Metal mine filling material concentration dynamic regulation method and system

Abstract

This invention discloses a method and system for dynamic control of backfill concentration in metal mines, belonging to the field of metal mine backfill technology. It includes: collecting a first type of physical signal output from a pipeline sensor and a second type of process parameter output from the backfill station preparation stage, respectively, to calculate the slurry mass concentration; performing a consistency check based on the absolute value of the deviation between the two concentration values; if the check fails, conflict arbitration is performed using signal confidence weights; the arbitration result is combined with the average concentration value when the check passes to construct a real-time rheological state characterization vector; performing a three-level state attribution judgment based on the relationship between the characterization vector and the three-level state thresholds; triggering a differentiated control response strategy according to the attribution result and performing feasibility verification under the constraints of a multi-constraint collaborative boundary management system; calculating the strength compliance confidence level based on actual strength data after each shift and writing it into a historical database. This method improves the reliability of concentration sensing and the comprehensiveness of rheological state characterization, achieving both strength compliance and cost control.

Description

Technical Field

[0001] This invention relates to the field of metal mine backfilling technology, specifically to a method and system for dynamic control of the concentration of metal mine backfill material. Background Technology

[0002] The concentration of backfill slurry in metal mines is a core process parameter for controlling the mechanical properties of the backfill and ensuring safe underground transport. Existing concentration control technologies mostly rely on single detection methods, such as online measurement using ultrasonic concentration meters installed on the transport pipeline, or indirect concentration estimation through material balance calculations using weighing and flow meters at the backfill station. When the gas content of the slurry fluctuates, the sensor probe scales, or the feed rate changes transiently, the output value of a single detection channel will show significant deviations, significantly reducing the reliability of the concentration feedback value. Directly relying on such signals for control is highly prone to misoperation. Current control methods typically employ a uniform shutdown protection strategy or a significant adjustment of the water supply when the concentration exceeds the limit, without establishing a graded response mechanism based on the degree of deviation of the slurry's rheological state. Backfill slurries with different deviations exhibit fundamentally different rheological characteristics and pipeline transport risks. Using intervention measures of the same scale often results in over-regulation or under-response, not only affecting production efficiency but also potentially causing insufficient backfill strength or pipe blockage accidents. Furthermore, the feasibility verification of actual control commands is often overlooked. There is a lack of systematic verification regarding whether adjustments to the proportioning parameters will cause issues such as excessive pipeline flow velocity, overloaded mixing power, or concentration exceeding the process window. The synergistic constraints between pipeline transport safety boundaries, mixing equipment capacity boundaries, and concentration constraints are not incorporated into real-time decision-making. Meanwhile, fluctuations in the physical properties of backfill materials and differences in mining conditions mean that the optimal proportioning parameters are not fixed for each shift. Existing concentration control methods fail to effectively utilize historical production data for feedback optimization. Setting parameters independently for each shift makes it difficult to inherit and iterate low-cost, high-reliability proportioning schemes from accumulated backfill strength compliance records, resulting in unclear cement usage expectations and significant fluctuations in the backfill strength compliance rate. Summary of the Invention

[0003] The present invention aims to provide a method and system for dynamic control of the concentration of filling material in metal mines, in order to solve the problems of poor reliability of concentration detection, lack of rheological state classification and multi-constraint collaborative verification of control strategies, and inability to use historical intensity data to optimize the proportioning parameters in the prior art.

[0004] To achieve the above objectives, this invention provides the following technical solution: This invention provides a method for dynamic control of backfill concentration in metal mines, comprising: collecting physical signals output from pipeline sensors and process parameters output from the backfill station preparation stage, and calculating the slurry mass concentration to obtain two concentration estimates based on different measurement principles. A consistency check is performed based on the absolute value of the deviation between the two concentration values. If the check fails, conflict arbitration is conducted using a signal confidence weighting method, and the arbitration result is combined with the average concentration value when the check passes to construct a real-time rheological state characterization vector, thereby solving the problem of single measurement methods being susceptible to environmental interference and having low data reliability. Based on the relationship between this characterization vector and three-level state thresholds, a three-level state attribution judgment is performed. Differentiated control response strategies are triggered according to the attribution result, and feasibility verification is performed under the constraints of a multi-constraint collaborative boundary management system, ensuring that the control actions meet both concentration and rheological requirements while always remaining within the safe range of pipeline transportation. After each shift, the strength compliance confidence level is calculated based on actual strength data and written into a historical database, providing traceable quality basis for the next shift. Before the next shift starts, the database is searched for the best historical mix design that matches the current working conditions. If the confidence level of the searched design does not reach the threshold, a closed-loop iterative convergence adjustment of the mix design parameters is initiated until the strength prediction value meets the standard. This allows the historical experience to be used to quickly approximate the best mix design, while closed-loop correction ensures that the strength of the filling body reliably meets the design specifications.

[0005] As a preferred approach, a first concentration estimate is obtained by interpolation from the first type of physical signals, composed of pipeline ultrasonic attenuation spectrum data and pipeline differential pressure data, using a pre-calibrated attenuation-concentration mapping table. A second concentration estimate is obtained by calculating the ratio of solid dry material mass to total mass using the second type of process parameters, composed of the filling station feed rate and water flow rate. Timestamps are then added to both concentration estimates. This synchronous acquisition of multi-source heterogeneous data lays the foundation for concentration fusion and verification.

[0006] As a technical solution of this invention, the absolute value of the deviation between the first and second estimated concentration values ​​is calculated. When the deviation is less than a preset deviation threshold, the arithmetic mean of the two concentration values ​​is taken as the fused concentration value. When the deviation is greater than or equal to the threshold, the signal-to-noise ratio of the ultrasonic attenuation spectrum data and the fluctuation amplitude of the feed rate sensor are obtained. A first confidence level and a second confidence level are assigned, and the two concentration values ​​are weighted and averaged using the confidence level as the weight to obtain the arbitration concentration value. The fused concentration value or arbitration concentration value is combined with the current pipeline flow rate and stirring torque value to construct a real-time rheological state characterization vector. This processing method can effectively eliminate instantaneous interference and data inconsistency, and improve the ability of the characterization vector to reflect the true state of the slurry.

[0007] Furthermore, concentration and flow rate components are extracted from the real-time rheological state representation vector, and the concentration deviation ratio and flow rate deviation are calculated. When the concentration deviation ratio is within the first deviation range and the flow rate deviation does not exceed the first deviation threshold, it is determined to be a Level 1 state, triggering a response strategy that maintains the current mixing parameters unchanged and continuously monitors. When the concentration deviation ratio or flow rate deviation is within the second deviation range, it is determined to be a Level 2 state, triggering a response strategy that adjusts the water addition in a single adjustment step. When the concentration deviation ratio is within the third deviation range or the flow rate deviation exceeds the third deviation threshold, it is determined to be a Level 3 state, triggering a response strategy that stops the filling operation and initiates pipeline flushing. Through the three-level state division, a graded response from fine-tuning to emergency shutdown is achieved, balancing production continuity and system safety.

[0008] In the above scheme, the multi-constraint collaborative boundary management system includes concentration constraint boundaries, rheological constraint boundaries, and pipeline transportation safety boundaries. The concentration constraint boundary is composed of the absolute range of fluctuation of the target concentration value, the rheological constraint boundary is composed of the critical flow velocity and critical yield stress values ​​of pipeline transportation, and the pipeline transportation safety boundary is composed of the pipeline pressure resistance limit and the maximum output power value of the delivery pump. The adjusted proportioning parameters under the secondary or tertiary states are sequentially mapped to the concentration value, pipeline flow velocity value, and delivery pump output power value, and compared with the above boundaries item by item. If all constraints are met, the verification is passed; if any constraint is not met, the adjustment amount in the proportioning parameters is halved by the step size and verified again. This halving is repeated until it passes or the adjustment amount is returned to zero. This constraint verification mechanism ensures that the control commands do not exceed the process and equipment safety boundaries, effectively preventing pipe blockage, segregation, and equipment overload.

[0009] As a further preferred option, at the end of each shift, the measured sequence of uniaxial compressive strength values ​​of the cast test blocks for that shift is obtained. The average concentration value and average feed rate value are extracted and matched with the predicted strength value interval under the same working condition in a pre-established strength prediction surface. The proportion of test blocks whose measured value sequence falls within this interval is calculated as the confidence level for strength compliance and is stored in association with the shift number, average concentration value, average feed rate value, and mix proportion parameters. This quantifies the strength quality evaluation into a confidence index, facilitating the retrieval and utilization of historical data.

[0010] More specifically, before the next shift starts, the design strength grade of the filling material and the pipeline transportation distance parameters of the stope to be filled are obtained. Based on this, records with the same stope conditions are screened from the historical database. The record with the highest confidence in achieving the strength standard and the lowest cement addition is selected as the historically optimal mix design, and its concentration and feed rate values ​​are used as the initial mix design parameters. This method quickly determines the initial parameters by leveraging historical best practices, avoiding blind trial mixing and shortening the adjustment time.

[0011] In terms of strength closed-loop iteration, if the confidence level of strength compliance calculated after the current shift is lower than the preset threshold, the water-cement ratio is decreased in fixed steps, while the amount of cement added is increased simultaneously. The predicted strength value is then recalculated using the strength prediction surface. A single iteration is completed when the predicted strength value meets the design grade requirements and the adjusted mix proportion parameters satisfy the multi-constraint boundaries. This result serves as the mix proportion parameter for the next shift. This iteration is repeated until the confidence level of strength compliance is not lower than the threshold. This iterative convergence process ensures strength compliance while minimizing the amount of cementitious materials used, achieving a dual optimization of economy and quality.

[0012] This invention also provides a dynamic control system for the concentration of backfill material in metal mines, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the aforementioned method. This system can integrate online pipeline signals and preparation process parameters, construct a rheological state characterization vector in real time, and perform three-level state judgment and graded control. It automatically verifies the feasibility of control under multiple constraint boundaries and uses historical data to match the optimal proportioning scheme and perform closed-loop iterative correction, thereby maintaining the stability of the backfill slurry concentration and the reliability of the backfill strength under complex working conditions.

[0013] The technical effects and advantages provided by the present invention in the above technical solution are as follows: The first type of physical signal, composed of pipeline ultrasonic attenuation spectrum data and pipeline differential pressure data, and the second type of process parameters, composed of filling station feed rate and water flow rate, are used to estimate concentration, resulting in first and second estimated concentration values. A consistency check is performed on the two concentration values. When the deviation exceeds the limit, the signal-to-noise ratio of the ultrasonic attenuation spectrum data and the fluctuation amplitude of the feed rate sensor are extracted and assigned confidence levels. Conflict arbitration is then performed using a confidence-weighted method to obtain an arbitration concentration value. If the check passes, the average concentration value is taken. The fused concentration value is combined with pipeline flow velocity and stirring torque values ​​to construct a real-time rheological state characterization vector. By jointly utilizing two types of heterogeneous signal sources and conducting reliability assessments, the concentration distortion caused by single sensors due to bubbles, scaling, or material fluctuations is overcome. The concentration characterization results are more robust, and the multi-dimensional structure of the rheological state vector can fully reflect the flow and resistance characteristics of the slurry, providing reliable input for state identification.

[0014] Based on the deviation ratio of the concentration component in the characterization vector from the preset target concentration and the deviation of the flow rate component from the preset flow rate threshold, the operating state is divided into three levels. Different response strategies are triggered for different levels, such as maintaining parameter monitoring, adjusting water volume in a single step, stopping operation and flushing the pipeline. The feasibility of the adjusted ratio parameters is verified in a multi-constraint collaborative boundary management system. If the verification fails, a feasible solution is approximated through a step-halving cycle. This hierarchical control and collaborative boundary verification mechanism allows small deviations to be stabilized through fine-tuning and rapid intervention in critical states, avoiding excessive disturbances or response lags caused by uniform coarse adjustments. At the same time, the collaborative verification of multiple constraint boundaries such as concentration, flow rate, and delivery power ensures that the control commands do not cause secondary risks such as pipeline overpressure or agitation overload at the execution level.

[0015] At the end of each shift, based on the measured uniaxial compressive strength sequence of the filled slurry casting test blocks from that shift, the predicted strength range under the same concentration and feed rate in the strength prediction surface is matched, the proportion falling into the range is calculated, the strength compliance confidence level is obtained, and the mix proportion parameters are associated and stored in the historical database. Before the start of the next shift, using the design strength level and pipeline transportation distance as search conditions, the record with the highest strength compliance confidence level and the lowest cement addition is selected from the historical records as the initial mix proportion scheme. If the confidence level is lower than the threshold at the end of the shift, closed-loop iterative convergence is initiated, the water-cement ratio is decreased and the cement addition is increased simultaneously, and the predicted strength is recalculated in the strength prediction surface until compliance is achieved. The adjustment result that satisfies multiple constraint boundaries is used as the mix proportion parameters for the next shift. By quantifying the shift strength results into compliance confidence levels and accumulating historical data, the mix proportion parameters can be matched and iteratively optimized based on real strength feedback. Under the premise of meeting the design strength level, the cement dosage is continuously reduced, the filling cost is reduced, and the consistency of strength compliance gradually improves with the accumulation of production data. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0017] Figure 1 This is a flowchart of a method for dynamically controlling the concentration of filling materials in metal mines; Figure 2 This is a flowchart of the dual-channel calculation and time synchronization of slurry mass concentration; Figure 3 This is a flowchart of the three-level classification and differentiated control strategy for the rheological state of filling slurry; Figure 4It is a dynamic change curve of the slurry mass concentration calculated by dual channels; Figure 5 It is a curve showing the dynamic change between the slurry concentration deviation ratio and the flow rate deviation. Figure 6 It is a curve showing the change between the mapped concentration value and the water supply flow rate adjustment after the ratio parameters are adjusted; Figure 7 This is a statistical chart showing the distribution of measured values ​​of uniaxial compressive strength of filling slurry for each shift and the confidence level of strength compliance. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0019] See Figure 1 This invention provides a method for dynamic control of filling material concentration in metal mines. The overall implementation scheme is as follows: During the filling operation, the first type of physical signal output from the pipeline sensor and the second type of process parameter output from the filling station preparation stage are collected. The mass concentration of the slurry is calculated using these two different sources of data. After obtaining the two concentration calculation values, the absolute value of the deviation between the two is calculated, and a consistency check is performed accordingly. If the check fails, conflict arbitration is performed with the signal confidence level as the weight to generate an arbitration concentration value; if the check passes, the average of the two concentration values ​​is directly calculated. The arbitration result or the average concentration value is combined with the current pipeline flow rate and stirring torque value to construct a real-time rheological state characterization vector. Based on the numerical relationship between each component in this characterization vector and the preset three-level state threshold, a three-level state classification judgment is performed to determine whether the current slurry is in a first-level state, a second-level state, or a third-level state. Differentiated control response strategies are triggered according to the state: in the first-level state, the current ratio parameters are maintained; in the second-level state, the water addition is adjusted in a single step; and in the third-level state, filling is stopped and pipeline flushing is initiated. Before issuing control commands at level two or three, the adjusted mix proportions are placed under a multi-constraint collaborative boundary management system for feasibility verification. Only commands that pass verification are output to the actuator. After each shift, the system calculates the confidence level of strength compliance based on the actual measured intensity data and writes the shift data into the historical database. Before the next shift starts, the system retrieves the historically optimal mix proportion scheme that matches the current working conditions from the historical database as the initial mix proportion parameters. If the confidence level of strength compliance in the previous shift does not reach the preset threshold, a closed-loop iterative convergence process is initiated, continuously adjusting the mix proportion parameters until the predicted intensity value meets the standard.

[0020] In specific implementation, please refer to Figure 2 The step of calculating the slurry mass concentration by collecting the first type of physical signals output by the pipeline sensors and the second type of process parameters output by the filling station preparation stage is implemented in the following way: The pipeline sensors are an ultrasonic attenuation spectrum measuring device and a differential pressure transmitter installed in the filling pipeline. The first type of physical signal consists of ultrasonic attenuation spectrum data output by the ultrasonic attenuation spectrum measuring device and pipeline differential pressure data output by the differential pressure transmitter. The ultrasonic attenuation spectrum data is a sequence of amplitude attenuation values ​​of ultrasonic waves of different frequencies propagating in the slurry, and the pipeline differential pressure data is the pressure difference between two measuring points in the pipeline. An attenuation-concentration mapping table is pre-established. The construction process of the attenuation-concentration mapping table is as follows: In a laboratory environment, multiple standard slurry samples with known mass concentrations are prepared. Using the same model of ultrasonic attenuation spectrum measuring device and differential pressure transmitter as on-site, the ultrasonic attenuation values ​​and corresponding pipeline differential pressure values ​​of each standard slurry sample with known mass concentration are measured at multiple frequency points. The ultrasonic attenuation values, pipeline differential pressure values, and mass concentrations are established and stored as a two-dimensional mapping table. Each pair of ultrasonic attenuation values ​​and pipeline differential pressure values ​​in the table uniquely corresponds to a slurry mass concentration value. During the actual filling process, ultrasonic attenuation spectrum data and pipeline differential pressure data are collected in real time. The measured values ​​of the attenuation coefficient at specific frequency points are extracted from the ultrasonic attenuation spectrum data. The measured values ​​of the attenuation coefficient at specific frequency points and the real-time pipeline differential pressure value are used as query conditions. The four neighboring nodes surrounding the query conditions are found in the attenuation-concentration mapping relationship table. Bilinear interpolation is performed on the mass concentration values ​​corresponding to the four neighboring nodes. The bilinear interpolation calculation output is recorded as the first concentration estimate.

[0021] The second type of process parameter output from the filling station preparation stage consists of the filling station feed rate and the filling station water flow rate. The filling station feed rate is collected by a mass flow meter installed at the discharge end of the dry material feeder, representing the mass flow rate of the solid dry material entering the mixing equipment per unit time. The filling station water flow rate is collected by an electromagnetic flow meter installed on the water supply pipeline, representing the mass flow rate of the water added to the mixing equipment per unit time. The solid dry material mass flow rate is determined based on the collected filling station feed rate. The unit is kilograms per second; the mass flow rate of water is determined based on the water supply flow rate of the filling station. The unit is kilograms per second. Second concentration estimate. The calculation formula is expressed as:

[0022] in, This represents the mass flow rate of the solid dry material. The mass flow rate of water is given. The calculated second concentration estimate is a dimensionless ratio, representing the mass percentage of dry solids in the slurry.

[0023] Timestamps are added to both the first and second estimated concentration values. These timestamps originate from the master clock of the filling control system and are formatted as a string representing year-month-day hour:minute:second.millisecond, with millisecond precision. The timestamp addition operation is performed at the same time the first and second estimated concentration values ​​are generated, binding the timestamps to the corresponding estimated concentration values ​​as data pairs. This synchronizes the first and second estimated concentration values ​​across time, enabling the pairing and retrieval of dual-channel concentration data at the same moment in subsequent processes.

[0024] See Figure 4 In the figure, the horizontal axis represents the time sequence (sampling point) during the filling operation, and the vertical axis represents the slurry mass concentration, with a value range of approximately 0.66 to 0.82. The legend identifies two concentration estimation curves, where the dashed blue curve represents the first estimated concentration value calculated based on ultrasonic attenuation spectrum data and pipeline differential pressure data, and the red dotted curve represents the second estimated concentration value calculated based on the filling station feed rate and water flow rate.

[0025] Judging from the trend of the curve, the overall fluctuation of the estimated value of the first concentration is greater than that of the estimated value of the second concentration. Especially in the range of approximately 90 to 130 sampling points, the estimated value of the first concentration shows a significant upward peak, reaching a maximum of about 0.80, while the estimated value of the second concentration changes relatively steadily in this range, remaining around 0.72.

[0026] Between approximately sampling points 330 and 360, the first concentration estimate showed a significant trough, reaching a low of approximately 0.67, far below the stable level of the second concentration estimate in the same range (approximately 0.71 to 0.73). This also reflects a substantial discrepancy between the two concentration estimates, necessitating the triggering of a conflict arbitration mechanism. Apart from these two abnormal intervals, the two concentration estimates maintained good consistency across the remaining time series, with concentration values ​​evenly distributed between 0.70 and 0.74 and minimal fluctuations. This indicates that the measurement status of the two data sources was stable during this period, resulting in a high degree of consistency in the estimates.

[0027] In practical implementation, the consistency check is performed based on the absolute value of the deviation between the two concentration values. If the check fails, conflict arbitration is performed using signal confidence weighting. The arbitration result is then combined with the average concentration value when the check passes to construct a real-time rheological state representation vector. This is achieved through the following steps: First and second estimated concentration values ​​generated at the same time are extracted from the data storage area with attached timestamps. The absolute value of the deviation between the first and second estimated concentration values ​​is calculated. The preset deviation threshold used to calculate the absolute deviation value is 0.03. The reason for setting the preset deviation threshold to 0.03 is that when the slurry mass concentration is expressed as a mass fraction, a deviation of 0.03 corresponds to a concentration difference of three percentage points. This difference exceeds the normal measurement error range, indicating that at least one of the two signals is significantly interfered with. The calculated absolute deviation value is compared with the preset deviation threshold. When the absolute deviation value is less than 0.03, the consistency check result is passed. The arithmetic mean of the first and second estimated concentration values ​​is calculated and used as the fused concentration value.

[0028] When the absolute value of the deviation is greater than or equal to 0.03, the consistency check result is considered unsuccessful, and a conflict arbitration procedure is initiated. The conflict arbitration procedure obtains the signal quality indices corresponding to the first type of physical signals and the second type of process parameters. The signal quality indices corresponding to the first type of physical signals are the signal-to-noise ratio (SNR) of the ultrasonic attenuation spectrum data. The SNR of the ultrasonic attenuation spectrum data is defined as the ratio of the effective signal power to the background noise power in the received ultrasonic signal. A higher SNR indicates a more stable operating state of the ultrasonic attenuation spectrum measurement device and less environmental interference in the measurement results. The signal quality indices corresponding to the second type of process parameters are the fluctuation amplitude of the feed rate sensor. The fluctuation amplitude of the feed rate sensor is defined as the ratio of the peak-to-valley difference of the feed rate measurement value within a preset sliding window to the average value of the feed rate measurement value within the window. A smaller fluctuation amplitude indicates a more stable output of the feed rate sensor.

[0029] A first confidence level is assigned based on the signal-to-noise ratio (SNR) of the ultrasonic attenuation spectrum data, with a value limited to between 0.3 and 0.7. When the SNR of the ultrasonic attenuation spectrum data is greater than or equal to 25 dB, the first confidence level is 0.7; when the SNR is less than or equal to 10 dB, the first confidence level is 0.3; and when the SNR is between 10 dB and 25 dB, the first confidence level is calculated using linear interpolation. A second confidence level is assigned based on the fluctuation amplitude of the feed rate sensor, with a value limited to between 0.3 and 0.7. When the fluctuation amplitude of the feed rate sensor is less than or equal to 0.05, the second confidence level is 0.7; when the fluctuation amplitude is greater than or equal to 0.15, the second confidence level is 0.3; and when the fluctuation amplitude is between 0.05 and 0.15, the second confidence level is calculated using linear interpolation. The first confidence level and the second confidence level reflect the degree of reliability of the first concentration estimate and the second concentration estimate in the conflict arbitration process, respectively.

[0030] After obtaining the first and second confidence levels, a weighted average is calculated on the first and second concentration estimates, using the first and second confidence levels as weights. The arithmetic result of the weighted average calculation is used as the arbitration concentration value. The weighted average calculation formula is expressed as follows:

[0031] in, For arbitration concentration values, As the first confidence level, This is the estimated value for the first concentration. As the second confidence level, This is the estimated value for the second concentration. Molecular portion. The numerator represents the confidence-weighted contribution of the first concentration estimate. The denominator represents the confidence-weighted contribution of the second concentration estimate. The sum of the first and second confidence levels is used to normalize the weighted contribution.

[0032] After concentration fusion or conflict arbitration is completed, the current pipeline flow velocity and stirring torque values ​​are obtained. The pipeline flow velocity value is collected in real time by an electromagnetic flowmeter installed in the filling pipeline, representing the linear velocity of the slurry flow in the pipeline. The stirring torque value is collected in real time by a torque sensor installed on the drive shaft of the stirring equipment, representing the torque required for the stirring blades to overcome the slurry resistance. The fused concentration value or arbitration concentration value is combined with the current pipeline flow velocity value and stirring torque value to form a real-time rheological state characterization vector containing three-dimensional components. In the real-time rheological state characterization vector, the first component is the slurry concentration value determined after consistency verification or conflict arbitration, the second component is the pipeline flow velocity value, and the third component is the stirring torque value. The three components are arranged in a fixed order in the vector and are labeled with a uniform timestamp.

[0033] In specific implementation, please refer to Figure 3 The process of performing three-level state attribution judgment based on the relationship between the real-time rheological state characterization vector and the three-level state threshold, and triggering differentiated control response strategies according to the attribution results, is implemented in the following way: Concentration and velocity components are extracted from the real-time rheological state characterization vector. The concentration component is the slurry concentration value determined after consistency verification or conflict arbitration in the real-time rheological state characterization vector, and the velocity component is the pipeline velocity value collected by the electromagnetic flowmeter in the real-time rheological state characterization vector. The process of calculating the concentration deviation ratio is as follows: Obtain the preset target concentration value, which is determined by the mix proportion scheme corresponding to the design strength grade of the filling body; calculate the difference between the concentration component and the preset target concentration value; divide the absolute value of the difference by the preset target concentration value, and the resulting quotient is the concentration deviation ratio. The process of calculating the velocity deviation is as follows: Obtain the preset velocity threshold, which is calculated based on the pipe diameter parameters and the empirical formula for critical settling velocity; calculate the difference between the velocity component and the preset velocity threshold; divide the absolute value of the difference by the preset velocity threshold, and the resulting quotient is the velocity deviation. The formulas for calculating the concentration deviation ratio and the flow rate deviation are expressed as follows:

[0034] in, This is the concentration deviation ratio. The concentration component is extracted from the real-time characterization vector of the rheological state. To preset the target concentration value, For flow velocity deviation, The velocity component is extracted from the real-time rheological state characterization vector. Preset flow rate threshold. Concentration deviation ratio. Reflects the relative degree of deviation of the current slurry concentration from the target concentration, and the flow rate deviation. It reflects the relative degree to which the current flow velocity in the pipeline deviates from the critical flow velocity.

[0035] The calculated concentration deviation ratio and flow rate deviation are compared with preset three-level intervals and thresholds for judgment. The first deviation interval is set as the range where the concentration deviation ratio is greater than or equal to 0 and less than 0.05, indicating that the concentration deviation ratio is within ±5%, and the concentration fluctuation is within an acceptable normal range. The first deviation threshold is set to 0.10, indicating that the flow rate deviation does not exceed 10%. When the concentration deviation ratio is within the first deviation interval and the flow rate deviation is less than or equal to 0.10, the current slurry is determined to be in the first-level state. The first-level state indicates that both the slurry concentration and the pipeline flow rate are within the normal fluctuation range, and the filling system is operating smoothly.

[0036] The second deviation range is set as follows: the concentration deviation ratio is greater than or equal to 0.05 and less than 0.15. This range indicates a moderate concentration deviation, with the concentration deviation ratio between 5% and 15%. The second deviation range is set as follows: the flow rate deviation is greater than 0.10 and less than or equal to 0.25. This range indicates a flow rate deviation, with the flow rate deviation between 10% and 25%. When either the concentration deviation ratio or the flow rate deviation falls within the second deviation range, the slurry is determined to be in a secondary state. A secondary state indicates a significant deviation in slurry concentration or pipeline flow rate, suggesting a deteriorating trend in the filling system's operating condition, requiring intervention and adjustment.

[0037] The third deviation range is set to a concentration deviation ratio greater than or equal to 0.15, indicating a significant concentration deviation of 15%. The third deviation threshold is set to 0.25, representing a flow rate deviation exceeding 25%. When the concentration deviation ratio is within the third deviation range, or the flow rate deviation is greater than 0.25, the current slurry is determined to be in a level three state. A level three state indicates a significant deviation in slurry concentration or pipe flow rate, and continued filling operations pose a risk of pipe blockage or segregation.

[0038] After determining the three-level state attribution, a differentiated control response strategy is triggered based on the attribution result. When the state is determined to be Level 1, the first control response strategy is triggered. This strategy maintains the current mixing parameters unchanged, performs no adjustments, collects signals from pipeline sensors and process parameters from the filling station preparation stage, and continuously updates the rheological state characterization vector and determines the three-level state attribution. When the state is determined to be Level 2, the second control response strategy is triggered. This strategy adjusts the water addition in single-step increments. The single-step increment is set to 8% of the current water flow rate. The rationale for setting the single-step increment to 8% is that an 8% change in water flow rate can cause an adjustment of approximately 1.5 to 2.5 percentage points in the slurry concentration. This adjustment is sufficient to pull the concentration deviation ratio back from the second deviation range to the first deviation range, while avoiding excessive adjustment that could cause system oscillations. When the concentration deviation ratio is in the second deviation range and the concentration component is higher than the preset target concentration value, the water flow rate is increased by a single adjustment step. When the concentration deviation ratio is in the second deviation range and the concentration component is lower than the preset target concentration value, the water flow rate is decreased by a single adjustment step. When only the flow velocity deviation is in the second deviation range and the concentration deviation ratio is in the first deviation range, the water flow rate is increased by a single adjustment step to reduce the slurry viscosity and restore the flow velocity towards the preset flow velocity threshold. When the condition is determined to be a level three state, the third control response strategy is triggered. The third control response strategy generates a control command to immediately stop the current filling operation, closes the valves of the dry material feeder and the water supply pipeline, and simultaneously starts the pipeline flushing program, turns on the flushing water pump and the pipeline flushing valve, injects flushing water into the filling pipeline and continuously delivers it to the pipeline outlet until all residual slurry in the pipeline is discharged.

[0039] See Figure 5 In the graph, the horizontal axis represents time in seconds, and the vertical axis represents the proportion or deviation, ranging from 0 to 0.35. The orange solid line represents the trend of concentration deviation ratio over time, and the green dashed line represents the trend of flow velocity deviation over time. The red dotted line, blue dotted line, and blue-green dotted line represent the key thresholds for concentration deviation ratio and flow velocity deviation, respectively: the upper limit of the first deviation interval is 0.05, the first deviation threshold is 0.10, the upper limit of the second deviation interval is 0.15, and the second deviation threshold is 0.25.

[0040] As shown in the figure, the concentration deviation ratio remained within the range of 0 to 0.05 for most of the time, occasionally fluctuating above 0.05, reaching around 0.15 in some periods. This indicates that the slurry concentration remained within the first deviation range most of the time, occasionally entering the second deviation range or even approaching the third-level range. The fluctuation range of the flow velocity deviation was larger than that of the concentration deviation ratio, exceeding the first deviation threshold of 0.10 multiple times, and significantly climbing above 0.25 in some periods, indicating that the pipeline flow velocity fluctuated greatly and repeatedly entered the second deviation range and exceeded the third-level state threshold.

[0041] In practical implementation, the feasibility verification steps under the constraints of the multi-constraint collaborative boundary management system are achieved through the following methods. The multi-constraint collaborative boundary management system consists of three sub-boundary systems: concentration constraint boundary, rheological constraint boundary, and pipeline transportation safety boundary. The concentration constraint boundary comprises an upper concentration limit and a lower concentration limit. The upper concentration limit is set as the preset target concentration value plus an absolute amplitude value, and the lower concentration limit is set as the preset target concentration value minus the same absolute amplitude value. The absolute amplitude value is set to 0.03, based on the fact that when the concentration fluctuation of the mine backfill slurry is within ±3 percentage points, the strength and fluidity of the slurry can meet the design requirements of the backfill and the safety requirements of pipeline transportation. Exceeding this range poses a risk of insufficient strength or pipe blockage. The rheological constraint boundary comprises the critical flow velocity value and the critical yield stress value for pipeline transportation. The critical flow velocity value for pipeline transportation is calculated using the Durand formula based on the particle size distribution, density, and pipe diameter of the backfill slurry. When the pipeline flow velocity is lower than the critical flow velocity value, solid particles will settle. The critical yield stress value is obtained through actual measurement and calibration using a rotational viscometer, based on the concentration of the filling slurry and the amount of cementitious material. When the yield stress of the slurry exceeds the critical yield stress value, the pipeline transport resistance will increase sharply. The pipeline transport safety boundary consists of the pipeline pressure resistance limit and the maximum output power of the transport pump. The pipeline pressure resistance limit is taken as 0.85 times the design pressure of the filling pipeline, based on the principle of retaining a 15% safety margin. The maximum output power of the transport pump is the rated power value indicated on the pump's nameplate.

[0042] The adjusted proportioning parameters for the secondary or tertiary states are obtained. These parameters are represented by the target water flow rate setpoint of the electromagnetic flowmeter on the water supply pipeline. The process of mapping the adjusted proportioning parameters to concentration values ​​is as follows: The current solid dry material mass flow rate is obtained. This mass flow rate is collected in real-time by the mass flowmeter at the outlet of the dry material feeder. The adjusted water flow rate setpoint is used as the water mass flow rate, and the slurry concentration is calculated. The process of mapping the adjusted proportioning parameters to pipeline flow velocity values ​​is as follows: The solid dry material mass flow rate is divided by the solid dry material density to obtain the solid dry material volumetric flow rate. The adjusted water flow rate setpoint is divided by the water density to obtain the water volumetric flow rate. The sum of the solid dry material volumetric flow rate and the water volumetric flow rate is used as the total slurry flow rate. The total slurry flow rate is divided by the cross-sectional area of ​​the filling pipeline to obtain the pipeline flow velocity value.

[0043] The process of mapping the adjusted proportioning parameters to the output power value of the delivery pump uses the following formula:

[0044] in, This represents the output power value of the delivery pump. The total pressure drop of the piping system, This represents the total flow rate of the slurry. The efficiency coefficient of the transfer pump. Total pressure drop of the piping system. It consists of two parts: friction loss and local resistance loss. Friction loss is calculated using Bingham's fluid pipeline resistance calculation formula. The required slurry yield stress and plastic viscosity are obtained by interpolating the concentration values ​​through a pre-calibrated concentration-rheological parameter relationship curve. Total slurry flow rate. This is the sum of the volumetric flow rates of the solid dry material and the water. (Efficiency coefficient of the transfer pump) The efficiency value of the pump under rated operating conditions is obtained from the performance curve provided by the pump manufacturer. When the pump's operating point deviates from the rated operating conditions, the value is corrected according to the performance curve.

[0045] After completing the three mappings, the obtained concentration values ​​are compared item by item with the upper and lower limits of the concentration constraint boundary. The pipeline flow velocity value is compared with the critical pipeline flow velocity value of the rheological constraint boundary. The pump output power value is compared with the maximum pump output power value of the pipeline delivery safety boundary. At the same time, the calculated slurry yield stress is compared with the critical yield stress value of the rheological constraint boundary. When the concentration value is greater than or equal to the lower limit and less than or equal to the upper limit, the pipeline flow velocity value is greater than or equal to the critical pipeline flow velocity value, the pump output power value is less than or equal to the maximum pump output power value, and the calculated slurry yield stress is less than or equal to the critical yield stress value, all mapped values ​​are determined to be within the corresponding boundaries, and the verification is marked as passed. The adjusted water flow rate setpoint corresponding to the proportioning parameters is then output to the programmable logic controller (PLC) actuator.

[0046] When the concentration value is less than the lower limit or greater than the upper limit, or the pipeline flow velocity is less than the critical flow velocity, or the pump output power is greater than the maximum output power, or the calculated slurry yield stress is greater than the critical yield stress, the verification is marked as failed. When verification fails, the feedwater flow rate adjustment is obtained for the current adjustment. The feedwater flow rate adjustment is defined as the difference between the adjusted feedwater flow rate setpoint and the original feedwater flow rate. The absolute value of the feedwater flow rate adjustment is halved by a step size. The new adjustment value after halving is equal to half the original feedwater flow rate adjustment value, and its direction is consistent with the original adjustment. The original feedwater flow rate is added to the new adjustment value to obtain the corrected feedwater flow rate setpoint. Using the corrected feedwater flow rate setpoint as the new proportioning parameter, the mapping to concentration, pipeline flow velocity, pump output power, and slurry yield stress is re-executed, and a new round of item-by-item comparison is performed with the concentration constraint boundary, rheological constraint boundary, and pipeline transport safety boundary. The process of halving, remapping, and comparing the water flow rate adjustment is repeated until a revised ratio parameter ensures that the concentration, pipeline flow velocity, pump output power, and slurry yield stress simultaneously meet their respective boundary constraints. In this iteration, the revised ratio parameter is marked as verified and output to the programmable logic controller (PLC) actuator. When the absolute value of the water flow rate adjustment is less than 0.01 kg / s after multiple halvings, it is considered that the water flow rate adjustment has returned to zero. The halving iteration process is stopped, the current ratio parameter adjustment is abandoned, the water flow rate before adjustment is maintained, and this water flow rate is output to the PLC actuator.

[0047] See Figure 6In the graph, the horizontal axis represents the time sequence, indicating consecutive time points of data sampling; the left vertical axis represents the mapped concentration value, ranging from approximately 0.69 to 0.77; and the right vertical axis represents the water supply flow rate adjustment, in kilograms per second, ranging from approximately -0.12 to 0.14. The blue solid line curve in the graph represents the dynamic change of the mapped concentration value, and the red dashed line represents the trend of the water supply flow rate adjustment. The graph also marks the upper and lower limits of the concentration; the upper limit is represented by a gray dotted dashed line with a value of 0.75, and the lower limit is represented by a gray dashed dotted line with a value of 0.69.

[0048] As shown in the figure, the mapped concentration value fluctuated between the upper and lower limits, remaining stable within the range of approximately 0.70 to 0.74. The value did not exceed the set upper limit of 0.75 and lower limit of 0.69, indicating that the current mixing parameters have been verified by the multi-constraint collaborative boundary management system, and the mapped concentration value meets the concentration constraint boundary conditions. At approximately time points 80 and 360, the feedwater flow rate adjustment showed significant positive peak abrupt changes, with maximum adjustments of approximately 0.13 kg / s and 0.10 kg / s, respectively. These corresponded to brief increases in the mapped concentration value, reflecting a single adjustment of the feedwater flow rate in the secondary state to correct the concentration deviation. At approximately time point 200, the feedwater flow rate adjustment abruptly turned negative to approximately -0.12 kg / s, resulting in a brief decrease in the mapped concentration value, reflecting that the system adjusted by reducing the feedwater flow rate based on the concentration deviation.

[0049] In practice, at the end of each shift, a sequence of measured uniaxial compressive strength values ​​for the cast test blocks prepared with the filling slurry is obtained. The preparation process for the cast test blocks with the filling slurry is as follows: During the filling operation, slurry samples are collected from the sampling port of the filling pipeline at fixed time intervals. The collected slurry samples are injected into a cubic mold with a side length of 70.7 mm. After curing under standard curing conditions for 28 days, the test blocks are removed. An axial load is applied to each test block using a pressure testing machine until the test block fails. The maximum axial pressure value at failure is recorded. The maximum axial pressure value is divided by the bearing area of ​​the test block to obtain the measured uniaxial compressive strength value. After pressure testing, a set of measured uniaxial compressive strength values ​​is obtained for all test blocks prepared in this shift. This set of measured uniaxial compressive strength values ​​is arranged in the order of the test block numbers to form a sequence of measured uniaxial compressive strength values.

[0050] Extract the average concentration and average feed rate values ​​during this shift's operation. The average concentration value is obtained by iterating through the concentration components in the real-time rheological state representation vector at all sampling times during the shift, summing the concentration components at all sampling times, and then dividing by the total number of sampling times; the resulting arithmetic mean is the average concentration value. The average feed rate value is obtained by iterating through the solid dry material mass flow rate at all sampling times during the shift, summing the solid dry material mass flow rates at all sampling times, and then dividing by the total number of sampling times; the resulting arithmetic mean is the average feed rate value.

[0051] The extracted average concentration and average feed rate values ​​are matched to a pre-established strength prediction surface. The strength prediction surface is a two-dimensional surface function. The construction process is as follows: Under laboratory conditions, for the same material ratio, filling slurry is prepared and cast into specimens under multiple combinations of concentration and feed rate values. After curing for 28 days, the uniaxial compressive strength under each combination is measured. The concentration value is used as the first independent variable axis, the feed rate value as the second independent variable axis, and the uniaxial compressive strength value as the dependent variable. A continuous surface is obtained by fitting using a bivariate polynomial regression method; this continuous surface is the strength prediction surface. On the strength prediction surface, for any pair of average concentration and average feed rate values, the function value of the strength prediction surface at that coordinate point is obtained. This function value is used as the center value. A preset strength tolerance range is extended upwards and downwards from the center value to form the predicted strength value interval. The preset strength tolerance range is ±10% of the center value.

[0052] Calculate whether each measured uniaxial compressive strength value in the uniaxial compressive strength measurement sequence falls within the predicted strength value range. Count the number of specimens in the uniaxial compressive strength measurement sequence whose values ​​are greater than or equal to the lower boundary of the predicted strength value range and less than or equal to the upper boundary of the predicted strength value range. Divide the counted number of specimens within the range by the total number of specimens in the uniaxial compressive strength measurement sequence; the resulting ratio is the confidence level of strength compliance. The formula for calculating the confidence level of strength compliance is:

[0053] in, To achieve the required strength confidence level, This represents the number of test blocks in the sequence of measured uniaxial compressive strength values ​​that fall within the range of predicted strength values. This represents the total number of test blocks in the sequence of measured uniaxial compressive strength values. The confidence level for strength compliance ranges from 0 to 1. The closer the confidence level is to 1, the higher the consistency between the actual strength and the predicted strength of the filling grout in that shift.

[0054] The shift number, average concentration value, average feed rate value, strength compliance confidence level, and proportioning parameters are linked into a single data record. The proportioning parameters specifically include the solid dry material mass flow rate, the water flow rate setpoint, and the water-cement ratio. This linked data record is then written to the historical database. The historical database can be a relational database or a time-series database, and each data record in the historical database is indexed using the shift number as the primary key.

[0055] Before the next shift starts, obtain the design strength grade of the backfill material and the pipeline transportation distance parameters for the stope to be backfilled. The design strength grade of the backfill material is provided by the mine's mining design department, and is expressed as the minimum uniaxial compressive strength value required for the backfill material at 28 days. The pipeline transportation distance parameter is the total length of the pipeline from the outlet of the mixing equipment at the backfilling station to the farthest end of the backfill pipeline within the stope to be backfilled. This pipeline transportation distance parameter is calculated from the backfill pipeline network layout diagram provided by the mine's surveying department. Using the design strength grade of the backfill material and the pipeline transportation distance parameter as search criteria, perform a query operation in the historical database to filter out all historical data records that have the same design strength grade of the backfill material as the stope to be backfilled and whose pipeline transportation distance parameters deviate from those of the stope to be backfilled within ±5%.

[0056] From all the selected historical data records, the best candidate record is chosen: the one with the highest confidence score for strength compliance. If multiple historical data records have the same highest confidence score for strength compliance, the record with the highest water-cement ratio among these records is selected as the candidate record. The highest water-cement ratio indicates the lowest cement addition. This candidate record is determined as the historically optimal mix design. The average concentration value stored in the historically optimal mix design is extracted as the initial concentration setpoint for the next shift, and the average feed rate value stored in the historically optimal mix design is extracted as the initial solid dry matter mass flow rate setpoint for the next shift. These two setpoints together constitute the initial mix design parameters for the next shift.

[0057] The calculated strength achievement confidence score after the current shift is compared with a preset confidence threshold. The preset confidence threshold is set at 0.85. This threshold is chosen because: a strength achievement confidence score below 0.85 indicates a significant deviation between the actual and predicted strength of the filling slurry prepared under the current mix design, requiring adjustment of the mix design parameters; a strength achievement confidence score greater than or equal to 0.85 indicates acceptable repeatability and reliability of the current mix design. When the strength achievement confidence score is greater than or equal to 0.85, the closed-loop iterative convergence process is not initiated, and the mix design parameters of the current shift are used in the next shift.

[0058] When the confidence level for achieving the strength target is below 0.85, the closed-loop iterative convergence process is initiated. The water-cement ratio (HCR) value in the current shift's mix design parameters is obtained. The HCR is defined as the ratio of the water flow rate to the solid dry material mass flow rate. The HCR is decreased in a fixed step size of 0.02. This fixed step size is chosen because for every 0.02 decrease in the HCR, the corresponding increase in cement addition is approximately 5% to 8%. This increment allows for a noticeable improvement in the strength of the filling slurry while avoiding a significant increase in filling costs due to excessive cement addition. Simultaneously, the cement mass flow rate in the solid dry material mass flow rate is increased proportionally. The increase is calculated using the mass balance relationship based on the fixed HCR step size. The adjusted HCR and adjusted cement mass flow rate are combined to form the adjusted mix design parameters.

[0059] The adjusted proportioning parameters are input into the strength prediction surface. The corresponding concentration and feed rate values ​​are then queried on the strength prediction surface to recalculate the predicted strength values. The recalculated predicted strength values ​​are compared with the strength requirements corresponding to the design strength grade of the filling material in the mine to be filled. When the recalculated predicted strength value is greater than or equal to the strength requirements corresponding to the design strength grade of the filling material, this iteration meets the strength compliance condition. The proportioning parameters that meet the strength compliance condition are submitted to the multi-constraint collaborative boundary management system for feasibility verification. The adjusted proportioning parameters are mapped to concentration values, pipeline flow rates, pump output power, and slurry yield stress. Each value is compared with the concentration constraint boundary, rheological constraint boundary, and pipeline transport safety boundary. When all mapped values ​​are within their corresponding boundaries, the iteration result is output as the proportioning parameters for the next shift, completing one iteration.

[0060] When the recalculated predicted strength value is less than the strength requirement value corresponding to the design strength grade of the filling body, or when the adjusted mix proportion parameters fail the verification by the multi-constraint collaborative boundary management system, the water-cement ratio is reduced again by a fixed step of 0.02 based on the previously adjusted water-cement ratio. Simultaneously, the cement mass flow rate is increased proportionally again. The predicted strength value is recalculated by substituting it into the strength prediction surface, and the multi-constraint collaborative boundary management system verification is re-executed. This process continues until the predicted strength value is greater than or equal to the strength requirement value corresponding to the design strength grade of the filling body after a certain iteration, and the mix proportion parameters pass the verification of all boundaries of the multi-constraint collaborative boundary management system. The mix proportion parameters after this iteration are then used as the mix proportion parameters for the next shift. The above iterative process of decreasing water-cement ratio, increasing cement mass flow rate, calculating predicted strength value, and verifying feasibility is repeated until the confidence level of strength achievement is greater than or equal to 0.85, at which point the closed-loop iterative convergence process stops.

[0061] See Figure 7The graph shows the uniaxial compressive strength (in MPa) on the horizontal axis and the frequency of the corresponding strength value on the vertical axis. The light blue bar chart shows the distribution of the measured uniaxial compressive strength of the slurry casting test blocks from this shift. The bar chart shows that the strength values ​​of the test blocks are mainly concentrated in the range of approximately 6.5 MPa to 10 MPa, exhibiting an approximately normal distribution trend, with the frequency reaching a peak near 8 MPa. The two red dashed lines in the graph represent the lower and upper boundaries of the strength prediction range, respectively. The range corresponds to ±10% of the center value of the predicted strength surface, with the lower boundary approximately 7.3 MPa and the upper boundary approximately 8.8 MPa. As can be seen from the bar chart, less than half of the test blocks fall within the predicted strength range, and the strength of most test blocks in the bar chart is lower than the lower boundary of the predicted range, indicating a significant deviation between the actual and predicted strength of the slurry used in this shift. The strength compliance confidence level marked in the upper left corner of the figure is 0.48, which is lower than the preset strength compliance confidence level threshold of 0.85. This indicates that the strength compliance rate of the slurry prepared in this shift is low, and the mixing scheme has not yet met the expected strength consistency requirements. The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application.

Claims

1. A method for dynamically controlling the concentration of filling material in metal mines, characterized in that, include: The first type of physical signal output from the pipeline sensor and the second type of process parameter output from the filling station preparation stage are collected to calculate the slurry mass concentration respectively. A consistency test is performed based on the absolute value of the deviation between the two concentration values. If the test fails, a conflict arbitration is performed using signal confidence weighting. The arbitration result is combined with the average concentration value when the test passes to construct a real-time rheological state characterization vector. Based on the relationship between the representation vector and the three-level state threshold, the three-level state attribution judgment is performed. The differentiated control response strategy is triggered according to the attribution result, and the feasibility verification is performed under the constraints of the multi-constraint collaborative boundary management system. After each shift, the confidence level of intensity compliance is calculated based on the actual intensity data and written into the historical database; Before the next shift starts, the database is searched for the historical best mix ratio scheme that matches the current working conditions. If the confidence level does not reach the threshold, a closed-loop iterative convergence is initiated to adjust the mix ratio parameters until the intensity prediction value meets the standard.

2. The method for dynamically controlling the concentration of filling material in a metal mine according to claim 1, characterized in that, The first type of physical signal output from the acquisition pipeline sensor and the second type of process parameter output from the filling station preparation stage are used to calculate the slurry mass concentration, specifically including: The first type of physical signal is composed of ultrasonic attenuation spectrum data and pipeline differential pressure data. The first type of physical signal is imported into a pre-calibrated attenuation-concentration mapping table, and the first concentration estimate is obtained by interpolation calculation. The second type of process parameters consists of the feeding rate and water flow rate of the filling station. The ratio of the mass of solid dry material to the total mass is calculated to obtain the estimated value of the second concentration. The first and second concentration estimates are each labeled with a corresponding timestamp.

3. The method for dynamically controlling the concentration of filling material in a metal mine according to claim 2, characterized in that, The consistency test is performed based on the absolute value of the deviation between the two concentration values. If the test fails, conflict arbitration is performed using signal confidence weighting. The arbitration result is then combined with the mean concentration value when the test passes to construct a real-time rheological state representation vector. Specifically, this includes: Calculate the absolute value of the deviation between the first concentration estimate and the second concentration estimate. If the absolute value of the deviation is less than the preset deviation threshold, the consistency test is passed. Take the arithmetic mean of the two concentration values ​​as the fused concentration value. If the absolute value of the deviation is greater than or equal to the preset deviation threshold, the consistency test fails. The signal-to-noise ratio of the ultrasonic attenuation spectrum data and the fluctuation amplitude of the feed rate sensor are obtained. The first confidence level and the second confidence level are assigned according to the signal-to-noise ratio and the fluctuation amplitude, respectively. The first and second concentration estimates are weighted by the first and second confidence levels, and the weighted average is used as the arbitration concentration value. By combining the fusion concentration value or arbitration concentration value with the current pipeline flow rate value and stirring torque value, a real-time rheological state characterization vector is constructed.

4. The method for dynamically controlling the concentration of filling material in a metal mine according to claim 3, characterized in that, The process of performing a three-level state attribution judgment based on the relationship between the representation vector and the three-level state threshold, and triggering a differentiated control response strategy based on the attribution result, specifically includes: Concentration and velocity components are extracted from the real-time rheological state representation vector. The concentration deviation ratio between the concentration component and the preset target concentration is calculated, and the velocity deviation of the velocity component relative to the preset velocity threshold is also calculated. When the concentration deviation ratio is in the first deviation range and the flow rate deviation does not exceed the first deviation threshold, it is determined to be a first-level state. When the concentration deviation ratio or the flow rate deviation is in the second deviation range, it is determined to be a level two state. When the concentration deviation ratio is in the third deviation range or the flow rate deviation exceeds the third deviation threshold, it is judged to be in a level three state. The first-level state trigger is a response strategy that keeps the current mixing parameters unchanged and continuously monitors them; the second-level state trigger is a response strategy that adjusts the water addition volume with a single adjustment step size; and the third-level state trigger is a response strategy that stops the current filling operation and starts pipeline flushing.

5. The method for dynamically controlling the concentration of filling material in a metal mine according to claim 4, characterized in that, The feasibility verification performed under the constraints of the multi-constraint collaborative boundary management system specifically includes: Obtain multi-constraint boundary conditions consisting of upper concentration limit, lower concentration limit, upper pipeline flow rate limit, and upper stirring power limit; and verify the adjusted proportioning parameters under the secondary or tertiary states by substituting them into each constraint boundary condition one by one. When the proportioning parameters simultaneously meet all constraint boundary conditions, the verification is marked as passed, and the adjusted proportioning parameters are output to the actuator. If the proportioning parameter does not meet any constraint boundary condition, it is marked as a failed verification. The adjustment amount in the proportioning parameter is halved by step size and then verified again. The above halving verification process is repeated until the verification passes or the adjustment amount is returned to zero.

6. The method for dynamically controlling the concentration of filling material in a metal mine according to claim 5, characterized in that, After each shift, the confidence level of intensity compliance is calculated based on the actual intensity data and written into the historical database, specifically including: At the end of each shift, obtain the sequence of measured values ​​of uniaxial compressive strength of the test blocks cast from the filling slurry prepared in that shift; Extract the average concentration value and average feed rate value corresponding to this shift, and match them with the predicted intensity value range under the same concentration value and the same feed rate value in the pre-established intensity prediction surface. Calculate the ratio of the number of test blocks falling into the predicted strength value range in the sequence of measured uniaxial compressive strength values ​​to the total number of test blocks, and use this ratio as the confidence level of strength compliance. The shift number, average concentration value, average feed rate value, strength compliance confidence level, and proportioning parameters are associated and stored in the historical database.

7. The method for dynamically controlling the concentration of filling material in a metal mine according to claim 6, characterized in that, The process of retrieving the historically optimal allocation scheme that matches the current operating conditions from the database before starting the next shift specifically includes: Before the next shift starts, obtain the design strength grade and pipeline transportation distance parameters of the filling body of the stope to be filled. Using the design strength grade and pipeline transportation distance parameters as search criteria, filter out historical records with the same stope conditions in the historical database. The record with the highest confidence level of strength compliance and the lowest cement addition in the mix proportion parameters is selected from the filtered historical records as the historical optimal mix proportion scheme; The concentration and feed rate values ​​from the historical optimal formulation scheme are extracted as the initial formulation parameters for the next shift.

8. The method for dynamically controlling the concentration of filling material in a metal mine according to claim 7, characterized in that, If the confidence level does not reach the threshold, the closed-loop iterative convergence adjustment of the ratio parameters is initiated until the intensity prediction value meets the standard, specifically including: The confidence level of achieving the intensity target calculated after the current shift ends is compared with the preset confidence threshold. When the confidence level of achieving the intensity target is lower than the preset confidence threshold, the closed-loop iterative convergence is initiated. The water-cement ratio in the mix proportion parameters of the current shift is decreased in fixed increments, while the amount of cement added is increased simultaneously. The adjusted mix proportion parameters are then substituted into the strength prediction surface to recalculate the predicted strength value. When the recalculated predicted strength value meets the design strength level requirement and the adjusted mix proportion parameters satisfy the multi-constraint collaborative boundary management system, a single iteration is completed, and the result of the single iteration is used as the mix proportion parameters for the next shift. Repeat the above iterative process until the strength reaches the set confidence level greater than or equal to the preset confidence threshold.

9. The method for dynamic control of filling material concentration in metal mines according to claim 8, characterized in that, The multi-constraint collaborative boundary management system includes concentration constraint boundaries, rheological constraint boundaries, and pipeline transportation safety boundaries. The feasibility verification performed under the constraints of this system specifically involves: The concentration constraint boundary is composed of the absolute range of the target concentration value fluctuation; the rheological constraint boundary is composed of the critical flow velocity value and critical yield stress value of pipeline transportation; and the pipeline transportation safety boundary is composed of the pipeline pressure resistance limit value and the maximum output power value of the delivery pump. The proportioning parameters are sequentially mapped to concentration values, pipeline flow rates, and pump output power values. Each value is then compared with the concentration constraint boundary, rheological constraint boundary, and pipeline transport safety boundary. The verification is successful when all mapped values ​​are within their corresponding boundaries.

10. A dynamic control system for the concentration of backfill material in a metal mine, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the method for dynamic control of filling material concentration in metal mines as described in any one of claims 1 to 9.

Images