A filling pipeline blockage real-time early warning and self-adaptive adjustment system and method

Abstract

The application relates to the technical field of mine filling monitoring, in particular to a filling pipeline blockage real-time early warning and self-adaptive adjustment system, which comprises a distributed pressure sensing device, an edge data preprocessing module, a pressure gradient space feature extraction module, a blockage risk multistage identification module, a cooperative self-adjustment execution module and a comprehensive management and control platform. The application captures full-line pressure signals by means of unevenly arranged sensors, extracts particle sedimentation features through edge computing, and constructs a pressure gradient space matrix in combination with gravity potential compensation; the flowing blocked state is identified through an evolution model, and a pump speed increase, pulse water injection or air plug disturbance and other hierarchical interventions are implemented by a linkage execution mechanism. The application can eliminate monitoring blind spots, capture weak signals in the early stage of flow state deterioration and realize automatic closed-loop adjustment, effectively solve the problems of blockage positioning difficulty and early warning lag, and guarantee the safety of the filling conveying system.

Description

Technical Field

[0001] This invention relates to the field of mining engineering technology, and in particular to a real-time early warning and adaptive adjustment system and method for filling pipeline blockage. Background Technology

[0002] In mine backfilling operations, existing technologies generally employ a single pump station outlet pressure monitoring model, which has significant limitations. Relying solely on pressure data from the pump station outlet, the system cannot capture real-time pressure changes across multiple sections within the pipeline, resulting in a significant warning lag. Alarms are often triggered only when blockages have created severe resistance, missing the early intervention window. Furthermore, traditional monitoring methods lack spatial dimension information, making it difficult to accurately pinpoint the specific section where a blockage occurs, greatly complicating subsequent maintenance work.

[0003] Specifically, the shortcomings of existing technologies are reflected in the following aspects: Firstly, the early warning lag is significant. For example, the online monitoring device and elimination method for slug flow in risers disclosed in patent application number CN201610821162.0 uses a single pressure threshold alarm mechanism that can only be triggered when the pressure exceeds a set value, failing to identify the spatiotemporal evolution characteristics of the pressure gradient. Secondly, the blockage location accuracy is insufficient. Although the real-time monitoring system for liquid-solid two-phase flow scanning imaging of a dredger disclosed in patent application number CN202311515876.5 introduces a multi-sensor layout, the sensor spacing is too large and it lacks pressure gradient analysis logic, making it difficult to achieve segment-level location. Thirdly, there is a lack of closed-loop regulation capability. Existing systems can only adjust pumping parameters manually after triggering an alarm, unable to automatically execute graded regulation procedures based on real-time operating conditions. For instance, when a partial blockage occurs in a pipeline, traditional systems cannot dynamically correct the pressure reference value, nor can they link dilution water injection or emergency pressure relief devices, leading to a continuous worsening of the blockage problem.

[0004] To address the aforementioned problems, this invention proposes a real-time early warning and adaptive adjustment system for blockage in filling pipelines. This system achieves comprehensive acquisition of instantaneous pressure data of the filling slurry within the pipeline, enabling real-time calculation of the pressure gradient and construction of a spatial distribution matrix, dynamically correcting the standard pressure gradient envelope, thus improving the accuracy and timeliness of early warnings. It can accurately classify pipeline states, providing a reliable decision-making basis for the collaborative self-regulating execution module. Furthermore, it achieves real-time early warning and adaptive adjustment for filling pipeline blockage problems, significantly improving the safety and efficiency of mine production. Summary of the Invention

[0005] The present invention aims to solve the problems of delayed early warning, inability to accurately locate blocked sections, and lack of closed-loop adaptive adjustment methods caused by the use of single pump station outlet pressure monitoring in the prior art.

[0006] This invention provides a real-time early warning and adaptive adjustment system for filling pipe blockage, comprising: The distributed pressure sensing device includes several pressure sensor groups deployed along the pipeline transport direction to collect instantaneous pressure data of the filling slurry in the pipeline in real time; the sensor deployment density is higher than that of ordinary straight pipe sections in key areas prone to deposition, such as pipe bends, uphill sections, diameter change sections, and long horizontal sections. The edge data preprocessing module, located within the acquisition station, is used to perform high-frequency sampling and noise reduction on the raw pressure signal, generating a standard data packet containing spatial location information, timestamps, and pressure amplitude. The pressure gradient spatial feature extraction module is used to calculate the real-time pressure gradient of the pipe segment corresponding to adjacent sensor groups, construct the pressure gradient spatial distribution matrix of the entire pipeline, and has a built-in material characteristic compensation mechanism to dynamically correct the standard pressure gradient envelope. The multi-level blockage risk identification module has a built-in flow state evolution model, which is used to perform spatiotemporal correlation analysis based on the pressure gradient spatial distribution matrix, and divide the pipeline state into safe operation state, flow obstruction state, warning critical state and structural blockage state. The collaborative self-regulating execution module communicates and controls with the programmable controller of the filling pump station, the dilution water inlet at key nodes of the pipeline, the emergency pressure relief valve, and the emergency disturbance device based on the gas-liquid two-phase flow principle via industrial Ethernet. It is used to execute a graded adjustment program based on the pipeline status identification results. The central integrated management and control platform integrates a three-dimensional virtual simulation environment, which is used to map real-time data collected by distributed pressure sensing devices onto a three-dimensional model of the mine pipeline network, display the pressure gradient distribution of the entire pipeline in the form of a heat map, and optimize the judgment threshold in the flow state evolution model based on historical operating data.

[0007] Preferably, the distributed pressure sensing device adopts a non-uniform arrangement: in deep well pipelines with a vertical depth of over 500 meters, the arrangement spacing of the sensor groups within 10 meters before and after the horizontal bend at the bottom of the pipeline is 5 meters, and the arrangement spacing in ordinary horizontal conveying sections is 50 meters.

[0008] Preferably, the edge data preprocessing module uses a wavelet transform transient feature extraction algorithm to extract particle sedimentation feature signals and encapsulates the feature signals as independent fields into a standard data packet.

[0009] Preferably, the pressure gradient spatial feature extraction module performs gravitational potential energy compensation on the vertical pipe section; the compensation calculation is as follows: the instantaneous pressure value of the upstream sensor is added to the static pressure correction value generated by the real-time density of the slurry, the gravitational acceleration and the height difference between two adjacent sets of sensors, the instantaneous pressure value of the downstream sensor is subtracted, and then divided by the axial distance between the two sensors to obtain the corrected pressure gradient.

[0010] Preferably, the pressure gradient spatial feature extraction module acquires feedback data from the flow meter and concentration meter of the filling station in real time; when the slurry concentration increases by 1%, the corresponding standard pressure gradient benchmark value is automatically increased according to the preset rheological correction model.

[0011] Preferably, the blockage risk multi-level identification module determines the following rule: when the pipe section pressure gradient continuously exceeds 1.2 times the rated gradient within the first time threshold and the downstream pressure gradient decreases, it is determined to be a state of flow obstruction and a location identifier is output.

[0012] Preferably, the multi-level blockage risk identification module can also identify: pressure waves propagate from downstream to upstream and their amplitude increases step by step, indicating downstream pipe diameter reduction; and local pressure gradients approaching zero and upstream pressure surging, indicating structural blockage.

[0013] Preferably, the hierarchical adjustment logic of the collaborative self-regulating execution module is as follows: When flow is obstructed, increase the output frequency of the main filling pump by 3% to 5%; When the warning reaches a critical state, open the nearest dilution water injection port upstream of the risky pipeline section and inject pulsed high-pressure water at a pressure no less than 1.5 times the pipeline's rated pressure; When structural blockage occurs, stop the pump and open the emergency pressure relief valve downstream of the affected pipeline section.

[0014] Preferably, the collaborative self-regulating execution module is equipped with an emergency disturbance device; when water dilution is ineffective under the warning critical state, the emergency disturbance device injects compressed air into the pipeline to form an air embolism to disturb the deposits and change the rheological properties of the slurry.

[0015] Preferably, the central integrated control platform displays the pipeline pressure gradient distribution using a three-dimensional heat map, with red representing high resistance anomaly areas and green representing normal areas, and optimizes the judgment threshold based on historical data and machine learning.

[0016] The beneficial effects of this invention are as follows: This invention achieves focused monitoring of high-risk sections in pipeline networks (such as elbows and bends at the bottom of deep wells) through a non-uniform deployment of distributed pressure sensing devices. By increasing sampling density, it eliminates spatial identification blind spots in traditional monitoring schemes, enabling blockage risks to be located to specific physical pipe segments. Secondly, the application of wavelet transform algorithms in the edge data preprocessing module allows the system to capture weak particle settling signals in the early stages of slurry flow deterioration, issuing early warnings before significant fluctuations in total pressure, greatly extending the emergency response window.

[0017] Furthermore, the pressure gradient spatial feature extraction module incorporates gravitational potential energy compensation and rheological dynamic correction mechanisms. By real-time accounting for the static pressure contribution of vertical pipe sections and the influence of slurry concentration on rheological properties, it solves the problem of false alarms and missed alarms caused by material fluctuations during long-distance transportation, improving the physical accuracy of the pressure gradient spatial distribution matrix. In addition, the collaborative self-regulating execution module adopts a hierarchical adjustment logic, achieving a leap from "passive alarm" to "active early warning and automatic intervention." Through the combined application of multiple methods such as pump speed compensation, pulsed water injection, and air embolism disturbance, early siltation can be eliminated without interrupting the filling operation, reducing the probability of a complete shutdown for cleaning. Attached Figure Description

[0018] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention.

[0019] In the attached image:

[0020] Figure 1 This is an overall flowchart of the real-time early warning and adaptive adjustment system for filling pipe blockage of the present invention; Figure 2 This is a flowchart of the distributed pressure sensing device module of the present invention; Figure 3 This is a flowchart of the edge data preprocessing module of the present invention; Figure 4 This is a flowchart of the pressure gradient spatial feature extraction module of the present invention; Figure 5 This is a flowchart of the multi-level blockage risk identification module of the present invention; Figure 6 This is a flowchart of the collaborative self-regulating execution module of the present invention; Figure 7 This is a flowchart of the central integrated management and control platform module of the present invention. Detailed Implementation

[0021] The technical solution of the present invention will now be described with reference to the accompanying drawings. However, the described embodiments are only some embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0022] like Figures 1 to 7As shown, the real-time early warning and adaptive adjustment system for blockage in filling pipelines provided by this invention is built on an industrial Ethernet backbone communication network. Through a four-layer architecture comprising a hardware sensing layer, a data processing layer, a logic analysis layer, and an execution control layer, it achieves deep coupling, bidirectional data interaction, and closed-loop command linkage between each layer. Ultimately, it constructs a comprehensive safety assurance system for high-concentration slurry transportation scenarios in mines, integrating real-time blockage risk perception, accurate identification, graded early warning, and adaptive adjustment. The system consists of distributed pressure sensing devices, an edge data preprocessing module, a pressure gradient spatial feature extraction module, a multi-level blockage risk identification module, a collaborative self-adjusting execution module, and a central integrated management and control platform, all connected via industrial Ethernet for bidirectional data interaction and command flow.

[0023] In terms of hardware deployment and physical connections, such as Figure 2 As shown, the core of the distributed pressure sensing device lies in several pressure sensor groups non-uniformly distributed along the pipeline's transport direction. The pipeline is typically a high-strength, wear-resistant alloy steel pipe or a ceramic-lined composite pipe, whose pressure-bearing capacity must meet the high-pressure requirements of deep well gravity filling or pumped filling. The pressure sensor groups are installed through precision openings pre-drilled in the pipeline wall. To ensure measurement accuracy and prevent wear or blockage of the sensor sensing surface by slurry particles, this embodiment uses a high-pressure sealing flange for fixation. The sealing flange has a through-hole in its center, through which the probe of the pressure sensor group passes, with its sensing diaphragm perfectly flush with the inner wall of the pipeline. This flush installation eliminates local turbulence caused by sensor protrusions or depressions, thereby obtaining accurate static pressure and dynamic pulsation signals. A high-pressure resistant sealing gasket is installed between the sealing flange and the outer wall of the pipeline, and they are secured with staggered high-strength bolts to ensure no leakage occurs under operating pressures exceeding 10 MPa.

[0024] The placement of the pressure sensor arrays was optimized using fluid dynamics simulations. In critical sections of the pipeline with geometric variations, such as bends, uphill sections, diameter changes due to pipe diameter variations, and sections prone to particle segregation in long-distance horizontal transport, the density of the pressure sensor arrays was significantly increased. For example, at bends where deep wells transition from vertical to horizontal, pressure sensor arrays were densely arranged at 1-meter intervals within a 5-meter range before and after the bend. In straight and stable horizontal transport sections, they were evenly distributed at intervals of 50 to 100 meters. This non-uniform deployment strategy enables focused monitoring of high-risk siltation areas without unnecessarily increasing hardware costs.

[0025] Each pressure sensor group is electrically connected to a nearby data acquisition station via shielded twisted-pair cables. The data acquisition station is fixed to the sidewall of the tunnel or the support structure of the pipeline using brackets. The data acquisition station integrates an edge data preprocessing module, which includes a high-precision analog-to-digital converter, a high-speed digital signal processor, and an industrial-grade communication chip. The housing of the data acquisition station is made of explosion-proof, corrosion-resistant, and waterproof die-cast aluminum, with a protection rating of IP67, to adapt to the humid, water-spraying, and dusty environment of underground mines.

[0026] like Figure 3 As shown, the acquisition station performs high-frequency sampling on the raw analog current signals (such as 4 to 20 mA signals) from the pressure sensor array, with a sampling frequency of no less than 1000 Hz. The edge data preprocessing module uses a built-in wavelet transform algorithm to decompose the sampling sequence in real time, removing periodic noise generated by the reciprocating motion of the filling pump plunger and extracting non-stationary pulsating characteristic signals caused by particle collisions and settling within the slurry. The processed data is encapsulated into a standard data packet containing sensor spatial topology coding, microsecond-level timestamps, instantaneous pressure average values, and pulsating energy feature vectors, and uploaded to the subsequent processing module via industrial Ethernet.

[0027] like Figure 4 As shown, in the data processing logic, the pressure gradient spatial feature extraction module runs within a central server or high-performance industrial gateway. This module extracts the pressure spatial distribution profile of the entire pipeline by parsing data packets from different acquisition stations. Its core calculation logic involves calculating the pressure difference between two adjacent pressure sensor groups and dividing it by the physical distance between their pipe axes to obtain the pressure loss per unit length in each section, i.e., the pressure gradient. The calculation logic follows the formula below: ; In the above formula, G i P represents the real-time pressure gradient of the i-th pipe segment, in megapascals per meter; i P represents the instantaneous pressure value measured by the i-th group of high-frequency pressure sensors located upstream; i+1 L represents the instantaneous pressure value measured by the (i+1)th adjacent high-frequency pressure sensor group located downstream; i This represents the physical axial distance between the two sets of sensors, which is pre-stored in the system's static geographic information database.

[0028] Based on this, the pressure gradient spatial feature extraction module constructs a pressure gradient spatial distribution matrix by synchronously analyzing the pressure gradient of all pipe sections along the entire pipeline. This matrix dynamically reflects the energy dissipation distribution of the slurry during the pipeline transportation process. For a specific filling slurry, the system sets a standard pressure gradient envelope under normal flow conditions. This envelope is not a static value but is dynamically generated by a material characteristic compensation mechanism introduced by the pressure gradient spatial feature extraction module. This module obtains real-time data from the flow meters and concentration meters at the filling station, including the slurry's mass concentration, volumetric flow rate, and coarse particle ratio. Based on the real-time material parameters, the system calculates the theoretical resistance loss and uses it as the benchmark for the standard pressure gradient envelope. For example, when the slurry concentration increases by 1%, the system automatically increases the corresponding standard pressure gradient benchmark value, thereby avoiding false alarms caused by material fluctuations.

[0029] For vertical or steeply inclined pipe sections, this module executes a gravitational potential energy compensation algorithm. For vertical pipe sections, the formula for calculating the pressure gradient is modified as follows: ; in, This refers to the real-time density of the slurry. It is the acceleration due to gravity. This represents the height difference between two adjacent sets of sensors. By introducing a gravity term correction, the system can more accurately reflect the effective resistance loss caused by friction, avoiding the misleading influence of gravity pressure on the recognition logic.

[0030] Specifically, the system pre-stores the three-dimensional spatial coordinates of each pressure sensor group. During calculation, based on the real-time slurry density value monitored by the filling station, it calculates the static pressure contribution between the two sensor groups due to the height difference, subtracts it from the measured pressure difference, and thus separates the dynamic loss caused purely by flow resistance. Furthermore, this module is linked with the filling pump station's control system via industrial Ethernet to obtain real-time flow and concentration parameters. Utilizing the built-in Herschel Bulkley rheological model, the rated resistance baseline is dynamically corrected based on concentration fluctuations, ensuring dynamic adaptability in the determination of pressure gradient anomalies.

[0031] like Figure 5As shown, the multi-level blockage risk identification module receives the aforementioned pressure gradient matrix and classifies the pipeline status using a spatiotemporal correlation analysis algorithm. When the pressure gradient in a specific section suddenly increases, and this increase is accompanied by an increase in upstream pressure and a decrease in downstream pressure, the module determines that the section has entered a flow obstruction state. If this pressure gradient anomaly shows a gradual upstream propagation, and the pressure fluctuation amplitude increases exponentially with time, it is identified as a critical warning state. In extreme cases, if the pressure gradient in a certain section tends to infinity (i.e., the downstream completely loses pressure signal) and the upstream pressure reaches the pumping safety limit, it is determined to be a structural blockage state. The risk report generated by the identification module includes the precise physical location of the blockage, the estimated remaining time window for complete blockage, and the recommended intervention intensity.

[0032] The collaborative self-regulating execution module is key to achieving closed-loop control. For example... Figure 6 As shown, this module executes a multi-level linkage strategy based on the risk identification results. In the case of obstructed flow, the execution module sends a frequency conversion command to the programmable logic controller (PLC) of the filling pump station via industrial Ethernet. Upon receiving the command, the main filling pump of the filling pump station increases its output pressure in preset frequency steps (e.g., increasing by 2Hz each time), using the instantaneously enhanced power source to drive the locally accumulated particles back into a suspended flow state. Simultaneously, the central integrated management platform monitors the pumping pressure feedback in real time; once the pressure gradient returns to the normal envelope range, it stops increasing the frequency.

[0033] If the system enters a critical warning state, simply increasing the pumping pressure is insufficient to eliminate the siltation. At this point, the execution module activates the dilution water injection port. The dilution water injection port is connected to the pipeline via a T-junction, with a high-pressure constant-pressure water tank or high-pressure pump set connected upstream. Upon receiving the command, the electric high-pressure ball valve of the dilution water injection port opens rapidly, injecting high-pressure pulsed dilution water into the pipeline. The injection pressure is designed to be 0.5 MPa to 1.0 MPa higher than the instantaneous pressure inside the pipeline to ensure that the dilution water can be forcibly injected into the slurry. The injection of dilution water can rapidly reduce the solid mass concentration of the slurry in local sections, change its rheological parameters, and reduce the yield stress, thereby causing the siltation layer to collapse under the action of hydraulic scouring.

[0034] For more severe operating conditions, the system is equipped with an emergency disturbance device. This device utilizes the principle of gas-liquid two-phase flow disturbance and includes a high-pressure air storage tank and a rapid electromagnetic pulse valve. If the resistance does not decrease after a preset time (e.g., 30 seconds) of the dilution water injection process, the emergency disturbance device instantaneously releases high-pressure compressed air into the pipeline. This compressed air forms a large air embolus in the dense slurry. During its high-speed movement, the embolus experiences violent volume expansion and contraction. This mechanical disturbance effectively breaks up the coarse-particle deposits that have already formed a skeletal structure, forcing the slurry to regain its fluidity. In the final stage, when a structural blockage is identified, to prevent pipeline rupture due to pressure buildup, the collaborative self-regulating execution module immediately instructs the filling pump station to stop emergency pumping and opens the emergency pressure relief valve located downstream of the risk section or in a low-lying area. The emergency pressure relief valve systematically discharges the high-pressure slurry into the emergency slurry collection pool, releasing the enormous elastic potential energy accumulated in the system.

[0035] like Figure 7 As shown, the central integrated control platform provides operators with an intuitive monitoring interface. Based on GIS and BIM technologies, the platform recreates the complex underground pipeline system at a 1:1 scale in a virtual 3D space. By mapping real-time collected pressure gradient data onto the 3D model, the platform generates a dynamic heatmap. In the heatmap, green represents safe operating sections with normal resistance, yellow represents flow obstruction sections with slight pressure gradient fluctuations, orange represents critical warning sections with significant siltation risk, and red highlights dangerous points that may experience blockage. Operators can click on any node on the heatmap to view the historical pressure curves, spectrum analysis diagrams, and feature vectors extracted by the edge processing module for that pressure sensor group.

[0036] Furthermore, the machine learning engine built into the central integrated management and control platform is capable of self-evolution. Each time the system completes an adaptive adjustment process, the engine automatically records data characteristics from the start of an alert to the recovery process, including the material ratio (cement, tailings, and water ratio), flow rate, ambient temperature, and the effectiveness of the adjustment methods. Through deep learning on this historical big data, the system can autonomously correct the judgment thresholds in the multi-level blockage risk identification module. For example, in winter, when the circulating water temperature decreases and the slurry viscosity increases, the system automatically raises the alarm baseline of the pressure gradient to reduce false alarms caused by environmental factors and improve the intelligence level of the early warning system.

[0037] In actual operation, the system exhibits extremely high synergy. Take a real-world application scenario as an example: when the filling pump station is delivering 72% concentrated tailings paste to the stope 1000 meters underground, fluctuations in the fineness of the tailings at the bottom of the material silo cause a decrease in the fluidity of the paste entering the pipeline. The pressure sensor group located at the bottom bend of the deep well first detects the increased high-frequency pressure pulsation. Subsequently, the pressure gradient between this section and the adjacent upstream sensor begins to exceed the preset 1.2 times envelope. The acquisition station immediately completes edge computing and sends the early warning signal to the central integrated control platform via industrial Ethernet.

[0038] The central integrated control platform, through logical judgment, confirmed that the flow was obstructed and immediately issued an instruction to the collaborative self-regulating execution module. The execution module first coordinated with the filling pump station to increase the pumping frequency and thrust. If, after 5 minutes of observation, the pressure gradient continued to rise to 1.5 times the envelope, the identification module automatically raised the risk level to the critical warning state. At this point, the execution module instructed the dilution water injection port, located 50 meters upstream of the bend, to open, injecting pulsed dilution water at a flow rate of 20 cubic meters per hour. Under the combined action of the dilution water and high-pressure pumping, the locally accumulated paste was diluted and pushed towards the stope. The heat map on the central integrated control platform gradually changed from orange to yellow, and finally back to green. The system then automatically closed the dilution water injection port and notified the filling pump station to resume its normal output frequency. The entire process achieved automated closed-loop regulation without human intervention.

[0039] The system's components are closely related in terms of location and coordination. The pressure sensor array, acting as the sensing endpoint, determines the positioning accuracy through its distribution density; the sealing flange ensures the physical authenticity of the sensing; the acquisition station, as an edge node, processes massive amounts of raw data, reducing the transmission burden on the industrial Ethernet; the pressure gradient spatial feature extraction module and the multi-level blockage risk identification module constitute the system's "brain," responsible for reconstructing the physical flow state within the pipe from abstract pressure values; and the collaborative self-regulating execution module acts as the "hands and feet," eliminating risks through precise control of the filling pump station, dilution water inlet, emergency disturbance device, and emergency pressure relief valve.

[0040] To further enhance system reliability, the distributed pressure sensing device also features self-diagnostic capabilities. Each data acquisition station periodically sends heartbeat packets to the central integrated management platform and monitors the loop resistance of the pressure sensor group. If a sensor fails or experiences signal drift, the system automatically retrieves values ​​from adjacent sensors, performs virtual data compensation using a linear interpolation algorithm, and simultaneously issues a maintenance reminder on the platform interface, ensuring the continuity of the early warning logic even in the event of a localized hardware failure. This fault-tolerant mechanism, combined with multiple safeguards at the physical, data link, and application layers, enables the system to operate stably for extended periods during high-intensity mine backfilling operations.

[0041] Regarding electrical connections and protection details, all devices connected to the industrial Ethernet are equipped with industrial-grade surge protectors to resist electromagnetic interference generated by the start-up and shutdown of large underground electrical equipment. The electrical interfaces between the data acquisition station and the actuators uniformly use metal aviation plugs, which are equipped with internal positioning keys and locking nuts to effectively prevent poor contact caused by vibration. The vibration damping bracket inside the control box is made of high-polymer damping rubber material, capable of absorbing mechanical vibrations in the frequency range of 10Hz to 500Hz, protecting the internal precision electronic components from the impact of filling pump pulsations.

[0042] This system, through the coordinated operation of the aforementioned modules, solves the pain points of traditional backfill monitoring, which suffers from the limitations of "seeing pressure but not being able to see distribution, and being unable to control blockages." It not only achieves meter-level accuracy in locating blockages but also transforms traditional "post-accident handling" into "in-operation self-healing" through proactive early warning and multi-level adaptive adjustment, significantly improving the safety and production efficiency of mine backfilling operations. Whether under long-distance gravity transport or high-pressure pumping conditions, the system demonstrates superior flow monitoring and risk management capabilities.

[0043] To enable those skilled in the art to fully understand and implement this invention, the specific implementation principles of this invention are further supplemented below with a specific application scenario.

[0044] Step 1, Real-time Operating Condition Monitoring and Benchmark Establishment: During the initial phase of mine backfilling operations, the backfilling pump station begins injecting high-concentration tailings slurry into the pipeline. At this time, pressure sensor arrays distributed throughout the pipeline are precisely positioned using sealed flanges, ensuring their sensing diaphragms directly contact the slurry and acquire pressure signals. The acquisition station performs high-frequency capture of these signals and eliminates pumping pulsation interference, transmitting the clean pressure data to the central integrated management platform via industrial Ethernet. The central integrated management platform, combining the current flow rate and slurry concentration, uses a built-in rheological model to calculate the theoretical pressure gradient envelope for each pipeline segment under normal flow conditions. By comparing the measured pressure difference between adjacent pressure sensor arrays with the calculated value after height difference compensation, the system establishes a dynamic benchmark for the pressure distribution along the entire pipeline. This ensures that subsequent anomaly detection is based on a dynamic comparison of current material characteristics, rather than a static fixed threshold, thereby improving the sensitivity of early warnings.

[0045] Step 2, during the precise location and feature extraction of risk sections, when the rheological properties of the pipeline deteriorate due to changes in tunnel temperature or excessive slurry residence time in the middle section, the pressure value measured by the pressure sensor group downstream of the affected section will decrease ahead of the pump station end due to increased resistance, while the pressure value measured by the pressure sensor group upstream of the section will increase due to pressure buildup. The central integrated control platform calculates the rate of change of pressure difference between adjacent sensors in real time through the pressure gradient spatial feature extraction module. When it is found that the slope of the pressure gradient between pressure sensor groups numbered N and N+1 continuously exceeds 1.5 times the baseline, the system immediately identifies the specific section as a potential blockage point through spatial topology coding. This identification mechanism based on local pressure gradient abrupt changes rather than single-point pressure over-limit allows the system to capture the "pseudoplastic flow obstruction" characteristics caused by particle settling before the slurry has completely stopped flowing and before the total pressure of the pump station has significantly increased, thus gaining a critical time window for subsequent adaptive adjustment.

[0046] Step 3, during the graded adaptive adjustment and dynamic compensation, upon detecting local flow obstruction, the central integrated control platform issues a first-level adjustment command to the filling pump station via industrial Ethernet. The frequency converter of the filling pump station drives the main pump to increase its speed, increasing the initial power at the pipeline inlet and attempting to forcibly push away the initially accumulated material by increasing the pressure gradient along the entire line. Simultaneously, the central integrated control platform monitors the feedback data from the pressure sensor groups at both ends of the obstructed section in real time. If the pressure gradient does not return within a preset time (e.g., 3 minutes), the system automatically triggers the second-level adjustment, which involves opening the dilution water injection port located upstream of the obstructed section. High-pressure dilution water is injected into the pipeline through a three-way nozzle, rapidly reducing the solid mass concentration of the local slurry through hydraulic dilution, thereby reducing the yield stress of the slurry. The slurry resumes high-speed flow under lower resistance, carrying away the accumulated coarse particles and achieving self-healing of the local flow pattern.

[0047] Step 4: During extreme blockage intervention and safety pressure relief, if the pressure sensor group downstream of the blocked section continues to drop or even approaches zero after dilution water injection, it indicates that a structural skeletal blockage has formed in the pipeline. At this time, the central integrated control platform immediately instructs the emergency disturbance device to activate, releasing a high-pressure gas pulse into the pipe instantaneously. The cavitation effect and mechanical vibration generated by the high-pressure gas in the slurry can severely disturb the already settled particle layer, destroying its physical connection structure. Simultaneously, if the upstream pressure of the blocked section is detected to be close to the pipeline's rated pressure limit, the system will forcibly shut down the filling pump station and simultaneously open the nearest emergency pressure relief valve to prevent pipe rupture. The opening of the emergency pressure relief valve allows the high-pressure potential energy accumulated in the pipe to be released in a directional manner, avoiding the destructive impact of pressure waves on the pipe wall and protecting the safety of the downhole working environment.

[0048] Step 5, Heatmap Mapping and Strategy Optimization: Throughout the adjustment process, the central integrated control platform converts the data transmitted from the pressure sensor array across the entire pipeline into real-time visual information, presenting it as a heatmap in the 3D pipeline network model. Through the variations in color intensity of the heatmap, operators can intuitively observe the propagation path of abnormal pressure waves within the pipeline and the flow recovery process after the implementation of adjustment measures. After adjustment, the central integrated control platform automatically stores the evolution characteristics of the blockage risk, slurry parameters, and response parameters of the dilution water injection port and emergency disturbance device into the database. By comparing the pressure gradient curves before and after adjustment using machine learning algorithms, the system autonomously optimizes the logical threshold for the next warning, enabling more precise closed-loop adaptive control when facing filling materials with different proportions and viscosities, ensuring continuous and stable operation of long-distance filling operations.

Claims

1. A real-time early warning and adaptive adjustment system for filling pipe blockage, characterized in that, include: The distributed pressure sensing device includes several pressure sensor groups deployed along the pipeline transport direction to collect instantaneous pressure data of the filling slurry in the pipeline in real time; the sensor deployment density is higher than that of ordinary straight pipe sections in key areas prone to deposition, such as pipe bends, uphill sections, diameter change sections, and long horizontal sections. The edge data preprocessing module, located within the acquisition station, is used to perform high-frequency sampling and noise reduction on the raw pressure signal, generating a standard data packet containing spatial location information, timestamps, and pressure amplitude. The pressure gradient spatial feature extraction module is used to calculate the real-time pressure gradient of the pipe segment corresponding to adjacent sensor groups, construct the pressure gradient spatial distribution matrix of the entire pipeline, and has a built-in material characteristic compensation mechanism to dynamically correct the standard pressure gradient envelope. The multi-level blockage risk identification module has a built-in flow state evolution model, which is used to perform spatiotemporal correlation analysis based on the pressure gradient spatial distribution matrix, and divide the pipeline state into safe operation state, flow obstruction state, warning critical state and structural blockage state. The collaborative self-regulating execution module communicates and controls with the programmable controller of the filling pump station, the dilution water inlet at key nodes of the pipeline, the emergency pressure relief valve, and the emergency disturbance device based on the gas-liquid two-phase flow principle via industrial Ethernet. It is used to execute a graded adjustment program based on the pipeline status identification results. The central integrated management and control platform integrates a three-dimensional virtual simulation environment, which is used to map real-time data collected by distributed pressure sensing devices onto a three-dimensional model of the mine pipeline network, display the pressure gradient distribution of the entire pipeline in the form of a heat map, and optimize the judgment threshold in the flow state evolution model based on historical operating data.

2. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 1, characterized in that, The distributed pressure sensing device adopts a non-uniform arrangement: in deep well pipelines with a vertical depth of over 500 meters, the sensor group is arranged with a spacing of 5 meters within 10 meters before and after the horizontal bend at the bottom of the pipeline, and the spacing is 50 meters in ordinary horizontal conveying sections.

3. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 1, characterized in that, The edge data preprocessing module uses a wavelet transform transient feature extraction algorithm to extract particle sedimentation feature signals and encapsulates the feature signals as independent fields into standard data packets.

4. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 1, characterized in that, The pressure gradient spatial feature extraction module performs gravitational potential energy compensation on the vertical pipe section. The compensation calculation is as follows: the instantaneous pressure value of the upstream sensor is added to the static pressure correction value generated by the real-time density of the slurry, the gravitational acceleration and the height difference between two adjacent sets of sensors, the instantaneous pressure value of the downstream sensor is subtracted, and then divided by the axial distance between the two sensors to obtain the corrected pressure gradient.

5. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 1, characterized in that, The pressure gradient spatial feature extraction module acquires feedback data from the flow meter and concentration meter of the filling station in real time; when the slurry concentration increases by 1%, the corresponding standard pressure gradient benchmark value is automatically increased according to the preset rheological correction model.

6. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 1, characterized in that, The blockage risk multi-level identification module has the following judgment rule: when the pressure gradient of the pipe section continuously exceeds 1.2 times the rated gradient within the first time threshold, and the downstream pressure gradient decreases, it is judged as a flow obstruction state and a location mark is output.

7. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 6, characterized in that, The multi-level blockage risk identification module can also identify: pressure waves propagate from downstream to upstream and their amplitude increases step by step, indicating downstream pipe diameter reduction; local pressure gradients approach 0 and upstream pressure surges, indicating structural blockage.

8. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 1, characterized in that, The hierarchical adjustment logic of the collaborative self-regulating execution module is as follows: When flow is obstructed, increase the output frequency of the main filling pump by 3% to 5%; When the warning reaches a critical state, open the nearest dilution water injection port upstream of the risky pipeline section and inject pulsed high-pressure water at a pressure no less than 1.5 times the pipeline's rated pressure; When structural blockage occurs, stop the pump and open the emergency pressure relief valve downstream of the affected pipeline section.

9. A real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 8, characterized in that, The collaborative self-regulating execution module is equipped with an emergency disturbance device; when water dilution is ineffective under the warning critical state, the emergency disturbance device injects compressed air into the pipeline to form an air embolism to disturb the deposits and change the rheological properties of the slurry.

10. The real-time early warning and adaptive adjustment system for filling pipe blockage according to claim 1, characterized in that, The central integrated control platform displays the pipeline pressure gradient distribution using a three-dimensional heat map, with red indicating high resistance anomaly areas and green indicating normal areas, and optimizes the judgment threshold based on historical data and machine learning.

Images