Variable frequency adjustment train negative pressure cleaning system based on pressure difference perception

Abstract

The invention relates to the technical field of industrial dust removal and pneumatic conveying, and discloses a variable frequency adjustment train negative pressure cleaning system based on differential pressure sensing, which comprises a central control unit, a negative pressure driving unit, a pneumatic compensation unit and a multi-dimensional differential pressure sensing array, the negative pressure driving unit, the gas-solid separation unit, the long-distance conveying pipeline and the flu type suction nozzle unit are sequentially communicated to form a negative pressure suction gas path; the central control unit collects pressure signals and calculates flow resistance fingerprint characteristic values, the flow resistance fingerprint characteristic values are compared with a fingerprint database so as to identify no-load, stable conveying, critical sedimentation and suction port blocking flow pattern states, and accordingly the frequency of the driving unit or the opening degree of the pneumatic compensation unit is adjusted so as to maintain stable suspension conveying of materials or relieve sedimentation blockage. The flow pattern can be accurately identified from the microcosmic flow field level, the pulse airflow is injected through the harmonic waves to damage the deposition layer, the bypass air supply self-healing is utilized to prevent blockage, the operation energy consumption is reduced, and the conveying efficiency is improved.

Description

Technical Field

[0001] This invention relates to the field of industrial dust removal and pneumatic conveying technology, specifically a variable frequency adjustable train negative pressure cleaning system based on differential pressure sensing. Background Technology

[0002] In the railway transportation sector, especially during the transport of bulk materials such as coal, large amounts of residual material and dust easily accumulate in train carriages and related areas. To ensure safe vehicle operation and meet environmental protection requirements, negative pressure cleaning systems are typically used to collect and process residues from the carriage floor and crevices. These systems generally consist of end-point suction nozzles, long-distance conveying pipelines, gas-solid separation devices, and high-pressure blowers, utilizing aerodynamic principles to suspend and transport materials to the dust collection point.

[0003] However, existing train negative pressure cleaning systems have technical limitations in actual operation. Traditional control strategies mostly employ constant power operation or PID feedback regulation based on a simple static negative pressure threshold. This control method lacks the ability to perceive the microscopic flow regime of the gas-solid two-phase flow inside the pipeline and cannot distinguish between complex operating conditions such as no-load airflow, stable suspended flow, and critical settling flow. To prevent pipeline blockage, the system often maintains extremely high wind speed and negative pressure under all operating conditions, resulting in the fan operating at high energy consumption for extended periods. This not only causes serious energy waste but also accelerates the wear and tear of equipment components.

[0004] Furthermore, during long-distance transportation, fluctuations in material physical properties (such as particle size and moisture content) and changes in pipeline resistance can easily cause airflow velocity to drop below the critical suspension velocity, leading to gradual material deposition at the bottom of the pipeline and the formation of dune flows. Existing monitoring methods typically rely on low-frequency sampling or average value readings from pressure transmitters. When an abnormally high pressure is detected, it often indicates that a substantial physical blockage has occurred in the pipeline. Adjusting the pressure at this point, simply increasing the negative pressure, can actually further compact the accumulated material, making it difficult for the airflow to clear the blockage itself. Manual intervention is usually required, affecting the continuity and efficiency of cleaning operations. Simultaneously, conventional suction nozzle designs primarily focus on physical suction, lacking optimized design for flow field pressure gradients. This makes it difficult to acquire high signal-to-noise ratio fluid characteristic signals under complex operating conditions, further limiting the possibility of achieving intelligent adaptive control of the system. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides a variable frequency control train negative pressure cleaning system based on differential pressure sensing. This system solves the problems of existing train negative pressure cleaning systems, which struggle to accurately sense the gas-solid two-phase flow state inside pipelines, resulting in low conveying efficiency, high energy consumption, and susceptibility to blockages.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a variable frequency control train negative pressure cleaning system based on differential pressure sensing, comprising a central control unit, and a negative pressure drive unit, a pneumatic compensation unit, and a multi-dimensional differential pressure sensing array electrically connected to the central control unit. The negative pressure drive unit, the gas-solid separation unit, the long-distance delivery pipeline, and the flu-type suction nozzle unit are sequentially connected along the airflow direction to form a negative pressure suction air path.

[0007] A pneumatic compensation unit is positioned between the flu-type suction nozzle unit and the long-distance delivery pipeline to regulate the auxiliary airflow entering the system. A multi-dimensional differential pressure sensing array is positioned at key nodes in the negative pressure suction air path to collect real-time pressure data within the system. The central control unit is configured to execute the following control logic: First, it acquires the pressure signal output from the multi-dimensional differential pressure sensing array and performs signal preprocessing; then, it performs time-domain statistical analysis on the preprocessed pressure signal to calculate the flow resistance fingerprint characteristic value, which characterizes the flow field; next, it compares the flow resistance fingerprint characteristic value with a preset flow resistance fingerprint database to identify the current flow pattern; finally, it generates control commands based on the identified flow pattern to adjust the operating frequency of the negative pressure drive unit or the valve opening of the pneumatic compensation unit to maintain stable suspension and delivery of materials within the pipeline or to relieve sedimentation and blockage within the pipeline.

[0008] In one specific embodiment, the flu-type suction nozzle unit adopts a double-circular straight tube to single-circular straight tube structure. This structure includes two parallel circular air inlet branches, a Y-shaped manifold throat, and a single-circular outlet straight tube. The two circular air inlet branches are the left and right air inlet branches, which are mirror-symmetrically distributed and their central axes form a predetermined angle in space. The rear ends of the left and right air inlet branches are connected to the inlet end of the Y-shaped manifold throat, and the rear end of the Y-shaped manifold throat is connected to the single-circular outlet straight tube. The sum of the cross-sectional areas of the left and right air inlet branches is greater than the cross-sectional area of ​​the single-circular outlet straight tube, making the Y-shaped manifold throat a throttling region. An inverted V-shaped guide ridge is provided at the intersection of the inner walls of the Y-shaped manifold throat, and the direction of the ridge line of the inverted V-shaped guide ridge is perpendicular to the plane defined by the left and right air inlet branches.

[0009] In the above embodiment, the multidimensional differential pressure sensing array includes a first pressure sensor, a second pressure sensor, and a third pressure sensor. The first pressure sensor is arranged at the inlet edge of the circular intake branch pipe and is configured to collect the ambient pressure at the inlet. The second pressure sensor is arranged at the geometric confluence center of the Y-shaped manifold throat or at a position within a predetermined downstream range and is configured to collect the throat pressure. The third pressure sensor is arranged at the beginning of the straight section of the long-distance delivery pipeline and is configured to collect the stable delivery pressure. An annular turbulence groove is provided on the inner wall of the outlet of the Y-shaped manifold throat, configured to induce the shedding of micro-vortices to enhance the pressure pulsation signal collected by the second pressure sensor.

[0010] Furthermore, the central control unit is internally configured with a model calculation module, a feature extraction module, a state identification module, and an instruction generation module. The model calculation module is configured to calculate the critical suspension velocity based on the material's physical properties and the particle force balance equation, and determine the reference operating frequency of the negative pressure drive unit. The feature extraction module is configured to process the pressure signal to remove the DC component, obtaining a dynamic pressure pulsation signal, and calculating the statistical characteristics of the dynamic pressure pulsation signal. The state identification module is configured to determine, based on the statistical characteristics, whether the system is in an unloaded operating state, a stable suspension conveying state, a critical settling warning state, or a suction port blockage state.

[0011] Specifically, the flow resistance fingerprint feature values ​​calculated by the feature extraction module include turbulence fluctuation intensity feature values ​​and skewness coefficients. The turbulence fluctuation intensity feature values ​​are obtained by calculating the square root of the average of the sum of squares of the differences between pressure sampling points and the average pressure value within the sampling window. The skewness coefficient is obtained by calculating the ratio of the average of the sum of cubes of the differences between pressure sampling points and the average pressure value within the sampling window to the cube of the turbulence fluctuation intensity feature value. The feature extraction module is also configured to perform windowing and weighting processing on the data sequence within the sampling window before calculation to suppress spectral leakage.

[0012] The judgment logic of the state identification module specifically includes: when the turbulent pulsation intensity characteristic value is less than the first threshold and the average negative pressure is lower than the set value, it is judged as an unloaded operation state; when the turbulent pulsation intensity characteristic value is in the medium-high amplitude range and the absolute value of the skewness coefficient is less than the skewness threshold, it is judged as a stable suspended conveying state; when the turbulent pulsation intensity characteristic value increases in the low frequency range and the skewness coefficient shows a positive shift, it is judged as a critical settlement warning state; when the average negative pressure increases to the high vacuum range and the turbulent pulsation intensity characteristic value drops to near zero, it is judged as a suction port blockage state.

[0013] When the status identification module determines that the system is in a critical settlement warning state, the command generation module is configured to generate a composite frequency command containing both DC and AC components. The composite frequency command is configured to superimpose a sinusoidal modulation wave that varies periodically with time onto a reference operating frequency. The modulation amplitude of the sinusoidal modulation wave is limited to a preset percentage range of the reference operating frequency. The frequency of the sinusoidal modulation wave is configured to match the acoustic natural frequency of the piping system between the flu-type suction nozzle unit and the gas-solid separation unit.

[0014] Preferably, the acoustic natural frequency is obtained through an acoustic modal sweep calibration program during the system initialization phase. The acoustic modal sweep calibration program includes: controlling the vector inverter in the negative pressure drive unit to superimpose a disturbance signal with linearly increasing frequency on the reference operating frequency; synchronously monitoring the pressure fluctuation amplitude response at the end of the long-distance delivery pipeline; identifying the disturbance frequency point corresponding to the maximum value of the pressure fluctuation amplitude, and locking the frequency point as the acoustic natural frequency.

[0015] When the state identification module determines that the system is in a blocked suction port state, the command generation module is configured to generate an opening signal to drive the bypass air supply valve of the pneumatic compensation unit to open. The bypass air supply valve adjusts its opening degree according to the target flow rate in the long-distance delivery pipeline, introducing auxiliary airflow to maintain the flow velocity in the pipeline above the critical suspension velocity. When the feature extraction module detects the recovery of the turbulent pulsation signal from the pressure sensor located at the throat of the flu-type suction nozzle unit, the command generation module controls the bypass air supply valve to gradually close.

[0016] Furthermore, for systems employing a dual-circular straight pipe to single-circular straight pipe structure, the multi-dimensional differential pressure sensing array includes independent pressure sensors in both the left and right intake branches. The central control unit is configured to monitor the absolute value of the pressure difference between the left and right intake branches. When the absolute value of the pressure difference continuously exceeds the imbalance threshold, the central control unit controls the negative pressure drive unit to execute an emergency stop and rapid rise sequence. The emergency stop and rapid rise sequence is configured to cause the fan speed to drop significantly within a preset time and then immediately increase to full speed, inducing reverse vortices and pressure shock waves at the Y-shaped confluence throat.

[0017] This invention provides a variable frequency control train negative pressure cleaning system based on differential pressure sensing. It has the following beneficial effects: 1. This invention acquires high-frequency pressure signals through a multi-dimensional differential pressure sensing array and calculates the turbulence pulsation intensity and skewness coefficient to construct a flow resistance spectral fingerprint. This enables accurate identification of flow patterns such as no-load, stable conveying, critical settling, and suction port closure. Compared with the traditional control method that only relies on the absolute negative pressure threshold, this method can capture the dynamic changes of gas-solid two-phase flow at the microscopic flow field level, effectively avoiding misjudgments caused by signal hysteresis. This allows the system to adjust the fan operating frequency in real time according to the actual working conditions, reducing operating energy consumption while ensuring the reliability of material conveying.

[0018] 2. This invention utilizes a harmonic injection pulse generation strategy. When critical material settling is detected, the negative pressure drive unit is driven to generate a pulsating airflow that matches the natural acoustic frequency of the pipeline. This pulsating airflow creates unsteady shear stress and acoustic resonance effects inside the pipeline, which can effectively disrupt the boundary balance of the particle layer deposited at the bottom of the pipe, causing it to resuspend and enter the mainstream area. This solves the material deposition problem in long-distance transportation with low energy consumption and improves the system's transportation efficiency and anti-clogging ability.

[0019] 3. This invention combines the structural features of the suction nozzle from a double-circle straight pipe to a single-circle straight pipe, and is equipped with differential backwashing and bypass air replenishment self-healing logic. When a blockage is detected on one side of the suction nozzle or a system blockage is detected, the inverter is controlled to execute an emergency stop and speed-up sequence or open the bypass valve. The pressure shock wave generated by fluid inertia is used to loosen the accumulated material or to introduce auxiliary airflow to maintain the conveying speed. This pneumatic self-cleaning and self-healing mechanism does not require additional mechanical unblocking components, preventing the entire pipe section from being blocked due to suction port blockage, and ensuring the continuity and safety of cleaning operations. Attached Figure Description

[0020] Figure 1 This is a schematic diagram of the overall system structure of the present invention; Figure 2 This is a flowchart of the two-phase flow resistance spectrum fingerprint extraction and flow pattern identification of the present invention; Figure 3 This is a logic block diagram of the pulse pneumatic delivery and adaptive control strategy of the present invention; Figure 4 This is a flowchart illustrating the overall system operation of the present invention.

[0021] The components include: 1. Influenza-type suction nozzle unit; 11. Circular air inlet branch pipe; 12. Y-shaped manifold throat; 13. Single circular air outlet straight pipe; 2. Multi-dimensional differential pressure sensing array; 21. First pressure sensor; 22. Second pressure sensor; 23. Third pressure sensor; 3. Long-distance delivery pipeline; 4. Pneumatic compensation unit; 41. Bypass air supply valve; 5. Gas-solid separation unit; 6. Negative pressure drive unit; 61. Centrifugal fan; 62. Vector frequency converter; 7. Central control unit; 71. Data acquisition module; 72. Model calculation module; 73. Feature extraction module; 74. Status identification module; 75. Command generation module. Detailed Implementation

[0022] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. 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.

[0023] See attached document Figure 1 This invention provides a variable frequency controlled train negative pressure cleaning system based on differential pressure sensing. This system is mainly used for negative pressure recovery of bulk materials (such as coal particles) in train carriages and related areas. The system is physically connected in sequence along the airflow direction, including: an influenza-type suction nozzle unit 1, a multi-dimensional differential pressure sensing array 2, a long-distance conveying pipeline 3, a pneumatic compensation unit 4, a gas-solid separation unit 5, a negative pressure drive unit 6, and a central control unit 7.

[0024] The flu-type suction nozzle unit 1, as the end effector of the system, is configured to contact or approach the surface to be cleaned to draw in a gas-solid mixture. In this embodiment, the flu-type suction nozzle unit 1 adopts a double-circular straight pipe to single-circular straight pipe structure. This structure includes two parallel circular inlet branches 11, the rear ends of which are connected to a Y-shaped confluence throat 12, and the rear end of the Y-shaped confluence throat 12 is connected to a single-circular outlet straight pipe 13. The sum of the cross-sectional areas of the circular inlet branches 11 is greater than the cross-sectional area of ​​the single-circular outlet straight pipe 13, so that the Y-shaped confluence throat 12 constitutes a throttling region for airflow acceleration and turbulent mixing. In another embodiment, the flu-type suction nozzle unit 1 is configured as an inverted cone single-pipe structure, characterized in that the diameter of the air inlet is larger than the diameter of the air outlet, and includes a conical acceleration section with a preset contraction angle in the middle. The above-mentioned specific geometric structure is not only used for material guidance, but also serves as a physical generator of fluid flow patterns, used to generate differentiated flow field pressure characteristics under different gas-solid mixing ratios.

[0025] The multi-dimensional differential pressure sensing array 2 consists of multiple high-frequency pressure transmitters used to acquire absolute pressure signals from key nodes of the system in real time. The multi-dimensional differential pressure sensing array 2 includes a first pressure sensor 21, a second pressure sensor 22, and a third pressure sensor 23. The first pressure sensor 21 is installed at the edge of the inlet of the circular air intake branch pipe 11 of the flu-type mouthpiece unit 1, configured to collect the ambient pressure at the inlet. The second pressure sensor 22 is installed at the narrowest point of the contraction section of the Y-shaped manifold throat 12 or the inverted cone single-tube structure, and is configured to collect the throat pressure at the point of highest flow velocity. The third pressure sensor 23 is installed at the beginning of the straight section of the long-distance conveying pipeline 3, and is configured to collect stable conveying pressure. The first pressure sensor 21, the second pressure sensor 22, and the third pressure sensor 23 are all electrically connected to the signal input terminal of the central control unit 7 via shielded signal lines, and the sampling frequency is set to no less than 500Hz in order to capture the high-frequency pulsating component of the airflow.

[0026] One end of the long-distance delivery pipe 3 is connected to the single-circle straight air outlet pipe 13 of the flu-type nozzle unit 1, and the other end extends to the gas-solid separation unit 5. The long-distance delivery pipe 3 is configured as a combination of flexible wear-resistant hose or rigid metal pipe to provide a delivery channel for gas-solid two-phase flow. The pneumatic compensation unit 4 is located at the connection between the flu-type nozzle unit 1 and the long-distance delivery pipe 3. The pneumatic compensation unit 4 includes a bypass air supply valve 41, which is an electrically proportional regulating valve. Its air inlet is connected to the atmospheric environment, and its air outlet is connected to the inside of the delivery pipe. The control terminal of the bypass air supply valve 41 is electrically connected to the control output terminal of the central control unit 7 to adjust the auxiliary air flow rate entering the long-distance delivery pipe 3 according to the command.

[0027] The gas-solid separation unit 5 is connected to the end of the long-distance conveying pipeline 3 and is configured to separate the recovered material (such as coal particles) from the airflow and allow it to settle into the dust collection box. The negative pressure drive unit 6 is connected to the outlet end of the gas-solid separation unit 5, providing adsorption and conveying power for the entire system. The negative pressure drive unit 6 includes a centrifugal fan 61 and a vector frequency converter 62. The intake port of the centrifugal fan 61 is connected to the gas-solid separation unit 5, and the exhaust port is open to the atmosphere or a silencer. The vector frequency converter 62 is electrically connected to the motor of the centrifugal fan 61 and is configured to adjust the motor speed according to the input frequency command. The vector frequency converter 62 has a dynamic torque response function and can respond to composite frequency commands containing both DC and AC components.

[0028] The central control unit 7 is configured as the core processing center of the system, and typically employs a programmable logic controller (PLC) or an industrial control computer. The central control unit 7 is electrically connected to the multi-dimensional differential pressure sensing array 2, the pneumatic compensation unit 4, and the negative pressure drive unit 6. Internally, the central control unit 7 includes a data acquisition module 71, a model calculation module 72, a feature extraction module 73, a state identification module 74, and an instruction generation module 75.

[0029] The data acquisition module 71 is configured to receive analog pressure signals from the multidimensional differential pressure sensing array 2 and convert them into digital pressure sequences. The model calculation module 72 internally stores a critical transport model based on particle stress analysis. This critical transport model is used to calculate the critical suspension velocity required to maintain a specific material in a suspended state within the long-distance transport pipeline 3. The model calculation module 72 calculates according to the following formula. : ; in: Represents gravitational acceleration; Represents the average equivalent particle size of the material being cleaned (coal particles); Represents the density of the material; Represents the density of the transported medium (air); This represents the aerodynamic drag coefficient. Model calculation module 72 calculates the aerodynamic drag coefficient based on this value. The reference operating frequency of the negative pressure drive unit 6 is determined by the cross-sectional area parameters of the long-distance conveying pipeline 3.

[0030] The feature extraction module 73 is configured to perform time-domain statistical analysis on the pressure sequence output by the data acquisition module 71, calculate the root mean square value of the pressure signal to characterize the turbulence intensity, and calculate the skewness coefficient of the pressure signal to characterize the flow pattern characteristics. The state identification module 74 is configured to compare the feature values ​​calculated by the feature extraction module 73 with a preset flow resistance fingerprint database to determine whether the current system is in an unloaded, stable conveying, critical settling, or suction port blocked state.

[0031] The instruction generation module 75 is configured to generate control signals based on the determination result of the state identification module 74 and send them to the vector inverter 62 and the bypass air supply valve 41. When in a critical settlement state, the instruction generation module 75 generates a signal containing a reference frequency. and high-frequency modulated waves The composite pulse command drives the centrifugal fan 61 to generate pulsating airflow; when the inlet is blocked, the command generation module 75 generates an opening signal to drive the bypass air supply valve 41 to open, so as to introduce compensating airflow to maintain the delivery speed in the pipeline.

[0032] In the embodiment employing a dual-circular straight pipe to single-circular straight pipe suction nozzle structure, the system is also equipped with differential balancing logic. In this case, the multi-dimensional differential pressure sensing array 2 has independent pressure sensors installed in each of the two circular intake branch pipes 11. The central control unit 7 determines whether a unilateral blockage exists by monitoring the absolute value of the pressure difference within the two circular intake branch pipes 11. When a unilateral blockage is detected, the command generation module 75 is configured to send a step-change frequency command to the vector inverter 62, utilizing the inertial effect of unsteady flow to generate a reverse turbulent airflow at the Y-shaped confluence throat 12.

[0033] In this embodiment, the flu-type suction nozzle unit 1 is constructed as an asymmetric dual-channel manifold assembly, designed to generate observable pressure gradients and turbulence characteristics in the internal flow field through specific geometric boundary conditions. The circular intake branch 11 includes a left intake branch, a right intake branch, a Y-shaped manifold throat 12, and a single circular outlet straight pipe 13. The left and right intake branches are mirror-symmetrically distributed, and their central axes form a predetermined angle in space. The included angle The value range is configured between 30 degrees and 60 degrees to ensure that the airflow generates shear mixing with specific frequency characteristics at the confluence, rather than excessive energy loss caused by vertical collision.

[0034] The ends of both the left and right intake branch pipes are hermetically welded to or integrally formed with the inlet end of the Y-shaped manifold throat 12. The internal flow channel cross-section of the Y-shaped manifold throat 12 gradually narrows along the airflow direction, and the cross-sectional area of ​​its outlet end is smaller than the sum of the cross-sectional areas of the left and right intake branch pipes. This variable cross-section design makes the Y-shaped manifold throat 12 a throttling element in terms of fluid dynamics, forcing the gas-solid two-phase flow to generate an acceleration effect when passing through this region, thereby amplifying the dynamic pressure component and static pressure pulsation amplitude in the throat region. An inverted V-shaped guide ridge is provided at the junction of the inner walls of the Y-shaped manifold throat 12, and the direction of the ridge line is perpendicular to the plane defined by the left and right intake branch pipes.

[0035] The angle of attack of the guide prism is configured to be less than the included angle. This is used to pre-form the streamlines before the two branch airflows merge, preventing excessive backflow zones at the confluence point that could lead to energy dissipation. Simultaneously, an annular turbulence groove is provided on the inner wall of the outlet of the Y-shaped confluence throat 12. The depth of this groove is 1% to 3% of the pipe diameter. This groove is used to artificially induce the shedding of micro-vortices at fixed points, enhancing the signal-to-noise ratio of the turbulent pulsation signal collected by the second pressure sensor 22, ensuring that effective flow resistance fingerprint features can still be extracted under low flow velocity conditions. A single-circular outlet straight pipe 13 is connected to the rear end of the Y-shaped confluence throat 12, and its inner diameter is consistent with the inner diameter of the outlet end of the Y-shaped confluence throat 12. This is used for flow rectification and connection to the subsequent long-distance delivery pipeline 3.

[0036] Based on this structure, the arrangement of the multidimensional differential pressure sensing array 2 has a strict geometric correspondence. The first pressure sensor 21 includes a first left branch sensor and a first right branch sensor, which are respectively located 10mm to 50mm above the inlet edge of the left and right intake branches. Each sensor communicates with the flow field inside the pipe through a static pressure tapping hole opened on the pipe wall. The axis of the pressure tapping hole is perpendicular to the cross-section of the pipe wall to eliminate the interference of dynamic pressure on the measurement.

[0037] The second pressure sensor 22 is located at the geometric center of the Y-shaped manifold throat 12 or within one pipe diameter downstream of it. This location corresponds to the region in the flow field where the flow velocity is highest and the pressure fluctuation is most intense. Pressure is collected from the left branch separately. Right branch pressure and confluence throat pressure This structure can physically reflect the load balance of the two intake branches and the overall solid-gas mixture ratio.

[0038] In this embodiment, the flu-type suction nozzle unit 1 is configured as an axisymmetric Venturi-type vacuum component, mainly composed of an inverted cone-shaped air intake hood, a hyperbolic constriction section, a throat flow stabilizing section, and a diffuser connecting section connected in sequence. The inverted cone-shaped air intake hood has an air inlet cross-sectional area of... Larger than the cross-sectional area of ​​the throat steady flow section The area ratio of the two Configured to be greater than 1.5 to ensure a pressure drop is formed inside the nozzle.

[0039] The hyperbolic contraction section is located between the inverted conical inlet hood and the throat stabilization section. Its inner wall profile is a smoothly contracting hyperbola or arc shape along the flow direction, used to guide the airflow to accelerate smoothly and suppress boundary layer separation. The throat stabilization section is a straight pipe with a uniform cross-section, its length configured to be 0.5 to 1 times the throat diameter, used to establish a stable maximum velocity region. The diffuser section is located after the throat stabilization section, with a small-angle expansion shape, used to convert part of the kinetic energy of the high-speed airflow into static pressure energy, thereby reducing the total flow resistance of the system.

[0040] In this inverted conical single-tube structure, the first pressure sensor 21 is located inside the skirt of the inverted conical air intake hood to monitor the vacuum level at the inlet. This value directly reflects the contact gap between the nozzle and the ground. The second pressure sensor 22 is located in the middle of the throat stabilization section, which is the point of lowest static pressure in the entire flow field. When the airflow containing coal particles flows through this structure, the mass effect of the coal particles changes the pressure loss coefficient of the fluid as it passes through the hyperbolic contraction section. At this time, the pressure difference between the first pressure sensor 21 and the second pressure sensor 22... It is no longer solely related to gas flow rate, but exhibits a non-linear mapping relationship with the solid concentration (solid-to-gas ratio) in the gas flow, thus endowing the mechanical structure with the physical function of measuring material concentration. All pressure taps are equipped with backflushing devices or anti-clogging filters to prevent coal dust from entering the sensing chamber of the pressure transmitter.

[0041] Specifically, a pressure tapping pipe connects the pressure tapping point to the pressure transmitter, and a three-way valve is installed on the pipe. The third end of the three-way valve is connected to a high-pressure pulse air source. The system is configured to trigger the high-pressure pulse air source to perform instantaneous reverse purging of the pressure tapping hole during non-sampling intervals or when signal hysteresis is detected. The purging pressure is set to 1.2 to 1.5 times the absolute value of the system's operating negative pressure to physically remove the coal powder filter cake accumulated at the edge of the pressure tapping hole. In addition, the pressure transmitter's sensing diaphragm uses a flush diaphragm structure, and the diaphragm surface is coated with a diamond-like carbon wear-resistant coating. During installation, the diaphragm surface is kept flush with the inner wall of the suction nozzle to avoid forming dead corners for dust accumulation and reduce secondary disturbances to the flow field.

[0042] See attached document Figure 2 The flow pattern identification method of this invention is not based on traditional threshold comparison, but rather on in-depth analysis of the time-frequency domain characteristics of high-frequency pressure signals to reconstruct the microscopic motion state of the gas-solid two-phase flow inside the pipeline. This process is periodically executed by the algorithm module inside the central control unit 7, and specifically includes three core stages: signal acquisition and preprocessing, flow resistance fingerprint feature calculation, and flow pattern state mapping.

[0043] During the signal acquisition and preprocessing stage, the central control unit 7 synchronously reads the analog voltage signal output by the multi-dimensional differential pressure sensing array 2 at a set high-frequency sampling rate (preferably 1kHz to 2kHz, and not less than 500Hz). To eliminate power frequency interference and electronic noise, the raw signal needs to undergo digital bandpass filtering, with the passband range set to 0.5Hz to 200Hz. This frequency band covers the slug flow frequency and turbulent pulsation frequency commonly found in gas-solid two-phase flows. Subsequently, the system performs DC component removal processing on the filtered pressure sequence to extract the pure dynamic pressure pulsation signal. ,in This represents the average pressure within the current sliding window. Before feature extraction, to suppress the spectral leakage effect caused by sliding window truncation, the system is configured to weight the data sequence within each sampling window using either a Hanning or Hamming window. The weighted signal sequence reduces sidelobe interference and improves subsequent computation. and The eigenvalues ​​provide the analytical resolution for subtle changes in flow patterns (such as low-frequency, subtle fluctuations caused by early settlement).

[0044] In the flow resistance fingerprint feature calculation stage, the feature extraction module 73 performs statistical analysis on the preprocessed pressure pulsation signal to calculate key fingerprint indicators characterizing the flow field properties. First, it calculates the turbulent pulsation intensity characteristic value of the pressure signal. This value physically corresponds to the energy level caused by vortex shedding and particle collisions in the flow field. The specific calculation uses the root mean square algorithm, targeting a length of... Discrete sampling window: ; in: For the first in the window One pressure sampling point. Next, the skewness coefficient of the pressure signal is calculated. The skewness coefficient (SCC) is used to describe the asymmetry of the probability density distribution of pressure fluctuations. In uniform suspended flow, pressure fluctuations approximately follow a normal distribution with a skewness coefficient close to zero; however, when flow instability or sedimentary layers occur, pressure fluctuations exhibit a significant unidirectional deviation. The SCC is calculated based on the following: ; Furthermore, for specific application scenarios, the feature extraction module 73 is also configured to perform a fast Fourier transform to calculate the energy ratio of the low-frequency band (1Hz-10Hz) to the high-frequency band (50Hz-100Hz). , as an auxiliary criterion.

[0045] During the manifold state mapping stage, the state identification module 74 uses the feature vector calculated above. The status is determined by combining a pre-set flow resistance fingerprint database. The determination logic specifically includes: when a flow resistance fingerprint is detected... The pressure average value is in the low amplitude range (less than the preset first threshold). When the temperature is low, the system is determined to be in an idle operating state; when In the mid-to-high amplitude range and When the absolute value is less than the preset skewness threshold, it indicates that the airflow contains material and is evenly mixed, and is judged to be in a stable suspension conveying state.

[0046] When detected Energy accumulation occurs in the low-frequency range, leading to an increase in values, and A significant positive offset indicates the formation of a moving sediment layer (dust flow) at the bottom of the pipeline. Airflow over the dune surface generates periodic, low-frequency, large fluctuations, marking a critical settling warning state. This state is a key point for system intervention, indicating that the current conveying velocity is approaching or below the material's critical suspension velocity. When an average negative pressure is detected... The vacuum level rose sharply to near the system's ultimate vacuum, but However, when the value abnormally drops to near zero, it indicates that the nozzle inlet is completely blocked, and the stagnant internal airflow causes the turbulent pulsation to disappear. At this point, it is determined to be a nozzle blockage state. Through the joint analysis of the above multi-dimensional characteristics, the system can accurately distinguish between a simple high negative pressure working state and a faulty blockage state, avoiding the misjudgment of traditional single negative pressure control logic.

[0047] See attached document Figure 3 This embodiment details how the central control unit 7 drives the negative pressure drive unit 6 and the pneumatic compensation unit 4 to perform unsteady airflow regulation based on the flow pattern identification results, in order to solve the settling and blockage problems in long-distance transportation. The control strategy mainly includes four execution subroutines: critical transportation reference setting, harmonic injection pulse generation, bypass air replenishment self-healing, and differential backwashing.

[0048] In the critical transport baseline setting stage, the model calculation module 72 first establishes a suspension velocity model based on preset material physical property parameters. To ensure that the material does not undergo gravity settling in the long-distance transport pipeline 3, the airflow must provide sufficient vertical component drag. The system calculates the critical suspension velocity based on the particle force balance equation. Based on calculations Based on the friction coefficient of the pipeline system and the performance curve of the fan, the controller determines the inverter reference frequency required to maintain the flow rate by looking up tables or interpolation. This reference frequency This represents the minimum energy supply level required to overcome pipeline resistance and maintain material suspension under steady-state flow conditions.

[0049] During the harmonic injection pulse generation stage, the command generation module 75 is configured to generate a composite control signal containing both DC and AC components to drive the centrifugal fan 61 to produce pulsating airflow. When the state identification module 74 determines that the system is in a critical settlement warning state, the command generation module 75 no longer outputs a constant signal. Instead, it constructs frequency instructions that change periodically over time. The mathematical expression of this instruction is configured as follows: ; in: represent The frequency command value is sent to the vector inverter 62 at all times; The reference frequency is the one calculated above; This represents the amplitude of the pulse modulation, and its value range is limited to... The speed should be between 5% and 15% to avoid excessive speed fluctuations that could damage the motor bearings. This represents the injected perturbation frequency. In this embodiment, The set value is not randomly selected, but is obtained through the acoustic modal sweep frequency calibration program during the system initialization phase. It is configured to match the low-order acoustic natural frequency of the pipeline system between the influenza-type nozzle unit 1 and the gas-solid separation unit 5, which is usually located in the low-frequency range of 0.5Hz to 5Hz.

[0050] Through the aforementioned harmonic injection, the flow rate and pressure output by the centrifugal fan 61 exhibit periodic fluctuations, forming an unsteady pulsating flow within the long-distance conveying pipeline 3. The physical effect of this pulsating flow is that it causes a gradient change in the airflow velocity along the time axis, thereby inducing additional unsteady shear stress in the boundary layer near the inner wall of the pipeline. This shear stress can disrupt the mechanical equilibrium of the coal particle layer deposited at the bottom of the pipeline, causing it to overcome static friction and re-enter the mainstream region. Furthermore, at the inlet of the flu-type suction nozzle unit 1, the pressure wave generated by the pulsating airflow creates a suction effect similar to breathing, which can disrupt the internal arch structure of the caking coal seam, assisting the suction nozzle in breaking the arch and adsorbing, thereby improving the system's conveying reliability without increasing average energy consumption.

[0051] During the bypass air replenishment self-healing phase, the system is configured to address the risk of material accumulation in the pipeline caused by high vacuum and low flow. When the status identification module 74 determines that the system is in a blocked suction port state (i.e., the suction nozzle is pressed against the ground or blocked by thick material), the instruction generation module 75 immediately sends an opening signal to drive the bypass air replenishment valve 41 to operate. The opening degree of the bypass air replenishment valve 41 is PID-regulated according to the target flow rate in the long-distance conveying pipeline 3 to ensure that even when the flow rate at the suction nozzle inlet approaches zero, the compensating airflow introduced from the bypass can maintain the actual air velocity in the pipeline at no less than the critical suspension velocity. This process ensures that materials already in the long-distance conveying pipeline 3 can continue to be conveyed to the gas-solid separation unit 5, preventing blockages or severe material accumulation accidents caused by a sudden drop in air velocity throughout the pipeline due to suction port blockage. When the feature extraction module 73 detects the pressure at the suction nozzle throat... After the pulsation characteristics are restored (indicating that the suction port is unobstructed), the bypass air supply valve 41 gradually closes, and the system returns to the normal suction and sweeping mode.

[0052] For embodiments employing a dual-circular straight tube to single-circular straight tube suction nozzle structure, the system is also equipped with differential backwash control logic. When the central control unit 7 monitors the pressure in the left branch... Pressure on the right branch When the absolute value of the difference continuously exceeds the preset imbalance threshold, it is determined that the suction nozzle is unilaterally unbalanced or blocked. At this time, the instruction generation module 75 controls the vector frequency converter 62 to execute an emergency stop and rapid rise sequence. Specifically, the controller first sends an instruction to cause the fan speed to drop sharply in a short time (e.g., within 0.5 seconds), using the airflow inertia to cause the negative pressure in the pipeline to drop rapidly; then immediately sends a full-speed rise instruction. During this process, due to the difference in fluid impedance between the unblocked side branch and the blocked side branch, the propagation speed and phase of the pressure wave in the two branches differ, thereby generating a transient lateral pressure gradient, pressure shock wave, or local reverse flow at the Y-shaped confluence throat 12. This unsteady fluid dynamic effect impacts and loosens the accumulated material on the blocked side, realizing a pneumatic self-cleaning function without mechanical moving parts.

[0053] See attached document Figure 4 This embodiment details the complete operating logic of the pulse-type train negative pressure cleaning system based on flow resistance spectrum fingerprint feedback in actual operation. This process is automatically completed by the control program executed inside the central control unit 7, and is mainly divided into the system initialization and modeling stage, the real-time flow condition monitoring and identification stage, and the multi-modal adaptive control stage.

[0054] During the system initialization and modeling phase, after the system is powered on, a self-test program is first executed to check the communication connection status and electrical response characteristics of each pressure sensor in the multi-dimensional differential pressure sensing array 2, the bypass air supply valve 41 of the pneumatic compensation unit 4, and the vector frequency converter 62 of the negative pressure drive unit 6. After the self-test passes, the model calculation module 72 reads the physical parameters of the working object pre-stored in the register, including coal particle density, average particle size, sphericity coefficient, and atmospheric density of the on-site environment. Based on the particle force balance equation constructed in the aforementioned embodiment, the model calculation module 72 calculates the critical suspension velocity under the current working condition. By combining the fluid resistance characteristic curve of the long-distance conveying pipeline 3, the lower limit of the inverter reference frequency that can maintain this minimum flow rate is derived in reverse. .

[0055] Subsequently, the system enters the acoustic modal sweep calibration subroutine. The instruction generation module 75 controls the vector inverter 62 in... A low-amplitude disturbance signal with linearly increasing frequency is superimposed on the signal, covering a scanning range of 0.5Hz to 10Hz. Simultaneously, the feature extraction module 73 monitors the pressure fluctuation amplitude response of the pressure sensor (i.e., the third pressure sensor 23) at the end of the long-distance transport pipeline 3. The system identifies the disturbance frequency point corresponding to the maximum value of the pressure fluctuation amplitude through spectrum analysis, and locks this frequency point as the acoustic disturbance frequency of the current pipeline system. The data is stored in a register as a fixed frequency parameter for the modulation wave in subsequent pulsed pneumatic delivery modes, ensuring that the pulsed airflow forms a standing wave resonance effect within the pipe. Simultaneously, the central control unit 7 collects static zero-point data from the first pressure sensor 21, the second pressure sensor 22, and the third pressure sensor 23 for calibration to eliminate the influence of ambient atmospheric pressure fluctuations on the relative pressure difference calculation.

[0056] Upon entering the real-time flow condition monitoring and identification phase, the system activates the negative pressure drive unit 6, which operates at a preset idle frequency. The data acquisition module 71 continuously acquires pressure data streams from each measuring point at a sampling frequency of no less than 500Hz. The feature extraction module 73 performs real-time calculations on the pressure sequence within the sliding time window, outputting feature values ​​including turbulence fluctuation intensity. skewness coefficient of pressure signal and average pressure Multidimensional feature vectors, including those from the eigenvectors.

[0057] The state identification module 74 inputs the feature vector into a preset flow resistance fingerprint database for matching and retrieval. If detected... If the value is lower than the preset aerodynamic noise threshold and the average negative pressure is low, the system determines it to be in an unloaded standby state. At this time, the command generation module 75 controls the vector inverter 62 to maintain the minimum energy consumption speed, and the bypass air supply valve 41 remains closed. If detected The value is in the medium to high amplitude range and When the solid-to-gas ratio approaches zero, the system determines that it is in a stable delivery state. At this point, the controller executes conventional PID control logic, fine-tuning the fan speed based on the real-time estimated solid-to-gas ratio to maintain the delivery velocity in the pipeline slightly higher than zero. .

[0058] During the multimodal adaptive control phase, the system executes targeted intervention strategies based on the special operating condition flags output by the state identification module 74. When the state identification module 74 captures... The energy in the low-frequency band is abnormally increased and When a positive offset occurs, the system determines that it has entered a critical settlement warning state. At this time, the command generation module 75 immediately switches to pulse transmission mode and generates a frequency command value superimposed with a sinusoidal modulation wave. The signal is sent to the vector inverter 62. The centrifugal fan 61 then outputs a pulsating airflow with a frequency matching the acoustic mode of the duct, using the shear force of the unsteady flow to disrupt the boundary of the particle deposition layer at the bottom of the duct, until the feature extraction module 73 provides feedback. A return to zero indicates that the sediment layer has been lifted and transported, and the system then slowly reduces the pulsation amplitude or switches back to steady-state control.

[0059] When the status identification module 74 detects the average pressure It rises sharply to the high vacuum range, and at the same time When the value drops sharply to near zero, the system determines that it has entered a suction port blockage state. The instruction generation module 75 immediately triggers the self-healing protection logic, sending a fully open or proportional adjustment instruction to the pneumatic compensation unit 4, driving the bypass air supply valve 41 to open rapidly. The external compensation airflow introduced by the bypass air supply valve 41 causes the flow velocity in the long-distance delivery pipeline 3 to quickly recover to normal. The above ensures that the material trapped in the pipeline continues to migrate to the gas-solid separation unit 5. During this period, the central control unit 7 continuously monitors the pressure fluctuation characteristics of the second pressure sensor 22 (the throat of the suction nozzle). Once the turbulent pulsation signal is detected to be restored, it indicates that the accumulated material at the suction port has been cleared. The system then gradually closes the bypass air supply valve 41 at a preset slope and simultaneously adjusts the frequency converter frequency to smoothly transition back to the normal operating mode.

[0060] Furthermore, for systems equipped with a dual-circular straight-tube to single-circular straight-tube suction nozzle, the central control unit 7 executes a differential balance monitoring subroutine in parallel. When the calculated absolute value of the pressure difference between the left and right branches continuously exceeds the set imbalance threshold (e.g., for a duration exceeding 2 seconds), the system interrupts the current control loop and inserts a backwashing action. The instruction generation module 75 sends a frequency drop instruction lasting several hundred milliseconds, followed immediately by a full-torque acceleration instruction. This sequence of actions induces a violent reverse vortex and pressure shock wave at the Y-shaped confluence point of the suction nozzle, using fluid dynamics to impact and loosen the material on the blocked side. After the backwashing action is completed, the system automatically resets to the real-time flow monitoring stage and continues to execute the main control loop until a shutdown instruction is received. The entire operation process, through the above closed-loop logic, achieves a dynamic optimal balance between energy consumption, conveying efficiency, and system reliability under complex and variable cleaning conditions.

Claims

1. A variable frequency controlled train negative pressure cleaning system based on differential pressure sensing, characterized in that, It includes a central control unit, as well as a negative pressure drive unit, a pneumatic compensation unit, and a multi-dimensional differential pressure sensing array that are electrically connected to the central control unit; The negative pressure drive unit, gas-solid separation unit, long-distance delivery pipeline and flu-type suction nozzle unit are connected in sequence along the airflow direction to form a negative pressure suction air path. The pneumatic compensation unit is disposed between the flu-type suction nozzle unit and the long-distance delivery pipeline, and the multi-dimensional differential pressure sensing array is disposed at the key node of the negative pressure suction air path; The central control unit is configured to execute the following control logic: The pressure signal output by the multidimensional differential pressure sensing array is acquired and preprocessed. Time-domain statistical analysis was performed on the preprocessed pressure signal to calculate the flow resistance fingerprint characteristic value that characterizes the flow field. The flow resistance fingerprint feature value is compared with a preset flow resistance fingerprint database to identify the current flow pattern state; Based on the identified flow pattern, control commands are generated to adjust the operating frequency of the negative pressure drive unit or the valve opening of the pneumatic compensation unit to maintain stable suspension and transport of materials in the pipeline or to relieve sedimentation and blockage in the pipeline.

2. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 1, characterized in that, The flu-type suction nozzle unit adopts a double-circle straight tube to single-circle straight tube structure; The dual-circle straight pipe to single-circle straight pipe structure includes two parallel circular air intake branches, a Y-shaped manifold throat, and a single-circle air outlet straight pipe. The two circular intake manifolds are the left intake manifold and the right intake manifold, respectively. The left intake manifold and the right intake manifold are mirror-symmetrically distributed, and the central axis of the left intake manifold and the central axis of the right intake manifold form a preset angle in space. The rear ends of the left intake branch pipe and the right intake branch pipe are connected to the inlet end of the Y-shaped manifold, and the rear end of the Y-shaped manifold is connected to the single circular outlet straight pipe. The sum of the cross-sectional areas of the left intake branch pipe and the right intake branch pipe is greater than the cross-sectional area of ​​the single circular outlet straight pipe, so that the Y-shaped confluence throat constitutes a throttling area. The inner wall of the Y-shaped manifold throat is provided with an inverted V-shaped guide ridge, and the ridge line of the inverted V-shaped guide ridge is perpendicular to the plane defined by the left intake branch pipe and the right intake branch pipe.

3. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 2, characterized in that, The multidimensional differential pressure sensing array includes a first pressure sensor, a second pressure sensor, and a third pressure sensor; The first pressure sensor is arranged at the inlet edge of the circular intake branch pipe and is configured to collect the ambient pressure at the inlet. The second pressure sensor is arranged at the geometric center point of the Y-shaped manifold throat or at a position within a preset range downstream of the geometric center point, and is configured to collect throat pressure; The third pressure sensor is arranged in the initial straight section of the long-distance conveying pipeline and is configured to collect stable conveying pressure; The inner wall of the outlet of the Y-shaped confluence throat is provided with an annular turbulence groove, which is configured to induce the shedding of micro-vortices to enhance the pressure pulsation signal collected by the second pressure sensor.

4. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 1, characterized in that, The central control unit is internally equipped with a model calculation module, a feature extraction module, a state recognition module, and an instruction generation module; The model calculation module is configured to calculate the critical suspension velocity based on the material physical property parameters and the particle force balance equation, and to determine the reference operating frequency of the negative pressure drive unit. The feature extraction module is configured to process the pressure signal to remove the DC component to obtain a dynamic pressure pulsation signal, and to calculate the statistical characteristics of the dynamic pressure pulsation signal. The state identification module is configured to determine whether the system is in an unloaded operation state, a stable suspended conveying state, a critical settling warning state, or a suction port blockage state based on the statistical characteristics.

5. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 4, characterized in that, The flow resistance fingerprint feature values ​​calculated by the feature extraction module include turbulence intensity feature values ​​and skewness coefficients; The characteristic value of the turbulence fluctuation intensity is obtained by calculating the arithmetic square root of the average of the sum of squares of the differences between the pressure sampling points and the average pressure within the sampling window. The skewness coefficient is obtained by calculating the ratio of the average of the sum of the cubes of the differences between the pressure sampling points and the average pressure within the sampling window to the cube of the turbulence fluctuation intensity characteristic value. The feature extraction module is also configured to perform windowing and weighting processing on the data sequence within the sampling window before calculation to suppress spectral leakage.

6. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 5, characterized in that, The determination logic of the state identification module specifically includes: When the turbulence pulsation intensity characteristic value is less than the first threshold and the average negative pressure is lower than the set value, it is determined to be the no-load operation state; When the turbulence fluctuation intensity characteristic value is detected to be in the medium-high amplitude range and the absolute value of the skewness coefficient is less than the skewness threshold, it is determined to be the stable suspension transport state. When the intensity characteristic value of the turbulence fluctuation is detected to increase in the low-frequency range and the skewness coefficient shows a positive shift, it is determined to be the critical settlement warning state. When the average negative pressure is detected to rise to the high vacuum range and the characteristic value of the turbulence pulsation intensity drops to near zero, it is determined that the suction port is in a blocked state.

7. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 6, characterized in that, The instruction generation module is configured to generate a composite frequency instruction containing both DC and AC components when the state identification module determines that the system is in the critical settlement early warning state. The composite frequency command is configured to superimpose a sinusoidal modulation wave that varies periodically with time onto the reference operating frequency; The modulation amplitude of the sinusoidal modulation wave is limited to a preset percentage range of the reference operating frequency; The frequency of the sinusoidal modulation wave is configured to match the acoustic natural frequency of the piping system between the flu-type nozzle unit and the gas-solid separation unit.

8. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 7, characterized in that, The acoustic natural frequency is obtained through the acoustic mode sweep calibration procedure during the system initialization phase; The acoustic mode sweep calibration procedure includes: The vector inverter in the negative pressure drive unit is controlled to superimpose a disturbance signal with linearly increasing frequency on the reference operating frequency; Simultaneously monitor the pressure fluctuation amplitude response on long-distance pipelines; Identify the disturbance frequency point corresponding to the maximum value of the pressure fluctuation amplitude, and lock the frequency point as the acoustic natural frequency.

9. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 6, characterized in that, The instruction generation module is configured to generate an opening signal to drive the bypass air supply valve of the pneumatic compensation unit to open when the state identification module determines that the system is in the state of the suction port blockage. The bypass air supply valve adjusts its opening degree according to the target flow rate in the long-distance delivery pipeline, and introduces auxiliary airflow to keep the flow velocity in the pipeline above the critical suspension velocity. When the feature extraction module detects the recovery of the turbulent pulsation signal of the pressure sensor located at the throat of the influenza-type suction unit, the instruction generation module controls the bypass air supply valve to gradually close.

10. The variable frequency control train negative pressure cleaning system based on differential pressure sensing according to claim 2, characterized in that, The multidimensional differential pressure sensing array is equipped with independent pressure sensors in the left intake branch pipe and the right intake branch pipe respectively; The central control unit is configured to monitor the absolute value of the pressure difference between the left intake manifold and the right intake manifold. When the absolute value of the pressure difference continues to exceed the imbalance threshold, the central control unit controls the negative pressure drive unit to execute an emergency stop and rapid rise sequence. The emergency stop and rapid rise sequence is configured to cause the fan speed to drop significantly within a preset time and then immediately increase to full speed, inducing reverse vortices and pressure shock waves at the Y-shaped confluence throat.

Images