A cable shaft fireproof plugging sealing performance detector

Abstract

This invention discloses a cable shaft fireproof sealing performance tester, comprising an integrated air intake structure, air delivery structure, and detection structure. The air intake structure connects to the pipe under test via a connecting sleeve, and a hydraulic cylinder drives a push rod to open a valve; the air delivery structure controls the gas flow rate via a regulating valve; the detection structure analyzes the gas concentration using a carbon dioxide concentration detector. A characteristic judgment device outputs a characteristic value H through laser scanning and pressure sensing, dynamically optimizing detection parameters; a prediction model generates concentration prediction values ​​based on historical data, replacing abnormal measured values ​​to maintain baseline stability; an error ratio evaluation system monitors the prediction error frequency and promptly triggers a sealing failure alarm; the user coordinate system setting supports custom detection benchmark points, simplifying the multi-point detection process. This integrated design solves the problems of complex assembly and low operating efficiency of traditional testing tools, significantly improving the accuracy and adaptability of sealing performance testing.

Description

Technical Field

[0001] This invention relates to the field of cable shaft construction technology, specifically to a cable shaft fireproof sealing performance tester. Background Technology

[0002] Cable shafts are channels in high-rise buildings used for vertically laying cables. Their dense wiring and confined spaces make them highly susceptible to fires caused by short circuits or overloads. Because the chimney effect of these shafts accelerates the spread of fire, the fireproof sealing and sealing performance testing is crucial.

[0003] Previously, when testing for air tightness, there was no convenient testing mechanism that could integrate air intake, delivery, and testing into one unit. Multiple tools were required for testing, and these tools needed to be installed and connected before testing. The operation was complicated and the testing efficiency was low. Summary of the Invention

[0004] To address the problems in the existing technology, the present invention provides a cable shaft fireproof sealing performance tester. This tester is pre-configured as a whole, which can conveniently integrate air intake, delivery and testing into one unit. There is no need to install and connect multiple tools before testing. It is easy to operate and has high testing efficiency during testing.

[0005] The technical solution adopted by this invention to solve its technical problem is a cable shaft fireproof sealing performance tester, including an air intake structure, an air supply structure, and a detection structure. The detection structure is provided at one end of the air intake structure, and the air intake structure and the detection structure are connected through the air supply structure. The device also includes:

[0006] A characteristic determination device, integrated into the inner wall of the docking sleeve on the air intake structure, is used to determine the characteristics of the pipeline to be tested in the cable shaft and output the characteristic value H.

[0007] An upstream detection device is installed at the inlet of the air intake pipe A on the air intake structure. It is used to detect the actual parameters of the gas input point, including the gas flow rate and initial concentration. The actual parameters are used as the upstream actual value Su.

[0008] A control device, embedded in the top of the detection chamber of the detection structure, adjusts the opening of the regulating valve on the gas supply structure and the sensitivity of the carbon dioxide concentration detector on the detection structure based on the characteristic value H and the actual value Su on the upstream side.

[0009] The prediction value determination unit is located inside the detection chamber. The prediction value determination unit predicts the current gas concentration based on the learned model.

[0010] An error evaluation unit, located within the control device, is used to calculate the proportion of times within a set time period when the difference between the predicted value and the measured concentration is less than a specified threshold.

[0011] The storage unit, integrated within the extension plate on the detection structure, stores learning information and detection data, including characteristic value H, historical concentration values, and prediction error.

[0012] The calibration unit, connected to the control device, calibrates the connection position of the intake pipe A based on the initial offset between the actual value Su on the upstream side and the target value; the calibration unit fine-tunes the position of the push rod through a hydraulic cylinder to ensure that the gas input point is aligned.

[0013] The user coordinate system setting unit is located in the housing of the detector and provides a user interface for setting the detection coordinate system, including the axial and radial reference points of the pipe.

[0014] The replacement mechanism is executed by the prediction value determination unit. When the prediction error is greater than or equal to a specified threshold, the measured concentration is replaced with the predicted value to maintain the normal detection benchmark.

[0015] Furthermore, the intake structure includes:

[0016] The docking sleeve has a lower convex plate on the outer side of its bottom end for connecting with the pipe to be tested, and an upper convex plate connected to the outer side of its top end. The bottom of the lower convex plate is provided with a first bolt through a mounting hole.

[0017] The upper connecting cylinder is installed on the upper end of the docking sleeve by bolts and an upper convex plate;

[0018] A rotating sleeve is fixed to the outer side of the top end of the upper connecting cylinder by a threaded rotation.

[0019] The fixed plate is mounted to the top of the rotating sleeve by a second bolt;

[0020] A hydraulic cylinder is connected to the bottom of the fixed plate and located inside the upper connecting cylinder;

[0021] A large-diameter sleeve is located at the center of the bottom end of the upper connecting sleeve;

[0022] Inlet pipe A is connected to the bottom of the large-diameter sleeve;

[0023] A connecting nut is threaded onto the outside of the intake pipe A;

[0024] The push rod is located inside the air intake pipe A, and its top end is connected to the power output end of the hydraulic cylinder through a connector and bolts;

[0025] The detection structure includes:

[0026] An extension plate is fixed to one end of the upper connecting cylinder;

[0027] The detection chamber is installed at the bottom of the extension plate;

[0028] The detection tube is connected to one side of the bottom of the detection chamber;

[0029] A carbon dioxide concentration detector is mounted on the detection tube, with its detection end located inside the detection tube;

[0030] The air delivery structure includes:

[0031] Air inlet pipe B is used to connect the upper connecting cylinder and the detection chamber;

[0032] A regulating valve, installed on the air inlet pipe B, is used to regulate the flow rate of gas entering the detection chamber;

[0033] The shut-off valve A is located on the air inlet pipe B near the regulating valve and the detection chamber.

[0034] Furthermore, the characteristic determination device includes:

[0035] A laser scanning module is installed on the inner wall of the docking sleeve, which emits a laser beam to scan the inner wall of the docking sleeve and obtain the inner diameter and surface roughness data.

[0036] The pressure sensing module is integrated at the front end of valve A in the intake pipe. It measures the gas input pressure at the front end of valve A in the intake pipe and calculates the material friction coefficient based on the inner diameter.

[0037] The data processing unit, embedded on the main control board of the characteristic determination device, converts the inner diameter, friction coefficient, and sealing layer thickness into characteristic values ​​H.

[0038] The characteristic value H is output to the control device to dynamically set the flow gain coefficient of the regulating valve.

[0039] Furthermore, the upstream detection device includes:

[0040] A camera and a flow sensor are mounted side-by-side inside the inlet flange of intake pipe A.

[0041] The camera is used to capture images of the gas input point and identify flow field distribution and foreign object blockage;

[0042] Flow velocity sensor, based on ultrasonic measurement of real-time flow velocity;

[0043] The data fusion unit integrates image data and flow velocity data into the upstream actual value Su.

[0044] Furthermore, the control device includes:

[0045] The microprocessor runs a PID control algorithm and receives the characteristic value H and the actual value Su from the upstream side.

[0046] The flow control module calculates the target opening degree V1 of the regulating valve based on the characteristic value H;

[0047] The sensitivity adjustment module sets the sampling frequency and threshold of the carbon dioxide concentration detector.

[0048] The feedback loop compares the measured concentration with the predicted value and dynamically fine-tunes the target opening V1.

[0049] Furthermore, the predicted value determination unit includes:

[0050] The learned model is complete, and the training data in the learned model consists of historical concentration sequences and flow command sequences under normal sealed conditions; and the learned model includes:

[0051] The input layer receives n historical concentration values ​​C1-Cn and n flow command values ​​Q1-Qn;

[0052] The output layer generates the current concentration prediction value P_C;

[0053] The replacement unit replaces the measured value with P_C and updates the data in the storage unit when |P_C - measured concentration| ≥ a specified threshold Th; the specified threshold Th is set based on the model mean absolute error and the detection repeatability accuracy.

[0054] Furthermore, the error evaluation unit includes:

[0055] A counter that records the number of times, N, the prediction error is less than Th within a set time T.

[0056] The ratio calculator calculates the ratio R = N / M, where M is the total number of tests.

[0057] The comparator outputs an error signal when R < the preset ratio R_set;

[0058] The preset ratio R_set is inversely proportional to the specified threshold Th. When Th increases, R_set decreases. This is automatically managed by the setting change unit. An abnormal signal triggers an audible and visual alarm.

[0059] Furthermore, the storage unit includes:

[0060] The prediction variable memory stores n sets of concentration data and flow instructions that are updated on a rolling basis;

[0061] Storage memory for recording predicted values, measured values, and error flags;

[0062] The learning information repository stores the normal state training dataset and model parameters;

[0063] The data compression module reduces the dimensionality of historical data for storage.

[0064] The retrieval interface allows the correction processing department to access the data optimization model.

[0065] Furthermore, the correction unit includes:

[0066] The offset calculation unit is based on the difference ΔS between the actual value Su on the upstream side and the target value S_target.

[0067] The coordinate system transformation unit converts ΔS into a hydraulic cylinder displacement command;

[0068] The actuator drives the hydraulic cylinder to fine-tune the position of the push rod, compensating for installation misalignment;

[0069] Feedback sensors monitor the corrected position in real time to ensure that ΔS < tolerance.

[0070] Furthermore, the user coordinate system setting unit includes:

[0071] The touchscreen interface displays a 3D model of the pipeline.

[0072] The coordinate input panel allows users to set the detection origin O, the X-axis, and the Y-axis;

[0073] The data output interface transmits the user coordinate system (UF) to the calibration unit and the prediction value determination unit; the calibration module automatically optimizes the UF parameters based on the measured data from the first shooting position P1.

[0074] The beneficial effects of this invention are:

[0075] Traditional testing requires assembling the air source device, delivery pipeline, and testing instrument in stages, which is time-consuming and prone to errors due to poor sealing. This design uses a rotating sleeve to fix the hydraulic cylinder, with the large-diameter sleeve directly connected to the air inlet pipe A, and the extension plate integrating the testing chamber, achieving "installation equals testing".

[0076] This adaptive control, driven by the characteristic value H, uses a characteristic determination device to output the characteristic value H in real time, and the control device dynamically sets the flow gain coefficient (K_H) and the sensitivity of the detector accordingly. Traditional methods use a fixed flow rate / sensitivity, which is difficult to handle different pipeline conditions. This application can achieve automatic adjustment of the flow rate according to the H value (reflecting the pipeline complexity): when the pipeline inner diameter is small (large H value), the flow rate is increased to improve the detection rate of minute leaks; when the friction coefficient is high (large H value), the sampling frequency of the concentration detector is increased to capture transient changes.

[0077] After learning, the model generates a predicted value P_C based on historical sequences (concentration C1-Cn, flow rate Q1-Qn); when |P_C - measured value| ≥ Th, the measured value is replaced with P_C. Gas turbulence interference is common in cable shafts, causing instantaneous fluctuations (±15%) in concentration detectors. Traditional threshold methods are prone to misjudging seal failure. This design uses an LSTM time-series model to predict the concentration under normal conditions, and a replacement mechanism to shield against abnormal disturbances: the threshold Th is dynamically set (Th = 0.9 * (MAE + R95)), combined with the model's average error and the equipment's repeatability accuracy; after replacement, the data in the storage section is updated, achieving online model optimization.

[0078] The error ratio evaluation system ensures reliability: the error evaluation unit calculates the ratio R of the predicted error < Th, and triggers an alarm when R < R_set; R_set is inversely proportional to Th (Th↑→R_set↓). Traditional methods rely on single error exceeding the limit alarm, which easily misses progressive leaks (such as slow cracking of the sealing layer). This user coordinate system simplifies multi-position detection: users set the axial (X-axis) and radial (Y-axis) reference points of the pipeline through the touch screen, generating a user coordinate system (UF); the calibration unit performs position compensation based on the UF. Cable shafts often require multi-point detection, and traditional methods require repeated adjustments to the equipment orientation. This design allows users to preset the detection origin: the coordinate system transformation unit converts the offset ΔS into hydraulic cylinder displacement commands; the calibration module automatically optimizes the UF parameters based on the initial detection data (point P1). Attached Figure Description

[0079] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0080] Figure 1 This is a schematic diagram of the overall structure of the present invention;

[0081] Figure 2 This is a schematic diagram of the rear view structure of the present invention;

[0082] Figure 3 This is a schematic diagram of the rear view of the dismantled inner structure of the present invention;

[0083] Figure 4 This is a schematic diagram of the cross-sectional structure of the present invention;

[0084] Figure 5 This is a schematic diagram of the front view of the structure after sectional cutting of the present invention.

[0085] In the diagram: 1. Connecting sleeve; 2. Lower cam; 3. First bolt; 4. Upper cam; 5. Upper connecting sleeve; 6. Limiting cam; 7. Rotating sleeve; 8. Second bolt; 9. Fixed plate; 10. Exhaust pipe A; 11. Safety valve; 12. Hydraulic cylinder; 13. Inspection port A; 14. Sealing block; 15. Embedded pipe; 16. Extension plate; 17. Inlet pipe B; 18. Regulating valve; 19. Shut-off valve A; 20. Detection pipe; 21. Carbon dioxide concentration detector; 22. Exhaust pipe B; 23. Shut-off valve B; 24. Inspection port B; 25. Door body; 26. Inlet pipe A; 27. Connecting nut; 28. Connector; 29. ​​Push rod; 30. Large diameter sleeve; 31. Detection chamber. Detailed Implementation

[0086] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0087] Please see Figure 1 , Figure 1 This is a schematic diagram of the structure of the present invention.

[0088] A cable shaft fireproof sealing performance tester includes an air intake structure, an air supply structure, and a detection structure. The detection structure is located at one end of the air intake structure, and the air intake structure and the detection structure are connected via the air supply structure. The air intake structure includes a docking sleeve 1, a lower convex plate 2 for connecting to the pipe to be tested on its outer bottom end, and an upper convex plate 4 connected to its outer top end. A first bolt 3 is installed at the bottom of the lower convex plate 2 through a mounting hole. An upper connecting sleeve 5 is bolted and connected to the upper convex plate 4 and mounted on the upper end of the docking sleeve 1. A rotating sleeve 7 is threaded... The upper connecting cylinder 5 is fixed to the outer side of its top end; the fixed plate 9 is installed on the top of the rotating sleeve 7 by the second bolt 8; the hydraulic cylinder 12 is connected to the bottom of the fixed plate 9 and located inside the upper connecting cylinder 5; the large-diameter sleeve 30 is located at the center of the bottom end of the upper connecting cylinder 5; the air inlet pipe A26 is connected to the bottom of the large-diameter sleeve 30; the mating nut 27 is installed on the outer side of the air inlet pipe A26 by thread; the push rod 29 is located inside the air inlet pipe A26, and its top end is connected to the power output end of the hydraulic cylinder 12 by the connector 28 and bolt.

[0089] The detection structure includes an extension plate 16 fixed to one end of the upper connecting cylinder 5; a detection chamber 31 installed at the bottom of the extension plate 16; a detection tube 20 connected to one side of the bottom end of the detection chamber 31; and a carbon dioxide concentration detector 21 mounted on the detection tube 20, with its detection end located inside the detection tube 20.

[0090] The gas delivery structure includes an air inlet pipe B17 for connecting the upper connecting cylinder 5 and the detection chamber 31; a regulating valve 18, which is disposed on the air inlet pipe B17 for regulating the flow rate of gas entering the detection chamber 31; and a shut-off valve A19, which is disposed on the air inlet pipe B17 near the regulating valve 18 and the detection chamber 31.

[0091] In the above, an exhaust pipe B22 is connected to one side of the top of the test chamber 31. A shut-off valve B23 is installed on the exhaust pipe B22. The shut-off valve B23 facilitates the discharge of gas from the inspection port A13 inside the test chamber 31 through the exhaust pipe B22 after it is opened.

[0092] Two sets of pre-embedded pipes 15 are pre-installed on one side of the top of the hydraulic cylinder 12 for connecting to the oil inlet and outlet of the hydraulic cylinder 12. Through the pre-embedded pipes 15, the end of the pre-embedded pipe 15 near the inner wall of the rotating sleeve 7 can be connected to the oil inlet and outlet of the hydraulic cylinder 12. The end of the pre-embedded pipe 15 near the outer wall of the rotating sleeve 7 is connected to the external pipe after the hydraulic cylinder 12 is rotated and locked, thereby avoiding the influence of the rotation of the rotating sleeve 7 on the connecting pipe of the hydraulic cylinder 12.

[0093] The top end of the hydraulic cylinder 12 is provided with an inspection port A13. A sealing block 14 is installed on the inner wall of the inspection port A13 by screws. The sealing block 14 can be easily removed by loosening the screws, so that the connecting pipe of the hydraulic cylinder 12 can be inspected through the inspection port A13.

[0094] A limiting protrusion 6 is provided on the outer periphery of the upper connecting cylinder 5 to limit the downward rotation position of the hydraulic cylinder 12. The limiting protrusion 6 facilitates the limitation of the degree of freedom of the hydraulic cylinder 12 to rotate downward.

[0095] The top end of the hydraulic cylinder 12 is equipped with an exhaust pipe A10, and a safety valve 11 is installed on the exhaust pipe A10. When the air pressure in the upper connecting cylinder 5 is high, the exhaust pipe A10 is automatically opened to exhaust the air.

[0096] A maintenance port B24 is provided on one side of the docking sleeve 1. A door 25 is fixed to the outside of the maintenance port B24 by screws. The door 25 can be opened easily by screws, so that the docking nut 27 inside the docking sleeve 1 can be rotated through the maintenance port B24.

[0097] The detector also includes a characteristic determination device, integrated into the inner wall of the docking sleeve 1, used to determine the characteristics of the pipeline to be tested in the cable shaft, including the inner diameter of the pipeline, the coefficient of friction of the material, and the thickness of the fireproof sealing layer; the characteristic determination device acquires pipeline characteristic data through laser scanning and pressure sensors, and outputs characteristic value H, which continuously increases as the inner diameter of the pipeline decreases or the coefficient of friction of the material increases, and is used to indicate the difficulty of gas leak detection.

[0098] An upstream detection device is installed at the inlet of the air inlet pipe A26 to detect the actual parameters of the gas input point, including gas flow rate and initial concentration. The upstream detection device includes a camera and a flow rate sensor, and its detection data is used as the upstream actual value Su.

[0099] The control device, embedded in the top of the detection chamber 31, adjusts the opening of the regulating valve 18 and the sensitivity of the carbon dioxide concentration detector 21 based on the characteristic value H and the actual value Su on the upstream side. The control device adjusts the gas flow rate according to the magnitude of the characteristic value H. When the characteristic value H increases, the gas flow rate increases to improve the detection accuracy.

[0100] The prediction value determination unit is located inside the detection chamber 31 and is electrically connected to the carbon dioxide concentration detector 21. The prediction value determination unit predicts the current gas concentration based on the learned model and multiple historical gas concentration values ​​and target flow values. When the absolute value of the difference between the predicted value and the measured concentration is greater than or equal to a specified threshold, the predicted value replaces the measured value for the next prediction.

[0101] An error evaluation unit, located within the control device, is used to calculate the proportion of times within a set time period where the difference between the predicted value and the measured concentration is less than a specified threshold; when the proportion is less than a preset proportion, an abnormal notification is triggered.

[0102] The storage unit, integrated within the extension plate 16, stores learning information and detection data, including characteristic value H, historical concentration values, and prediction error.

[0103] The calibration unit 133, connected to the control device, calibrates the connection position of the intake pipe A26 based on the initial offset between the actual value Su on the upstream side and the target value; the calibration unit finely adjusts the position of the push rod 29 through the hydraulic cylinder 12 to ensure that the gas input point is aligned.

[0104] The user coordinate system setting unit is located in the housing of the detector and provides a user interface for setting the detection coordinate system, including the axial and radial reference points of the pipeline.

[0105] The replacement mechanism is executed by the prediction value determination unit. When the prediction error is greater than or equal to a specified threshold, the measured concentration is replaced with the predicted value to maintain the normal detection benchmark.

[0106] In the above, the characteristic determination device is fixed to the inner wall of the docking sleeve 1, adjacent to the lower convex plate 2; the upstream detection device is embedded in the inlet flange of the air inlet pipe A26; the control device and the predicted value determination unit are integrated on the circuit board on the top of the detection chamber 31; the error evaluation unit and the storage unit are located on the chip inside the control device; the calibration unit is connected to the hydraulic cylinder 12 via a cable; and the user coordinate system setting unit is set on the outer shell touch screen.

[0107] The characteristic determination device outputs characteristic value H to quantify the detection difficulty; the upstream detection device provides initial data Su as the detection benchmark; the control device dynamically adjusts parameters to ensure that the detection adapts to the pipeline characteristics; the prediction value determination unit predicts the concentration by learning the model, reducing abnormal interference; the error evaluation unit monitors the error ratio and triggers alarms in a timely manner; the storage unit records data for model optimization; the correction unit eliminates installation offset; the user coordinate system setting unit simplifies the setting of detection points; and the replacement mechanism ensures prediction stability.

[0108] The characteristic determination device first determines the pipeline characteristics and outputs a characteristic value H; the control device adjusts the initial parameters based on the characteristic value H; the upstream detection device collects input point data; the gas enters the detection chamber 31 through the gas delivery structure; the prediction value determination unit predicts the concentration based on historical data; the measured value is compared with the predicted value, and the measured value is replaced if the error is large; the error evaluation unit evaluates the error ratio and triggers an anomaly; the storage unit saves the data for learning; the calibration unit calibrates the position before detection; and the user coordinate system setting unit allows for custom references.

[0109] The characteristic determination device includes: a laser scanning module that emits a laser beam to scan the inner wall of the docking sleeve 1 to obtain the inner diameter and surface roughness data; a pressure sensing module that measures the gas input pressure and calculates the material friction coefficient based on the inner diameter; and a data processing unit that converts the inner diameter D, friction coefficient F, and sealing layer thickness T into a characteristic value H, using the formula H = K1*(1 / D) + K2*F + K3*T, where K1, K2, and K3 are weighting coefficients preset according to the pipe type. The characteristic value H is output to the control device for dynamically setting the flow gain coefficient of the regulating valve. As the characteristic value H increases, the gain coefficient increases, and the flow regulation range expands. The laser scanning module is installed on the inner wall of the docking sleeve 1, facing the pipe inlet; the pressure sensing module is integrated at the front end of the inlet pipe A26 valve; and the data processing unit is embedded in the main control board of the characteristic determination device.

[0110] In the above process, laser scanning precisely quantifies the geometric characteristics of the pipeline; pressure sensing derives the frictional resistance; and data processing outputs a characteristic value H, which directly affects the control parameters. Before detection, laser scanning and pressure sensing acquire data; the data processing unit calculates the characteristic value H; and the control device receives the characteristic value H and adjusts the subsequent detection sensitivity.

[0111] The upstream detection device includes: a camera to capture images of the gas input point and identify flow field distribution and foreign object blockage; a flow velocity sensor to measure real-time flow velocity based on ultrasound; and a data fusion unit to integrate image data and flow velocity data into an upstream actual value Su, including flow velocity Vu and initial concentration Cu. The upstream actual value Su is transmitted to a prediction value determination unit as a prediction input variable. When the deviation between the upstream actual value Su and a preset benchmark is >10%, a correction unit is triggered to recalibrate the position of the intake pipe A26. The camera and flow velocity sensor are installed side-by-side inside the inlet flange of the intake pipe A26; the data fusion unit is located inside the control box of the upstream detection device.

[0112] In the above, the camera provides visual verification to ensure no blockage; the flow rate sensor quantifies the input conditions; and the data fusion output Su serves as the prediction benchmark. Upon detection startup, the upstream detection device 30u acquires the actual upstream value Su; this data is used for prediction model initialization and error correction.

[0113] The control device includes: a microprocessor that runs a PID control algorithm and receives the characteristic value H and the actual value Su from the upstream side; a flow control module that calculates the target opening V1 of the regulating valve 18 based on the characteristic value H, using the formula V1 = V0 + K_H * H, where V0 is the reference opening and K_H is the gain coefficient of H; a sensitivity adjustment module that sets the sampling frequency and threshold of the carbon dioxide concentration detector 21, increasing the sampling frequency and decreasing the threshold as the characteristic value H increases; and a feedback loop that compares the measured concentration with the predicted value to dynamically determine the target opening V1.

[0114] In the above, the microprocessor and flow control module are embedded in the top PCB board of the detection chamber 31; the sensitivity adjustment module is integrated into the interface circuit of the carbon dioxide concentration detector 21. The microprocessor realizes adaptive control; the flow module ensures that the gas flow matches the pipeline characteristics; the sensitivity module optimizes the detection accuracy; and the feedback loop improves stability. After the characteristic value H is input, the control device calculates the target opening V1; the regulating valve 18 performs the opening adjustment; the sensitivity is set synchronously, and the feedback data during detection is used for real-time optimization.

[0115] The prediction value determination unit includes: a learned model, with training data consisting of historical concentration sequences and flow command sequences under normal sealed conditions; an input layer that receives n historical concentration values ​​C1-Cn and n flow command values ​​Q1-Qn; an output layer that generates the current concentration prediction value P_C; and a replacement unit that replaces the measured value with P_C and updates the storage unit data when |P_C - measured concentration| ≥ a specified threshold Th; the specified threshold Th is set based on the model mean absolute error MAE and the detection repeatability accuracy, Th = 0.9 * (MAE + R95), where R95 is the 95th percentile of the repeatability accuracy distribution.

[0116] In the above, after the model is learned, it runs in the FPGA chip of the prediction value determination unit 102; the input layer and the output layer are connected to the storage unit 106 via a bus; the permutation unit is a software module. During detection, historical data is input into the model; P_C is output; it is compared with the actual measurement; if the error is large, permutation is performed; and the updated data is used for the next prediction.

[0117] The error evaluation unit includes: a counter that records the number of times N the prediction error is less than Th within a set time T; a proportional calculator that calculates the proportion R = N / M, where M is the total number of detections; a comparator that outputs an abnormal signal when R < a preset proportion R_set; the preset proportion R_set is inversely proportional to a specified threshold Th, and R_set decreases when Th increases, which is automatically managed by the setting change unit; the abnormal signal triggers an audible and visual alarm.

[0118] The counter, proportional calculator, and comparator are integrated into the coprocessor of the control unit; the setting change unit is a software module. The counter quantifies the error frequency; the proportional calculator evaluates the detection reliability; the comparator identifies abnormalities; and the inverse proportional adjustment balances the sensitivity. After each detection, the error evaluation unit updates the count, calculates R, compares it with R_set, and alarms when abnormalities occur.

[0119] The storage unit includes: a prediction variable storage unit for storing n sets of concentration data and flow instructions that are updated on a rolling basis; a storage unit for recording predicted values, measured values, and error flags; a learning information base for storing normal state training datasets and model parameters; a data compression module for dimensionality reduction storage of historical data; and a retrieval interface for the correction processing unit to access the data optimization model.

[0120] The prediction variable storage and the storage storage are located within the extension board 16; the learning information library is synchronously backed up on the cloud server. The variable storage supports real-time prediction; the storage storage archives detection records; and the learning library enables continuous model optimization. In this application, detection data is written to the prediction variable storage; the prediction results are stored in the storage storage; and the correction processing unit periodically updates the model from the learning library.

[0121] The correction unit includes: an offset calculation unit, based on the difference ΔS between the actual upstream value Su and the target value S_target; a coordinate transformation unit, which converts ΔS into a displacement command for the hydraulic cylinder 12; an actuator, which drives the hydraulic cylinder 12 to fine-tune the position of the push rod 29 to compensate for installation offset; and a feedback sensor, which monitors the corrected position in real time to ensure that the difference ΔS < tolerance. The offset unit and the coordinate transformation unit are located within the control device; the actuator is connected to the hydraulic cylinder 12, and the feedback sensor is a displacement encoder for the push rod 29. After upstream detection, the difference ΔS is calculated; a displacement command is generated; the hydraulic cylinder 12 adjusts its position; and feedback is used to verify the correction effect.

[0122] The user coordinate system setting unit includes: a touchscreen interface displaying a 3D model of the pipeline; a coordinate input panel allowing the user to set the detection origin O, the X-axis (pipeline axial direction), and the Y-axis (radial direction); a data output interface transmitting the user coordinate system UF to the calibration unit 1 and the prediction value determination unit; and a calibration module that automatically optimizes the UF parameters based on measured data from the first shooting position P1. The touchscreen is mounted on the front of the detector housing; the data interface connects to the control device via USB. After the user sets the UF, the calibration unit performs calibration based on this reference; the prediction model aligns with the UF data.

[0123] The correction processing unit accesses the learning information stored in the storage unit. When the prediction error is less than Th for k consecutive times, the weights of the learned model are updated. The correction is based on the gradient descent algorithm to reduce prediction bias. The setting change unit monitors the R value of the error evaluation unit and automatically adjusts the specified threshold Th and the preset ratio R_set. When Th increases, R_set decreases to maintain constant detection sensitivity. The adjustment formula is Th_new = Th_old * (1 + ΔR), R_set_new = R_set_old / (1 + ΔR), where ΔR is the change in R.

[0124] The principle of this application is as follows: The mating sleeve 1 is fixed to the inlet of the pipe to be tested via the lower convex plate 2 and the first bolt 3. The laser scanning module emits a beam to scan the inner wall of the pipe, obtaining the inner diameter D and surface roughness. The pressure sensing module measures the gas pressure at the front end of the inlet pipe A26 and calculates the material friction coefficient F based on the inner diameter. The data processing unit integrates the inner diameter D, the friction coefficient F, and the preset sealing layer thickness T to generate the characteristic value H using the formula: H=K1*(1 / D)+K2*F+K3*T.

[0125] The characteristic value H quantifies the difficulty of detection. The larger the characteristic value H is, the higher the risk of pipeline leakage, providing a benchmark parameter for subsequent control.

[0126] The camera captures the flow field image at the inlet of intake pipe A26 to identify foreign object blockages; the ultrasonic flow velocity sensor measures the gas flow velocity Vu and initial concentration Cu in real time, and merges them into the actual value Su on the upstream side. Su is compared with the target value S_target to calculate the offset ΔS; the position of the push rod 29 is finely adjusted by hydraulic cylinder 12 to ensure that the gas input point is accurately aligned with ΔS < tolerance.

[0127] Hydraulic cylinder 12 pushes push rod 29 to open the valve of the pipeline to be tested, and gas enters the upper connecting cylinder 5 through air inlet pipe A26. The target opening degree V1 of regulating valve 18 is calculated based on the characteristic value H using the formula: V1=V0+K_H*H; when the characteristic value H increases, the sampling frequency of carbon dioxide concentration detector 21 is increased simultaneously and the alarm threshold is reduced.

[0128] Gas enters the detection chamber 31 through the inlet pipe B17, and the concentration value is measured in real time by the carbon dioxide concentration detector 21. n sets of historical concentrations C1-Cn and flow commands Q1-Qn are input into the learned model; the current concentration prediction value P_C is output; if the absolute error between the measured concentration and P_C is greater than or equal to the threshold Th, the measured value is replaced by P_C using a replacement mechanism.

[0129] The counter records N times the prediction error is less than Th within a set time T;

[0130] The calculation ratio R = N / MM represents the total number of detections; if R < preset ratio R_set, an audible and visual alarm is triggered. Th is inversely proportional to R_set; as Th increases, R_set decreases, ensuring constant sensitivity.

[0131] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

1. A cable shaft fireproof plugging sealing detection instrument, comprising an air inlet structure, an air feeding structure and a detection structure, one end of the air inlet structure is provided with the detection structure, the air inlet structure and the detection structure are connected through the air feeding structure, characterized in that, Also includes: A characteristic determination device, integrated into the inner wall of the docking sleeve on the air intake structure, is used to determine the characteristics of the pipeline to be tested in the cable shaft and output the characteristic value H. An upstream detection device is installed at the inlet of the air intake pipe A on the air intake structure. It is used to detect the actual parameters of the gas input point, including the gas flow rate and initial concentration. The actual parameters are used as the upstream actual value Su. A control device, embedded in the top of the detection chamber of the detection structure, adjusts the opening of the regulating valve on the gas supply structure and the sensitivity of the carbon dioxide concentration detector on the detection structure based on the characteristic value H and the actual value Su on the upstream side. The prediction value determination unit is located inside the detection chamber. The prediction value determination unit predicts the current gas concentration based on the learned model. An error evaluation unit, located within the control device, is used to calculate the proportion of times within a set time period when the difference between the predicted value and the measured concentration is less than a specified threshold. The storage unit, integrated within the extension plate on the detection structure, stores learning information and detection data, including characteristic value H, historical concentration values, and prediction error. The calibration unit, connected to the control device, calibrates the connection position of the intake pipe A based on the initial offset between the actual value Su on the upstream side and the target value; the calibration unit fine-tunes the position of the push rod through a hydraulic cylinder to ensure that the gas input point is aligned. The user coordinate system setting unit is located in the housing of the detector and provides a user interface for setting the detection coordinate system, including the axial and radial reference points of the pipe. The replacement mechanism is executed by the prediction value determination unit. When the prediction error is greater than or equal to a specified threshold, the measured concentration is replaced with the predicted value to maintain the normal detection benchmark.

2. The cable shaft fireproof sealing performance tester according to claim 1, characterized in that, The air intake structure includes: The docking sleeve has a lower convex plate on the outer side of its bottom end for connecting with the pipe to be tested, and an upper convex plate connected to the outer side of its top end. The bottom of the lower convex plate is provided with a first bolt through a mounting hole. The upper connecting cylinder is installed on the upper end of the docking sleeve by bolts and an upper convex plate; A rotating sleeve is fixed to the outer side of the top end of the upper connecting cylinder by a threaded rotation. The fixed plate is mounted to the top of the rotating sleeve by a second bolt; A hydraulic cylinder is connected to the bottom of the fixed plate and located inside the upper connecting cylinder; A large-diameter sleeve is located at the center of the bottom end of the upper connecting sleeve; Inlet pipe A is connected to the bottom of the large-diameter sleeve; A connecting nut is threaded onto the outside of the intake pipe A; The push rod is located inside the air intake pipe A, and its top end is connected to the power output end of the hydraulic cylinder through a connector and bolts; The detection structure includes: An extension plate is fixed to one end of the upper connecting cylinder; The detection chamber is installed at the bottom of the extension plate; The detection tube is connected to one side of the bottom of the detection chamber; A carbon dioxide concentration detector is mounted on the detection tube, with its detection end located inside the detection tube; The air delivery structure includes: Air inlet pipe B is used to connect the upper connecting cylinder and the detection chamber; A regulating valve, installed on the air inlet pipe B, is used to regulate the flow rate of gas entering the detection chamber; The shut-off valve A is located on the air inlet pipe B near the regulating valve and the detection chamber.

3. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The characteristic determination device includes: A laser scanning module is installed on the inner wall of the docking sleeve, which emits a laser beam to scan the inner wall of the docking sleeve and obtain the inner diameter and surface roughness data. The pressure sensing module is integrated at the front end of valve A in the intake pipe. It measures the gas input pressure at the front end of valve A in the intake pipe and calculates the material friction coefficient based on the inner diameter. The data processing unit, embedded on the main control board of the characteristic determination device, converts the inner diameter, friction coefficient, and sealing layer thickness into characteristic values ​​H. The characteristic value H is output to the control device to dynamically set the flow gain coefficient of the regulating valve.

4. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The upstream detection device includes: A camera and a flow rate sensor are mounted side-by-side inside the inlet flange of intake pipe A. The camera is used to capture images of the gas input point and identify flow field distribution and foreign object blockage; Flow velocity sensor, based on ultrasonic measurement of real-time flow velocity; The data fusion unit integrates image data and flow velocity data into the upstream actual value Su.

5. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The control device includes: The microprocessor runs a PID control algorithm and receives the characteristic value H and the actual value Su from the upstream side. The flow control module calculates the target opening degree V1 of the regulating valve based on the characteristic value H; The sensitivity adjustment module sets the sampling frequency and threshold of the carbon dioxide concentration detector. The feedback loop compares the measured concentration with the predicted value and dynamically fine-tunes the target opening V1.

6. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The predicted value determination unit includes: The learned model is complete, and the training data in the learned model consists of historical concentration sequences and flow command sequences under normal sealed conditions; and the learned model includes: The input layer receives n historical concentration values ​​C1-Cn and n flow command values ​​Q1-Qn; The output layer generates the current concentration prediction value P_C; The replacement unit replaces the measured value with P_C and updates the data in the storage unit when |P_C - measured concentration| ≥ a specified threshold Th; the specified threshold Th is set based on the model mean absolute error and the detection repeatability accuracy.

7. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The error evaluation unit includes: A counter that records the number of times, N, the prediction error is less than Th within a set time T. The ratio calculator calculates the ratio R = N / M, where M is the total number of tests. The comparator outputs an error signal when R < the preset ratio R_set; The preset ratio R_set is inversely proportional to the specified threshold Th. When Th increases, R_set decreases. This is automatically managed by the setting change unit. An abnormal signal triggers an audible and visual alarm.

8. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The storage unit includes: The prediction variable memory stores n sets of concentration data and flow instructions that are updated on a rolling basis; Storage memory for recording predicted values, measured values, and error flags; The learning information repository stores the normal state training dataset and model parameters; The data compression module reduces the dimensionality of historical data for storage. The retrieval interface allows the correction processing department to access the data optimization model.

9. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The correction unit includes: The offset calculation unit is based on the difference ΔS between the actual value Su on the upstream side and the target value S_target. The coordinate system transformation unit converts ΔS into a hydraulic cylinder displacement command; The actuator drives the hydraulic cylinder to fine-tune the position of the push rod, compensating for installation misalignment; Feedback sensors monitor the corrected position in real time to ensure that ΔS < tolerance.

10. The cable shaft fireproof sealing and sealing performance tester according to claim 1, characterized in that, The user coordinate system setting unit includes: The touchscreen interface displays a 3D model of the pipeline. The coordinate input panel allows users to set the detection origin O, the X-axis, and the Y-axis; The data output interface transmits the user coordinate system (UF) to the calibration unit and the prediction value determination unit; the calibration module automatically optimizes the UF parameters based on the measured data from the first shooting position P1.

Images