitation Initiation (Bit 0-1 = 01): Cavitation energy mutation rate ΔE > 15% detected with spectral matching > 90%. Abnormal Pressure (Bit 2-4 = 001): The pressure sensor measurement exceeds the preset threshold by ±15%. Low Liquid Level (Bit 5 = 1): The liquid level is less than 20% of the safety threshold. Spool Displacement (Bit 6-7 = 01): The valve opening deviation from the set value is > ±5%. Scenario: High pressure and cavitation cause a sudden drop in the liquid level, resulting in uncontrolled valve spool displacement.

[0210] Code 0b10001010: Severe Cavitation (Bit 0-1 = 10): ΔE > 30%, cavitation has caused severe erosion. Temperature Abnormal (Bit 2-4 = 010): The temperature exceeds the set value by ±10°C. Liquid Level Normal (Bit 5 = 0): The liquid level is within the safe range. Torque Exceeded (Bit 6-7 = 10): The electric actuator torque exceeds the rated value of 500 N·m. Scenario: High temperature causes material expansion, resulting in torque exceeding the limit and accelerating cavitation.

[0211] Code 0b00110000: Cavitation normal. Bit 0-1 = 00: No cavitation characteristics detected. Abnormal flow rate. Bit 2-4 = 100: Flow rate deviates ±20% from the set value. Low liquid level. Bit 5 = 1: Insufficient liquid level triggers an alarm. Normal valve core. Bit 6-7 = 00: Displacement and torque are within safe ranges. Scenario: A sudden change in flow causes the liquid level to drop, but the valve's mechanical state remains intact.

[0212] Code 0b11000001: Cavitation Initiation Bit0-1 = 01: Early cavitation signal detected. Normal Pressure Bit2-4 = 000: No pressure limit exceeded. Normal Liquid Level Bit5 = 0: Liquid Level is stable. Torque Exceeded Bit6-7 = 11: Reserved.

[0213] Code 0b00011111: Microcrack Fault Special Code: 0x1F is sent directly by the main controller, indicating a crack probability >70%. Logical External Code: This code is independent of the 8-bit dynamic rules and is used to quickly respond to structural damage. Torque is limited to 50% of the rated value, and the audible and visual alarms are activated.

[0214] In this embodiment, the cavitation level Bit0-1: 01 / 10 indicates incipient / severe, respectively, triggering a reduction in opening or shutdown. Physical quantity abnormality Bit2-4: Mutually exclusive marks 001=pressure, 010=temperature, 100=flow, and the priority is overwritten in the order of detection. Liquid level and valve core status: low liquid level Bit5=1 is an independent mark; valve core status Bit6-7 distinguishes displacement tolerance 01 or torque limit 10. By combining different abnormal states, the coding system can achieve precise positioning of more than 32 types of faults and guide the control system to perform operations such as load reduction, alarm or emergency shutdown. For example, when cavitation incipient ΔE=18% and pressure limit exceedance occur at the same time, the code 0b01001001 is generated, and the main controller sends a complete fault message including the location identifier to the upper computer via the CAN bus.

[0215] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. The control method of the intelligent sensing modular ball valve is characterized in that: The following steps are involved: S1. Multimodal data acquisition: The acoustic sensor array collects the real-time acoustic pressure signal of the fluid in the valve body. The physical quantity sensor group obtains pressure, temperature, flow, and liquid level data. The valve core state detection sensor monitors the valve core opening and closing degree and the electric actuator torque. S2. Early Identification of Cavitation Faults: Based on the spectral characteristics of the acoustic pressure signal from the acoustic sensor array, the critical energy mutation threshold at the initial stage of cavitation is calculated to identify the initial cavitation signal. The specific steps of step S2 include: S21, the acoustic sensor array is embedded in the valve body wall in a spiral distribution, with a spiral angle range of 15°-30°; the spacing between adjacent sensors is 0.3-0.5 times the valve body flow diameter; the three sensors are evenly spaced 120° apart, and the installation depth is 1 / 3 of the valve body wall thickness; the sound pressure signal is transmitted to the main controller and wavelet packet decomposition is performed using the Daubechies4 wavelet basis function; S22. High-frequency energy calculation: Extract the energy value of the 8-20kHz frequency band. The calculation formula is: ； in: : kth wavelet packet decomposition coefficient; : The total number of high-frequency sub-bands, the default N=5; when the flow sensor detects a flow rate greater than 80% of the rated value, N is adjusted to 8; S23. Energy mutation rate determination: Calculate the energy mutation rate of adjacent sampling periods: ； in: : High-frequency energy value at the current moment; : High-frequency energy value at the last sampling moment; : Sampling interval, the default is 0.1 seconds, synchronized with the PLC processor clock; S24. Cavitation initiation judgment rules: When If the sampling period exceeds the limit for three consecutive times, the initial alarm will be triggered; Under complex working conditions, that is, when the pressure sensor detects a pressure fluctuation of ±10%, or the flow sensor detects a sudden flow change of >30%, the threshold is adjusted to 20%; S25, frequency domain feature matching: The main controller calls the cavitation noise spectrum template pre-stored in the acoustic feature library with a frequency resolution of 1kHz and calculates the current spectrum matching degree: ； in: : current spectrum amplitude; : Template spectrum amplitude; when the matching degree is >90%, cavitation initiation is confirmed and a fault code is generated. S3. Microcrack Fault Detection: A crack initiation probability model is constructed by combining displacement data from the valve core status detection sensor with the resonance frequency shift from the acoustic sensor array. S4 multi-sensor data fusion: The abnormal signal of steps S1-S3 is input into the data fusion module of the main controller to generate a fault type code; S5. Fault and Alarm: Trigger an alarm based on abnormal conditions and send an alarm signal to the host computer. Abnormal conditions include cavitation detection abnormalities and microcrack detection abnormalities.

2. The control method of the intelligent sensing modular ball valve according to claim 1, characterized in that: The specific steps of step S3 include: S31. Sensor data acquisition: The valve core status detection sensor monitors the valve core displacement and electric actuator torque in real time. The displacement sensor has an accuracy of ±0.1mm and a range of 0-100mm. The Hall sensor has a torque range of 0-500N·m and a sampling frequency of 1kHz. The acoustic sensor array detects the valve body resonance frequency offset. , frequency resolution 0.5Hz; S32, Torque Fluctuation Analysis: Torque Sampling Value Based on Hall Sensor , calculate the standard deviation of torque fluctuation within the window period: ； in: : torque value of the i-th sampling; : average torque value within the window period; : Sampling window length, the default value is 50, which corresponds to a 50ms time window; S33, frequency shift detection: The acoustic sensor array calculates the current resonant frequency through the FFT analysis module of the main controller With reference frequency Offset: ； in: : The resonant frequency of the valve body without cracks is provided by the factory calibration value; S34, crack probability calculation: combined and Construct a crack initiation probability model: ； in: : material fatigue coefficient; : Coupling parameter of torque fluctuation and frequency shift; S35, fault judgment and output: When When the fault occurs, it is determined to be micro-crack initiation; the main controller sends the fault code 0x1F to the alarm through the CAN bus and adjusts the driving torque of the electric actuator to a safety threshold, which is reduced to 70% of the rated value by default.

3. The control method of the intelligent sensing modular ball valve according to claim 2, characterized in that: The generation logic of the fault type code in step S4 includes the following steps: S41. Coding structure definition: Define 8-bit binary fault code, format: 0bXXXXXXXX, the meaning of each bit is as follows: Bit0-1, indicates the cavitation level; 00: normal state; 01: Cavitation initiation, corresponding to ΔE>15%; 10: Severe cavitation, corresponding to ΔE>30%; Bit2-4, indicates abnormal physical quantity; 001: Abnormal pressure, indicating that the pressure sensor exceeds the limit by ±15%; 010: Temperature abnormality, indicating that the temperature sensor is > set value ±10℃; 100: Flow abnormality, indicating that the flow sensor deviates from the set value by ±20%; Bit5, indicating abnormal liquid level; 0: normal; 1: The detection value of the liquid level sensor is lower than the safety threshold; Bit6-7, indicating the valve core status; 00: Normal; 01: The valve core opening and closing degree is out of tolerance, indicating that the displacement sensor detection error is >±5%; 10: The electric actuator torque exceeds the limit, indicating that it is greater than the rated value; S42. Formulate dynamic code generation rules: The main controller receives sensor data in real time through the data fusion module and updates the code according to the following logic: When the pressure sensor detection value exceeds , default 0.2MPa; or , by default, when the pressure is 10MPa, Bit2 is forcibly set; when the liquid level sensor data fails, that is, when the signal is lost for more than 3 seconds, Bit5 is set to 1 and the backup power supply is activated; when the displacement data of the valve core status detection sensor deviates from the set value by more than ±5%, Bit6-7 are set to 01.

4. The control method of the intelligent sensing modular ball valve according to claim 3, characterized in that: In step S5, the specific steps for handling abnormal cavitation detection and abnormal microcrack detection include: S54. Cavitation detection abnormality handling: S541, trigger condition: the acoustic sensor array detects the energy mutation rate of the sound pressure signal , and the acoustic feature library matching degree> 90%; S542, alarm execution: the main controller sends the cavitation level code through the data transmission line; the alarm triggers a high-frequency buzzer, which continues until the cavitation disappears; S543, linkage control: automatically reduce the opening of the electric actuator to 70% of the current value; S55. Microcrack detection abnormality handling: S551, Trigger condition: Crack initiation probability model output ; S552, alarm execution: The main controller sends the fault code 0x1F, and the alarm starts the sound and light alarm; the torque of the electric actuator is limited to 50% of the rated value; the host computer locates and records the crack position through the valve core displacement sensor data.

5. An intelligent sensing modular ball valve for implementing the control method according to any one of claims 1 to 4, characterized in that: It includes valve body module and control module, including: The valve body module includes a valve core, a valve body, a drive mechanism and a multimodal sensor. The drive mechanism includes an electric actuator. The multimodal sensor includes an acoustic sensor array, a physical quantity sensor group and a valve core state detection sensor. The acoustic sensor array consists of three MEMS acoustic sensors, which are embedded in the inner wall of the valve body in a spiral distribution. They collect fluid acoustic pressure signals and connect to the control module through the SPI interface. The physical quantity sensor group is arranged at the inlet or outlet of the valve body, including a pressure sensor, a temperature sensor, a flow sensor and a liquid level sensor, and is connected to the PLC processor through a standard industrial interface; The valve core status detection sensor integrates a displacement sensor and a Hall sensor to monitor the valve core opening and closing degree and the electric actuator torque respectively, and is connected to the main controller through the CAN bus; The control module includes a main controller and a PLC processor, both of which are powered by a 24V power line and interconnected via a data transmission line. The main controller is integrated with a data fusion module and an acoustic feature library, and the control module is connected to the host computer for communication.

6. The intelligent sensing modular ball valve according to claim 5, characterized in that: The control module also includes an alarm and a backup power supply. The alarm is connected to the GPIO port of the main controller through an independent line, and the backup power supply is connected to the 24V power line and the alarm power supply circuit respectively through a dual redundant circuit.

7. The intelligent sensing modular ball valve according to claim 6, characterized in that: The three MEMS acoustic sensors of the acoustic sensor array are distributed at a helix angle of 15°-30°, the axial spacing between adjacent sensors is 0.3-0.5 times the valve body flow channel diameter, the installation depth of each sensor is 1 / 3 of the valve body wall thickness and is evenly distributed circumferentially at an interval of 120°.

Images