A closed loop management system and method for concentrate blending operations

Abstract

The present application relates to the technical field of operation management, in particular to a kind of concentrate ore blending operation closed-loop management system and method, system includes sampling parameter acquisition module generates sampling verification parameter set, depth consistent verification module completes dry base weight boundary determination and binds assay measured value, impurity limit reverse calculation module calculates remaining impurity quota and theoretical configuration amount, and single portion locking module locks planned feed weight, and weight interval checking module generates voucher closed record according to belt scale difference.This application, by dry base weight, theoretical weight interval, impurity quota, planned weight and measured weight are included in the same closed-loop verification chain, make sampling depth and assay data more stable, out-of-limit impurity batch feeding is constrained by contract quota, and only closed voucher is formed after on-site feeding is matched with planned weight, thereby reducing deviation lag exposure, reducing account caliber difference, improving the certainty of ore blending correction basis and operation traceability.

Description

Technical Field

[0001] This invention relates to the field of operation management technology, and in particular to a closed-loop management system and method for concentrate blending operations. Background Technology

[0002] The field of operations management technology involves the entire process management of production operations, from planning, task assignment, resource allocation, process execution, on-site recording, status tracking, quality verification, anomaly handling to result archiving. Its core aspects include the identification and classification of work objects, the division of work batches, the scheduling of personnel and equipment, the requisition and circulation of materials, the constraint of process sequence, the registration of work parameters, the backfilling of on-site sampling and inspection data, the confirmation of work deviations, and the correspondence between work records and production ledgers. This field typically organizes production activities around specific criteria such as work plans, work tickets, weighing records, inspection results, inventory ledgers, and on-site handover records, so that each work link forms a traceable management chain according to the established process.

[0003] The traditional concentrate blending operation closed-loop management system refers to a management system for the concentrate blending process before mineral processing or smelting. It involves planning the blending ratio, executing the blending, sampling and testing, verifying results, and correcting the ratio for concentrates from different sources with varying grades, moisture content, and impurity levels. The technical aspects it addresses include concentrate batch information registration, confirmation of inventory quantities in ore bins or material yards, input of test values ​​for major components such as copper, lead, zinc, iron, and sulfur, recording of moisture and impurity indicators, setting target grades for furnace or beneficiation, calculating the amount of concentrate used per batch, collecting weighing data from belt scales or weighbridges, and using grab buckets or feeders. The traditional method involves arranging the feeding sequence, recording the mixing and stockpiling operations, registering the resampled test results after the ore blending is completed, and readjusting the usage of subsequent batches based on the remaining inventory and test reports when the target grade is deviated from the target grade. The ore blending personnel determine the feeding ratio of each batch according to the target component range and the test values ​​of each concentrate batch. Then, the on-site personnel execute the ore blending according to the bin number, batch number, weight, feeding time, and material taking sequence, and fill in the weighing results, retest results, and abnormal records back into the operation log.

[0004] Traditional management systems rely on a sequential process of notification forms, test reports, inventory ledgers, weighing sheets, and shift records for concentrate blending operations. The blending ratio is determined by personnel based on the target components and batch test values, and then executed on-site according to warehouse number, batch number, weight, and material collection order. There is a lack of continuous verification between sampling depth, sample weight, test values, impurity limits, planned feed weight, and the actual difference between the belt scale and the actual weight. Operational deviations are easily exposed with delays, and document backfilling can easily lead to discrepancies. Subsequent batch corrections rely on experience-based judgment, making it difficult to reliably form a traceable closed-loop documentation. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide a closed-loop management system and method for concentrate blending operations. The technical solution is as follows: On the one hand, a closed-loop management system for concentrate blending operations is provided, the system comprising: The sampling parameter acquisition module acquires the cross-sectional area of ​​the sample groove, the effective sampling length, the wet sample weight, the moisture content and the target depth number, converts the wet sample weight into the dry basis weight, calculates the theoretical weight range corresponding to the target depth number and constructs the sampling verification parameter set. The depth consistency verification module, based on the sampling verification parameter set, calls the support vector machine model to compare the dry basis weight with the theoretical weight range, outputs the boundary judgment result, and when the dry basis weight is within the theoretical weight range, establishes a mapping relationship between the test value and the target depth number, extracts the impurity content measurement value to generate data-bound samples; The impurity limit calculation module calculates the remaining available impurity quota based on the measured impurity content, the contract impurity limit, the impurity safety retention amount, and the impurity quota occupancy value. When the measured impurity content is greater than the contract impurity limit, it calculates the theoretical configuration amount and generates the feeding capping parameters. The batching share locking module compares the candidate batching share with the theoretical configuration amount based on the feeding capping parameters to determine the planned feeding weight, and constructs a batching weight mapping set by combining the batching operation order number; The weight range verification module calculates the difference in feeding weight based on the weight mapping set of the order number, according to the starting cumulative value and the ending cumulative value of the belt scale. When matching the planned feeding weight, it performs range locking on the ore blending operation order number and generates a voucher closure record.

[0006] As a further aspect of the present invention, the sampling parameter acquisition module includes: The dry basis conversion submodule obtains the wet sample weight and moisture content, multiplies the wet sample weight by the moisture content to calculate the moisture content value, and subtracts the moisture content value from the wet sample weight to generate the dry basis weight. The interval calculation submodule obtains the cross-sectional area of ​​the sample groove, the effective sampling length and the target depth number, multiplies the cross-sectional area of ​​the sample groove by the effective sampling length to obtain the volume value, calculates the fluctuation limit of the volume value in combination with the preset allowable error ratio, performs matching and alignment processing between the fluctuation limit and the target depth number to generate the theoretical weight interval. The parameter aggregation submodule calls the dry basis weight, the theoretical weight range, and the target depth number to extract the dry basis weight and the theoretical weight range as evaluation reference values. The evaluation reference values ​​are written into the layer mapping node associated with the target depth number, and the data structure items of the layer mapping node are integrated to construct a sampling verification parameter set.

[0007] As a further aspect of the present invention, the process of matching and aligning the fluctuation limit with the target depth number specifically involves: acquiring the historical calibration records of the sampling equipment and the ore material type index; extracting the wear allowance parameter from the historical calibration records; determining the corresponding loose coefficient based on the ore material type index; multiplying the wear allowance parameter by the loose coefficient to obtain a basic deviation value; adding a preset fixed mechanical tolerance to the basic deviation value to obtain a preset allowable error ratio; multiplying the volume value by the preset allowable error ratio to obtain a volume compensation value; adding the volume value and the volume compensation value to obtain a volume upper limit; subtracting the volume value from the volume compensation value to obtain a volume lower limit; combining the volume upper limit and the volume lower limit to constitute the fluctuation limit; extracting the depth level identifier of the target depth number; establishing a data mapping relationship between the depth level identifier and the fluctuation limit; converting the fluctuation limit with the data mapping relationship into a numerical range format; and outputting the theoretical weight range.

[0008] As a further aspect of the present invention, the depth consistency verification module includes: The boundary calculation submodule parses the sampling verification parameter set, extracts the dry basis weight, theoretical weight range and target depth number, calls the support vector machine model to calculate the numerical deviation between the dry basis weight and the boundary endpoint of the theoretical weight range, compares the numerical deviation with the preset classification hyperplane distance limit value, determines the data point distribution status, and generates the boundary determination result. The assay mapping submodule extracts the target depth number from the boundary determination result, collects the assay values ​​attached to the node attributes, mounts the assay values ​​to the main index path of the target depth number, performs key-value alignment and merging processing, and generates an assay association matrix. The sample extraction submodule parses the underlying element fields of the test correlation matrix, reads the impurity content measurement value of the impurity element identifier, merges and encapsulates the impurity content measurement value with the associated target depth number and writes it into the memory of a preset structure to generate a data-bound sample.

[0009] As a further aspect of the present invention, the process of comparing the numerical deviation with the preset classification hyperplane distance limit specifically involves: extracting historical weight features and historical moisture features to construct a training matrix; mapping the training matrix to a high-dimensional space based on a kernel function, solving for the maximum class margin to generate weight coefficients and bias coefficients, and establishing the support vector machine model using the weight coefficients and bias coefficients; inputting the dry basis weight and the theoretical weight range into the support vector machine model; obtaining the spatial vector after mapping the boundary endpoint values ​​of the dry basis weight and the theoretical weight range to the high-dimensional space, calculating the vertical distance between the spatial vector and the classification hyperplane of the support vector machine model, and using the vertical distance as the numerical deviation; obtaining the weighing error parameter and the ore blending accuracy parameter; multiplying the weighing error parameter and the ore blending accuracy parameter to generate a basic product term, adding the basic product term to the historical discrete variance to generate a boundary benchmark value; configuring the boundary benchmark value as the preset classification hyperplane distance limit value; and comparing the numerical magnitude of the numerical deviation with the preset classification hyperplane distance limit value.

[0010] As a further aspect of the present invention, the impurity limit inverse calculation module includes: The quota calculation submodule parses the data binding sample, extracts the impurity content measurement value, obtains the contract impurity upper limit, the impurity safety retention amount and the impurity quota occupancy value, subtracts the impurity safety retention amount and the impurity quota occupancy value from the contract impurity upper limit, and generates the remaining impurity available quota through arithmetic difference operation; The allocation inverse calculation submodule compares the measured impurity content with the contract impurity upper limit. When the measured impurity content is greater than the contract impurity upper limit, it reads the remaining available impurity quota, obtains the conversion ratio constant, divides the remaining available impurity quota by the conversion ratio constant, performs a quotient operation, and generates the theoretical allocation quantity. The capping construction submodule collects the candidate batch identifiers for the theoretical configuration quantity, combines the theoretical configuration quantity with the candidate batch identifiers to construct a restriction mapping dictionary, writes the restriction mapping dictionary into the business configuration table node, and generates the capping parameters for material feeding.

[0011] As a further aspect of the present invention, the process of obtaining the conversion ratio constant specifically involves subtracting the contract impurity upper limit from the measured impurity content value to obtain the impurity overflow concentration value; inputting the impurity overflow concentration value into a preset classification matrix for matching, and extracting the overflow amplification coefficient associated with the impurity overflow concentration value; and multiplying the proportion base by the overflow amplification coefficient to set the conversion ratio constant. The process of dividing the remaining available amount of impurities by the conversion ratio constant to perform a quotient operation specifically involves dividing the remaining available amount of impurities by the conversion ratio constant to generate an initial configuration limit; when the initial configuration limit is greater than the lower limit of material input, the initial configuration limit is set as the theoretical configuration amount.

[0012] As a further aspect of the present invention, the allocation share locking module includes: The share parsing submodule parses the feeding capping parameters, extracts the theoretical configuration amount, obtains candidate feeding shares, performs memory alignment between the candidate feeding shares and the theoretical configuration amount, and combines the candidate feeding shares and the theoretical configuration amount to generate a feeding comparison parameter group. The weight setting submodule calls the feeding comparison parameter group, reads the candidate feeding share and the theoretical configuration amount, compares the candidate feeding share and the theoretical configuration amount, and when the candidate feeding share is greater than the theoretical configuration amount, sets the theoretical configuration amount as the target feeding value and generates the planned feeding weight. The order number mapping submodule collects the ore blending operation order number for the planned material feeding weight, establishes a mapping relationship between the ore blending operation order number and the planned material feeding weight, uses the ore blending operation order number as the primary key identifier and the planned material feeding weight as the associated value to perform data binding, and constructs an order number weight mapping set.

[0013] As a further aspect of the present invention, the weight range verification module includes: The mapping parsing submodule parses the order number weight mapping set, obtains the ore blending operation order number and planned feeding weight, collects the belt scale start cumulative value and belt scale end cumulative value, merges the belt scale start cumulative value and belt scale end cumulative value, and generates a cumulative value data packet. The difference calculation submodule calls the cumulative value data package to read the starting cumulative value and the ending cumulative value of the belt scale, subtracts the starting cumulative value from the ending cumulative value of the belt scale, performs an arithmetic difference operation to obtain the numerical difference, and generates the feeding weight difference. The voucher locking submodule performs numerical matching based on the difference in material input weight and the planned material input weight. When the difference in material input weight is equal to the planned material input weight, it extracts the ore blending operation order number, performs interval locking on the ore blending operation order number, changes the associated attribute to the locked state, and generates a voucher closure record by combining the locked state with the ore blending operation order number.

[0014] On the other hand, a closed-loop management method for concentrate blending operations, wherein the closed-loop management method for concentrate blending operations is executed based on the aforementioned closed-loop management system for concentrate blending operations, includes the following steps: S1: Obtain the cross-sectional area of ​​the sample groove, effective sampling length, wet sample weight, moisture content and target depth number, convert the wet sample weight to dry basis weight, calculate the theoretical weight range corresponding to the target depth number and construct the sampling verification parameter set; S2: Based on the sampling verification parameter set, call the support vector machine model to compare the dry basis weight with the theoretical weight range, output the boundary judgment result, and when the dry basis weight is within the theoretical weight range, establish the mapping relationship between the test value and the target depth number, extract the impurity content measurement value to generate data binding samples; S3: Calculate the remaining available impurity quota based on the measured impurity content, the contract impurity limit, the impurity safety retention amount, and the impurity quota occupancy value, and calculate the theoretical configuration amount when the measured impurity content is greater than the contract impurity limit to generate feeding capping parameters; S4: Based on the feeding capping parameters, compare the candidate feeding share with the theoretical configuration amount to determine the planned feeding weight, and construct a weight mapping set by combining the ore blending operation order number; S5: Based on the weight mapping set of the order number, calculate the difference in feeding weight according to the starting cumulative value and the ending cumulative value of the belt scale. When matching the planned feeding weight, perform range locking on the ore blending operation order number and generate a voucher closure record.

[0015] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: This invention converts the cross-sectional area of ​​the sample groove, effective sampling length, moisture content, and wet sample weight into a dry basis weight and theoretical weight range at the target depth. This makes the accuracy of the sampling depth an entry constraint before binding the test values. The test impurity values ​​are then linked with the contract impurity limit, the safety retention amount, and the occupied quota to calculate the remaining available quota. Batches exceeding the limit are compressed into the theoretical configuration amount and the planned feeding weight is locked. Only after the cumulative difference of the on-site belt scale is closed with the planned weight is a voucher record formed. This creates a continuous verification chain between sampling, testing, quota, feeding, and vouchers, reducing the delayed exposure of deviations and the difference in document caliber, and improving the certainty and traceability of the basis for subsequent batch corrections. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of a closed-loop management system for concentrate blending operations provided in an embodiment of the present invention; Figure 2 This is a schematic diagram of the system framework of the present invention; Figure 3 This is a flowchart of the sampling parameter acquisition module in this invention; Figure 4 This is a flowchart of the depth consistency verification module in this invention; Figure 5 This is a flowchart of the impurity limit back calculation module in this invention; Figure 6 This is a flowchart of the order allocation share locking module in this invention; Figure 7 This is a flowchart of the weight range verification module in this invention; Figure 8 This is a flowchart of a closed-loop management method for concentrate blending operations provided in an embodiment of the present invention. Detailed Implementation

[0018] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0019] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0020] like Figure 1-2 As shown, this embodiment of the invention provides a closed-loop management system for concentrate blending operations. The system includes a sampling parameter acquisition module, a depth consistency verification module, an impurity limit back calculation module, a blending order share locking module, and a weight range verification module. The sampling parameter acquisition module acquires the cross-sectional area of ​​the sample groove, the effective sampling length, the wet sample weight, the moisture content, and the target depth number. It converts the wet sample weight into a dry basis weight based on the moisture content, calculates the theoretical weight range corresponding to the target depth number based on the cross-sectional area of ​​the sample groove and the effective sampling length, and integrates the target depth number, the dry basis weight, and the theoretical weight range to construct a sampling verification parameter set. The depth consistency verification module parses the sampling verification parameter set to extract the dry basis weight, theoretical weight range and target depth number, calls the support vector machine model to compare the dry basis weight and theoretical weight range and outputs the boundary judgment result, obtains the test value, and when the boundary judgment result determines that the dry basis weight is within the theoretical weight range, it constructs the mapping relationship between the test value and the target depth number, extracts the impurity content measurement value in the test value and constructs the data binding sample; The impurity limit reverse calculation module parses the data binding sample to extract the impurity content measurement value, obtains the contract impurity upper limit, impurity safety retention amount and impurity quota occupied value, subtracts the impurity safety retention amount and impurity quota occupied value from the contract impurity upper limit to calculate the remaining impurity available quota, and calculates the theoretical configuration amount based on the remaining impurity available quota when the impurity content measurement value is greater than the contract impurity upper limit, and generates the feeding capping parameters. The batching share locking module parses the feeding capping parameters to extract the theoretical configuration amount, obtains the candidate feeding share, and sets the theoretical configuration amount as the planned feeding weight when the candidate feeding share is greater than the theoretical configuration amount. It also obtains the ore blending operation order number and combines the ore blending operation order number with the planned feeding weight to construct an order number weight mapping set. The weight range verification module parses the order number weight mapping set to obtain the ore blending operation order number and the planned feeding weight, collects the starting cumulative value and ending cumulative value of the belt scale, calculates the difference between the ending cumulative value and the starting cumulative value of the belt scale to obtain the feeding weight difference, and performs a range locking operation for the ore blending operation order number when the feeding weight difference matches the planned feeding weight, generating a voucher closure record.

[0021] Specifically, such as Figure 2 , 3 As shown, the sampling parameter acquisition module includes: The dry basis conversion submodule obtains the wet sample weight and moisture content, multiplies the wet sample weight by the moisture content to calculate the moisture content value, and subtracts the moisture content value from the wet sample weight to generate the dry basis weight. The dry basis conversion submodule activates the data receiving channel of the underlying system bus and periodically scans the data register of the weighing sensor mounted at the bottom of the belt conveyor via an industrial programmable logic controller (PLC) to acquire the real-time raw analog electrical signal of the weight under high-frequency sampling. The dry basis conversion submodule performs pre-processing data cleaning on this raw analog electrical signal. Specifically, it uses a sliding time window mechanism to extract 20 consecutive data sampling points, performs median filtering on the data sampling points within the window to remove sudden interference peak data frames caused by mechanical vibration, and then converts it to a double-precision floating-point digital format using built-in analog-to-digital conversion logic to generate accurate wet sample weight parameters. The dry basis conversion submodule sends a data reading command to the microwave moisture analyzer installed on the side wall of the feeding hopper via a serial communication protocol to obtain the raw moisture percentage sequence data obtained by the device based on microwave attenuation measurement. The dry basis conversion submodule removes null values ​​and illegal out-of-bounds characters from the raw moisture percentage sequence data, performs an arithmetic mean calculation, and generates a stable moisture content parameter. The dry basis conversion submodule allocates an independent numerical computation sandbox space in memory, calls the built-in floating-point multiplication logic component, uses the wet sample weight parameter obtained after cleaning and conversion as the multiplier and the moisture content parameter as the multiplicand, performs a product operation, and generates a moisture content parameter to characterize the absolute mass of moisture. Next, the dry basis conversion submodule calls the floating-point subtraction logic component, sets the wet sample weight parameter as the minuend, sets the newly generated moisture content parameter as the subtrahend, performs an arithmetic difference operation, and derives the final dry basis weight parameter that reflects the pure physical mass of the material. For example, when the dry basis conversion submodule obtains a wet sample weight parameter of 1550.5 kg through the aforementioned acquisition and analysis operations, and the moisture content parameter obtained by the microwave moisture analyzer is 12.5%, the dry basis conversion submodule first calls the multiplication operation logic component to multiply 1550.5 kg and 12.5%, resulting in a moisture content parameter of 193.8125 kg. Then, the dry basis conversion submodule calls the subtraction operation logic component to subtract 193.8125 kg from 1550.5 kg, deriving a dry basis weight parameter of 1356.6875 kg. The execution node of this operation logic is that the dry basis conversion submodule calculates the dry basis weight parameter result based on the wet sample weight parameter and the moisture content parameter, thereby triggering the subsequent data integration operation of the parameter aggregation submodule.

[0022] The interval calculation submodule obtains the cross-sectional area of ​​the sample groove, the effective sampling length and the target depth number. It multiplies the cross-sectional area of ​​the sample groove by the effective sampling length to obtain the volume value. It calculates the fluctuation limit of the volume value in combination with the preset allowable error ratio. It performs matching and alignment processing between the fluctuation limit and the target depth number to generate the theoretical weight interval. The process of matching and aligning the fluctuation limit with the target depth number is as follows: First, obtain the historical calibration records of the sampling equipment and the ore material type index. Extract the wear allowance parameter from the historical calibration records and determine the corresponding loosening coefficient based on the ore material type index. Multiply the wear allowance parameter by the loosening coefficient to obtain the basic deviation value. Add the preset fixed mechanical tolerance to the basic deviation value to obtain the preset allowable error ratio. Multiply the volume value by the preset allowable error ratio to obtain the volume compensation value. Add the volume value and the volume compensation value to obtain the upper volume limit. Subtract the volume value from the volume compensation value to obtain the lower volume limit. Combine the upper and lower volume limits to form the fluctuation limit. Extract the depth level identifier of the target depth number, establish a data mapping relationship between the depth level identifier and the fluctuation limit, convert the fluctuation limit with the data mapping relationship into a numerical range format, and output the theoretical weight range. The interval calculation submodule reads the sampling hardware specification data stored in the local configuration file, obtains the sample slot cross-sectional area parameter and the effective sampling length parameter, and receives the target depth number parameter issued by the central scheduling platform. The interval calculation submodule calls the built-in multiplication logic register to multiply the sample slot cross-sectional area parameter and the effective sampling length parameter to derive the volume value parameter. Subsequently, the interval calculation submodule accesses the equipment ledger relational database through structured query language to retrieve the historical calibration record of the corresponding sampling equipment and the current ore material type index. The interval calculation submodule extracts the wear allowance parameter obtained through laser ranging comparison from the historical calibration record, and matches the corresponding loosening coefficient parameter in the material attribute configuration table using the ore material type index as a foreign key. The interval calculation submodule sends the wear allowance parameter and the loosening coefficient parameter into the arithmetic logic unit for product operation to generate the basic deviation value parameter. Next, the interval calculation submodule reads the system's factory-preset fixed mechanical tolerance parameter, performs an addition operation on the fixed mechanical tolerance parameter and the basic deviation value parameter, and derives the preset allowable error ratio parameter. The interval calculation submodule multiplies the volume numerical parameter with a preset allowable error ratio parameter, outputting the volume compensation numerical parameter. The submodule then calls the addition and subtraction interfaces to add the volume numerical parameter and the volume compensation numerical parameter to derive the upper limit volume parameter, and subtracts them to derive the lower limit volume parameter. These upper and lower limit parameters are combined to form the fluctuation limit parameter. Finally, the submodule extracts the depth level identifier parameter (containing leading zeros) from the target depth number parameter using string truncation logic. It constructs a data mapping relationship between the depth level identifier parameter and the fluctuation limit parameter in an in-memory hash table, serializes the fluctuation limit parameter with this mapping relationship into a closed interval format, and outputs the theoretical weight interval parameter.

[0023] Table 1 Historical Sampling Feature Calibration Parameters

[0024] Table 1 lists the calibration data for different material types. For example, when the sample groove cross-sectional area parameter is 0.05 square meters and the effective sampling length parameter is 2.4 m, the interval calculation submodule calculates the volume parameter as 0.12 cubic meters. If the read wear allowance parameter is 0.02 and the looseness coefficient parameter is 1.15, the interval calculation submodule calculates the basic deviation parameter as 0.023, adds the fixed mechanical tolerance parameter of 0.05, and obtains the preset allowable error ratio parameter as 0.073. Multiplying the volume parameter of 0.12 cubic meters by 0.073 yields the volume compensation parameter of 0.00876 cubic meters. The upper limit of the volume parameter is 0.12876 cubic meters, and the lower limit of the volume parameter is 0.11124 cubic meters. The execution node of this calculation logic is that the interval calculation submodule calculates the theoretical weight interval result based on the hardware specifications and deviation characteristic parameters, thereby triggering the subsequent operation of the spatial boundary verification of the parameter aggregation submodule.

[0025] The parameter aggregation submodule calls the dry basis weight, theoretical weight range and target depth number, extracts the dry basis weight and theoretical weight range as evaluation reference values, writes the evaluation reference values ​​into the layer mapping node associated with the target depth number, integrates the data structure items of the layer mapping node, and constructs the sampling verification parameter set. The parameter aggregation submodule opens a memory listening port and calls the dry basis weight parameter, theoretical weight range parameter, and target depth number parameter through a message queue subscription mechanism. Using data serialization conversion rules, the submodule removes the data type tags and verification timestamps attached to these parameters, extracting the dry basis weight parameter and theoretical weight range parameter in their pure numerical state as evaluation reference value parameters. The submodule performs path traversal in the document node tree of the underlying non-relational database, locating the layer mapping node associated with the target depth number parameter as the primary key. By executing a data write command, the submodule overwrites the extracted evaluation reference value parameters into the child node attribute segments of this layer mapping node. Subsequently, the submodule scans all data structure items within this layer mapping node, performing a memory pointer merging operation on the discrete depth coordinate parameters, time series parameters, and the newly written evaluation reference value parameters to construct a sampling verification parameter set with a strict hierarchical relationship. For example, when the input dry basis weight parameter is 1356.6875 kg, the theoretical weight range parameter is 0.11124 to 0.12876 cubic meters, and the target depth number parameter is D-2026-003, the parameter aggregation submodule extracts 1356.6875 kg and 0.11124 to 0.12876 cubic meters as evaluation reference values, persists them to the layer mapping node corresponding to D-2026-003, and aggregates them to generate a sampling verification parameter set with a size of 128 bytes. The execution node of this operation logic is that the parameter aggregation submodule performs data aggregation operation based on discrete verification indicators to obtain the sampling verification parameter set result, which in turn triggers the subsequent data set parsing operation of the boundary calculation submodule.

[0026] Specifically, such as Figure 2 , 4 As shown, the depth consistency verification module includes: The boundary calculation submodule parses the sampling verification parameter set, extracts the dry basis weight, theoretical weight range and target depth number, calls the support vector machine model to calculate the numerical deviation between the dry basis weight and the boundary endpoint of the theoretical weight range, compares the numerical deviation with the preset classification hyperplane distance limit value, determines the data point distribution status, and generates the boundary determination result. The process of comparing the numerical deviation with the preset classification hyperplane distance limit is as follows: First, historical weight and moisture features are extracted to construct a training matrix. Then, the training matrix is ​​mapped to a high-dimensional space based on a kernel function. The maximum class margin is calculated to generate weight and bias coefficients. A support vector machine (SVM) model is established using these coefficients. The dry-basis weight and theoretical weight range are input into the SVM model. The boundary endpoints of the dry-basis weight and theoretical weight range are mapped to a spatial vector in the high-dimensional space. The vertical distance between this spatial vector and the classification hyperplane of the SVM model is calculated, and this vertical distance is used as the numerical deviation. Weighing error parameters and ore blending accuracy parameters are obtained. The weighing error parameters and ore blending accuracy parameters are multiplied to generate a basic product term. This basic product term is added to the historical discrete variance to generate a boundary benchmark value. This boundary benchmark value is configured as the preset classification hyperplane distance limit value. Finally, the numerical deviation is compared with the preset classification hyperplane distance limit value. The boundary calculation submodule calls the data parsing function to deserialize and decompose the sampled verification parameter set, extracting the dry basis weight parameters, theoretical weight range parameters, and target depth number parameters. When constructing the support vector machine model, the boundary calculation submodule extracts calibrated historical weight and moisture characteristic parameters from the archived records of the historical database for the first 30 natural days, concatenating these two types of parameters into a two-dimensional training matrix data structure containing 2000 rows of data. The boundary calculation submodule calls the Gaussian radial basis kernel function operation component to perform spatial dimension upscaling on the two-dimensional coordinates of this two-dimensional training matrix data structure. The Gaussian radial basis kernel function operation component first performs a difference operation on the data point coordinates and calculates the squared Euclidean distance, then divides the squared Euclidean distance by the squared value of the Gaussian kernel width selected through grid search and multiplies it by a constant of -1, and finally performs a natural exponentiation operation to generate kernel matrix data in a high-dimensional feature space. Using kernel matrix data, the boundary calculation submodule calls the sequence minimization optimization operator to solve the structural risk minimization objective function. With the penalty factor hyperparameter set to 2.5, iteratively updates the weight coefficient sequence and bias coefficient parameters, formally establishing the support vector machine (SVM) model. The boundary calculation submodule inputs the endpoint values ​​of the dry basis weight parameters and the theoretical weight range parameters into the SVM model. The SVM model also calls the Gaussian radial basis kernel function to map these parameters to a high-dimensional feature space, generating a spatial vector. It then calls the analytical geometric spatial distance calculation logic to derive the vertical distance parameter from this spatial vector to the SVM model's classification hyperplane, directly setting this vertical distance parameter as the numerical deviation parameter. The boundary calculation submodule retrieves the weighing error parameters and ore blending accuracy parameters from the system environment library, performs a multiplication operation to generate the basic product term parameter, then reads the historical discrete variance parameter, adds the basic product term parameter to the historical discrete variance parameter to generate the boundary benchmark numerical parameter, and directly assigns this boundary benchmark numerical parameter as the preset classification hyperplane distance boundary value parameter. The boundary calculation submodule uses comparison logic to determine the relationship between the numerical deviation parameter and the preset classification hyperplane distance limit value parameter. If the numerical deviation parameter is less than the limit value, the data point distribution status parameter is determined to be within the normal confidence region, and the corresponding boundary determination result parameter is generated. For example, if the weighing error parameter is set to 0.05 and the ore blending accuracy parameter is set to 0.95, multiplying them yields a basic product term parameter of 0.0475. The historical variance parameter is set to 0.12, and adding them together yields a limit benchmark parameter of 0.1675, thus determining the preset classification hyperplane distance limit value parameter to be 0.1675. If the vertical distance calculated by the support vector machine model, i.e., the numerical deviation parameter, is 0.145, since 0.145 is less than 0.1675, the boundary calculation submodule determines it to be a normal distribution and generates the result.The execution node of this operation logic is that the boundary calculation submodule derives the boundary judgment result parameters based on multi-dimensional input features, which in turn triggers the subsequent operation of the test data association in the test mapping submodule.

[0027] The test mapping submodule extracts the target depth number from the boundary judgment result, collects the test values ​​attached to the node attributes, mounts the test values ​​to the main index path of the target depth number, performs key-value alignment and merging processing, and generates a test association matrix. The assay mapping submodule receives the incoming boundary judgment result parameters. After confirming that the parameters are within the normal confidence region, it extracts the corresponding target depth number parameter from the request context. The assay mapping submodule connects to the database interface of the laboratory information management system via a remote procedure call protocol, initiates batch query commands based on the target depth number parameters, and collects the assay measurement parameters attached to the device node attributes. The assay mapping submodule constructs a main index path string for the target depth number parameters in local memory, and uses a dictionary data structure to mount the obtained assay measurement parameters as key-value pairs to the memory address of this main index path string. The assay mapping submodule iterates through all key-value pairs under this memory address, performs memory alignment and data merging processing, eliminates duplicate key names caused by multiple assays, and generates a assay association matrix parameter in a unified format. For example, when the extracted target depth number parameter is D-2026-003, the test measurement parameters obtained by the test mapping submodule from the external system include an iron content of 62.5% and a sulfur content of 0.03%. The test mapping submodule loads and merges these indicators in key-value pairs to generate a test correlation matrix parameter containing a two-dimensional array structure. The execution node of this operation logic is that the test mapping submodule merges data based on the depth index to obtain the test correlation matrix result, which in turn triggers the element stripping subsequent operation of the sample extraction submodule.

[0028] The sample extraction submodule parses the underlying element fields of the assay correlation matrix, reads the impurity content measurement value of the impurity element identifier, merges and encapsulates the impurity content measurement value with the associated target depth number and writes it into the memory of the preset structure to generate data-bound samples. The sample extraction submodule invokes the matrix parser to perform a bidirectional row and column traversal scan of the assay correlation matrix parameters, extracting the underlying element field parameters hidden within the underlying network communication data packet payload. The submodule uses regular expression matching logic to scan the underlying element field parameters, searching for impurity element identifier parameters that conform to the preset impurity naming rules, and reads the specific impurity content measurement parameters at the corresponding column offset address. The submodule allocates a preset structure memory space in system memory, performs byte-level concatenation of the read impurity content measurement parameters with the target depth number parameter carried in the context, and packages them into a standard binary data block using a structure encapsulation function, thereby generating highly portable data-bound sample parameters. For example, when parsing the matrix, the submodule identifies the impurity element identifier parameter as sulfur, with a corresponding impurity content measurement parameter of 200.0 kg, and the associated target depth number parameter as D-2026-003. The submodule encapsulates and concatenates these two parameters within a structure, generating a 64-byte data-bound sample parameter. The execution node of this operation logic is that the sample extraction submodule obtains the data binding sample parameter result based on the impurity identifier matching operation, which in turn triggers the subsequent operation of the ore blending limit verification of the quota calculation submodule.

[0029] Specifically, such as Figure 2 , 5 As shown, the impurity limit inverse calculation module includes: The quota calculation submodule parses the data binding sample, extracts the impurity content measurement value, obtains the contract impurity upper limit, impurity safety retention amount and impurity quota occupied value, subtracts the impurity safety retention amount and impurity quota occupied value from the contract impurity upper limit, and generates the remaining impurity available quota through arithmetic difference calculation; The process of obtaining the impurity safety retention amount and impurity quota occupancy value is as follows: read the impurity fluctuation variance and risk penalty multiplier in the preset sample data set, multiply the impurity fluctuation variance and risk penalty multiplier to set the impurity safety retention amount; multiply the total dry basis weight of the feed with the average impurity content measurement value to determine the impurity quota occupancy value. The quota calculation submodule calls the data packet unpacking function to parse the data-bound sample parameters, accurately extracting the impurity content measurement parameters in floating-point state. The quota calculation submodule accesses the business contract database table to obtain the contract impurity upper limit parameter effective for the day, and retrieves the impurity fluctuation variance parameter and risk penalty multiplier parameter from the parameter configuration center. The quota calculation submodule calls the multiplication operation logic to multiply the impurity fluctuation variance parameter and the risk penalty multiplier parameter, setting the resulting value as the impurity safety retention amount parameter. Simultaneously, the quota calculation submodule collects the total dry basis weight parameter of the current batch, reads the average impurity content measurement parameter from the accumulated record, multiplies the total dry basis weight parameter and the average impurity content measurement parameter to determine the impurity quota occupancy value parameter. The quota calculation submodule calls the continuous subtraction operation register, using the contract impurity upper limit parameter as the initial minuend, first subtracting the calculated impurity safety retention amount parameter, then subtracting the impurity quota occupancy value parameter from the difference, and deriving the remaining available impurity quota parameter through continuous arithmetic subtraction. For example, if the contractual impurity limit parameter is 500.0 kg, the impurity fluctuation variance parameter is 2.0 kg, and the risk penalty multiplier parameter is 2.5, multiplying these two parameters yields an impurity safety retention parameter of 5.0 kg. If the total dry basis weight parameter of the feed is 10000.0 kg, and the average impurity content measurement parameter is 0.02, multiplying these parameters yields an impurity quota occupancy parameter of 200.0 kg. The quota calculation submodule subtracts 5.0 kg from 500.0 kg to obtain 495.0 kg, and then subtracts 200.0 kg from 495.0 kg to obtain the final remaining available impurity quota parameter of 295.0 kg. The execution node of this calculation logic is that the quota calculation submodule calculates the remaining available impurity quota parameter based on the contract threshold and consumption history, thereby triggering the subsequent operation of the ore weight conversion in the allocation back-calculation submodule.

[0030] The allocation inverse calculation submodule compares the measured impurity content with the contract impurity limit. When the measured impurity content is greater than the contract impurity limit, it reads the remaining available impurity quota, obtains the conversion ratio constant, divides the remaining available impurity quota by the conversion ratio constant, performs a quotient operation, and generates the theoretical allocation quantity. The specific process of obtaining the conversion ratio constant is as follows: subtract the contract impurity upper limit from the measured impurity content to obtain the impurity overflow concentration value; input the impurity overflow concentration value into a preset classification matrix to perform matching, and extract the overflow amplification coefficient associated with the impurity overflow concentration value; multiply the ratio base by the overflow amplification coefficient to set the conversion ratio constant; The process of dividing the remaining available impurities by the conversion ratio constant to perform the quotient operation is as follows: the remaining available impurities are divided by the conversion ratio constant to generate the initial configuration limit; when the initial configuration limit is greater than the lower limit of material input, the initial configuration limit is set to the theoretical configuration amount. The dosage inversion submodule calls the built-in numerical comparator component to compare the read impurity content measurement parameter with the contract impurity upper limit parameter. When the comparator outputs a logic high-level signal indicating that the impurity content measurement parameter is greater than the contract impurity upper limit parameter, the dosage inversion submodule triggers an exception handling branch and reads the remaining impurity available quota parameter generated by the pre-calculation. The dosage inversion submodule calls the subtraction logic to subtract the contract impurity upper limit parameter from the impurity content measurement parameter to obtain the impurity overflow concentration value parameter. The dosage inversion submodule uses this impurity overflow concentration value parameter as an index key to input a preset classification matrix in memory, and uses a binary search algorithm to match the corresponding concentration limit interval, thereby extracting the overflow amplification coefficient parameter strongly correlated with the impurity overflow concentration value parameter. The dosage inversion submodule obtains the preset ratio base parameter from the global configuration environment, performs a product operation on the ratio base parameter and the overflow amplification coefficient parameter, and obtains the conversion ratio constant parameter. The quantity inversion submodule calls the floating-point division unit, using the remaining available impurity allowance parameter as the dividend and the conversion ratio constant parameter as the divisor to perform a quotient operation, generating the initial configuration limit parameter. The quantity inversion submodule compares the initial configuration limit parameter with the material input lower limit parameter specified by the production safety baseline. If the initial configuration limit parameter is greater than the material input lower limit parameter, the initial configuration limit parameter is directly set and output as the theoretical configuration quantity parameter.

[0031] Table 2. Association Table of Impurity Overflow Classification Matrix

[0032] As shown in Table 2, the conversion index for the corresponding state can be obtained through matrix lookup. For example, when the input impurity content measurement parameter is 550.0 kg and the contract impurity upper limit parameter is 500.0 kg, since 550.0 kg is greater than 500.0 kg, the subtraction is performed to obtain the impurity overflow concentration value parameter of 50.0 kg. The dosage inverse calculation submodule matches 50.0 kg in the medium conversion range within the matrix and extracts the overflow amplification factor parameter of 1.2. If the base ratio parameter is 2.0, the two are multiplied to obtain the conversion ratio constant parameter of 2.4. The dosage inverse calculation submodule extracts the remaining impurity available quota parameter of 295.0 kg calculated in the pre-calculation, divides it by 2.4 to obtain the initial configuration limit parameter of 122.9166 kg. The lower limit value parameter for feeding is set to 100.0 kg. Since 122.9166 kg is greater than 100.0 kg, the dosage inverse calculation submodule determines the theoretical configuration quantity parameter to be 122.9166 kg. The execution node of this operation logic is that the allocation inverse calculation submodule performs dynamic division based on the overflow penalty mechanism to obtain the theoretical allocation result, which in turn triggers the subsequent operation of the safety dictionary generation of the capping construction submodule.

[0033] The capping construction submodule collects the identifiers of the batches to be fed into the theoretical configuration quantity, combines the theoretical configuration quantity with the identifiers of the batches to be fed into the theoretical configuration quantity to construct a restriction mapping dictionary, writes the restriction mapping dictionary into the business configuration table node, and generates the feeding capping parameters. The capping construction submodule, based on the theoretical configuration quantity parameters output during the quantity calculation stage, activates the communication interface of the local job scheduling system and collects the identifier parameters of the candidate feeding batches in a standby state through polling. The capping construction submodule calls a dictionary generator library function, using the candidate feeding batch identifier parameters as the core key name and the theoretical configuration quantity parameters as the core key value, to construct a limit mapping dictionary parameter for limiting the material ratio ceiling using key-value pair association. The capping construction submodule calls the write transaction of the distributed database, performs encryption and obfuscation operations on the limit mapping dictionary parameter using an advanced encryption standard algorithm, and then writes it into the business configuration table node of the core database architecture. After the write transaction is successfully committed, the effective feeding capping parameters are generated in the system's global broadcast space. For example, if the collected candidate feeding batch identifier parameter is BATCH-2026-001 and the theoretical configuration quantity parameter is 122.9166kg, the capping construction submodule combines them into a limit mapping dictionary parameter, encrypts and writes it into the business configuration table node, and then broadcasts the feeding capping parameters representing the maximum allowable feeding limit for that batch. The execution node of this operation logic is that the capping construction submodule constructs the material capping parameter result based on the quota limit instruction, which in turn triggers the memory alignment of the share parsing submodule to prepare for subsequent operations.

[0034] Specifically, such as Figure 2 , 6 As shown, the allocation share locking module includes: The share parsing submodule parses the feeding capping parameters, extracts the theoretical configuration quantity, obtains candidate feeding shares, performs memory alignment between the candidate feeding shares and the theoretical configuration quantity, combines the candidate feeding shares and the theoretical configuration quantity, and generates a feeding comparison parameter set. After the share parsing submodule detects a broadcast event for the material feeding cap parameter via the message bus, it initiates its built-in parameter decoding logic to decrypt and parse the material feeding cap parameter, accurately extracting the theoretical configuration quantity parameter that is currently active. Simultaneously, the share parsing submodule sends a data retrieval request to the front-end operator terminal or automated scheduling system to obtain the preliminary planned candidate material feeding share parameters in the current production plan. To ensure efficient memory processing in a 64-bit operating system environment, the share parsing submodule calls a memory padding operation, filling the data storage blocks of the candidate material feeding share parameters and the theoretical configuration quantity parameters with blank bytes, ensuring strict memory alignment of their starting addresses. After alignment, the share parsing submodule encapsulates and combines the memory pointers of the candidate material feeding share parameters and the theoretical configuration quantity parameters into a contiguous array structure, generating a material feeding comparison parameter set for subsequent decision-making comparison. For example, the extracted theoretical configuration quantity parameter is 122.9166 kg, and the candidate material feeding share parameter obtained from the production scheduling system is 150.0 kg. The share parsing submodule aligns these two floating-point values ​​in memory and packages them into a pointer array as the material feeding comparison parameter group. The execution node of this operation logic is that the share parsing submodule performs a memory parsing and reorganization operation based on the planned production scheduling data to obtain the material feeding comparison parameter group result, which in turn triggers the final threshold decision and subsequent operations of the weight setting submodule.

[0035] The weight setting submodule calls the feeding comparison parameter group, reads the candidate feeding share and the theoretical configuration amount, compares the candidate feeding share and the theoretical configuration amount, and when the candidate feeding share is greater than the theoretical configuration amount, sets the theoretical configuration amount as the target feeding value and generates the planned feeding weight. The weight setting submodule calls the material comparison parameter group residing in the memory area via a pointer address, and extracts the candidate material share parameters and theoretical configuration quantity parameters one by one using the built-in parameter reading interface. The weight setting submodule calls the hardware-level logic comparator to perform a strict numerical comparison operation on the candidate material share parameters and theoretical configuration quantity parameters. When the hardware-level logic comparator returns a status bit indicating that the candidate material share parameter is greater than the theoretical configuration quantity parameter, the weight setting submodule activates the forced overwrite mechanism, ignores the original candidate material share parameter, and directly sets the theoretical configuration quantity parameter as the target material value parameter. Then, based on this target material value parameter, the planned material weight parameter officially issued to the underlying equipment is generated. For example, the weight setting submodule reads a candidate feed share parameter of 150.0 kg and a theoretical configuration quantity parameter of 122.9166 kg. Comparing these, it determines that 150.0 kg is greater than 122.9166 kg, triggering a safety cutoff mechanism. This forces the target feed value parameter to be reduced and set to 122.9166 kg, resulting in an output planned feed weight parameter of 122.9166 kg. The execution node of this calculation logic is that the weight setting submodule performs a cutoff judgment operation based on the safety upper limit threshold to obtain the planned feed weight result, which in turn triggers the task flow binding of the order number mapping submodule for subsequent operations.

[0036] The order number mapping submodule collects the ore blending operation order number for the planned material feeding weight, establishes a mapping relationship between the ore blending operation order number and the planned material feeding weight, uses the ore blending operation order number as the primary key identifier and the planned material feeding weight as the associated value to perform data binding, and constructs an order number weight mapping set. Upon receiving the officially issued planned feed weight parameters, the order number mapping submodule invokes its internally integrated snowflake algorithm generator to generate a globally unique ore blending operation order number parameter based on the current timestamp, machine number, and auto-incrementing sequence. Within the document index partition of the relational database system, the order number mapping submodule establishes a foreign key mapping relationship between the ore blending operation order number parameter and the input planned feed weight parameter using Structured Query Language. During the mapping process, the order number mapping submodule mandates that the ore blending operation order number parameter be declared as a primary key identifier parameter, while simultaneously attaching the planned feed weight parameter as a non-nullable associated numerical parameter to it. By committing to the database and executing a binding transaction operation, a complete order number weight mapping set is constructed in persistent storage. For example, for the input planned feed weight parameter 122.9166kg, the order number mapping submodule generates a unique ore blending operation order number parameter ORE-2026-999, declares it as a primary key in the database, and attaches it as a foreign key to 122.9166kg. The constructed order number weight mapping set is then successfully written to the physical disk. The execution node of this operation logic is that the order number mapping submodule generates a document mounting operation based on the unique identifier to obtain the order number weight mapping set result, which in turn triggers the production execution tracking subsequent operation of the mapping parsing submodule.

[0037] Specifically, such as Figure 2 , 7 As shown, the weight range verification module includes: The mapping and parsing submodule parses the order number weight mapping set, obtains the ore blending operation order number and planned feeding weight, collects the belt scale's initial cumulative value and belt scale's final cumulative value, merges the belt scale's initial cumulative value and belt scale's final cumulative value, and generates a cumulative value data package. The mapping and parsing submodule retrieves and parses the written order number weight mapping set through the database connection pool, and obtains the associated ore blending operation order number parameters and planned feeding weight parameters through reverse addressing logic. During actual ore blending production, the mapping and parsing submodule connects in real-time to the on-site belt scale control box via the industrial open platform communication architecture protocol. It collects the first set of weight accumulation pulse data at the moment the belt scale starts feeding, converting it into the belt scale's initial cumulative value parameter. After the feeding stop command is triggered, it collects the final weight accumulation pulse data at the moment the belt scale stops running, converting it into the belt scale's final cumulative value parameter. The mapping and parsing submodule performs a merging and stacking operation on the ore blending operation order number parameter, the belt scale's initial cumulative value parameter, and the belt scale's final cumulative value parameter, according to a fixed byte length, generating a cumulative value data packet parameter for data verification. For example, the parsed ore blending operation order number is ORE-2026-999, the planned feed weight is 122.9166 kg, the collected initial cumulative value of the belt scale is 15000.0 kg, and the collected final cumulative value of the belt scale at the end of feeding is 15122.9166 kg. The mapping and parsing submodule pushes and merges the above data onto the stack, generating a cumulative value data packet parameter of size 32 bytes. The execution node of this operation logic is that the mapping and parsing submodule captures the cumulative data on site based on the communication bus, performs the push and merge to obtain the cumulative value data packet result, and then triggers the closed-loop quality verification subsequent operation of the difference calculation submodule.

[0038] The difference calculation submodule calls the cumulative value data package to read the starting cumulative value and the ending cumulative value of the belt scale, subtracts the starting cumulative value from the ending cumulative value of the belt scale, performs an arithmetic difference operation to obtain the numerical difference, and generates the feeding weight difference. The difference calculation submodule receives the cumulative value data packet parameters transmitted by the message middleware, calls the unpacking and disassembly interface, and reads the starting and ending cumulative value parameters of the belt scale from the preset offset memory address. The difference calculation submodule then calls the built-in high-precision floating-point subtraction core, sets the ending cumulative value parameter of the belt scale to the minuend state, and sets the starting cumulative value parameter to the subtrahend state, performs an arithmetic difference operation mechanism, and thus obtains a numerical difference parameter that truly reflects the net material input within the current operation interval. This numerical difference parameter is then directly encapsulated and converted into a material input weight difference parameter.

[0039] Table 3 Comparison of Control Parameters for Feeding Process

[0040] As shown in Table 3, the flow parameters for difference calculation are fully displayed through a table structure. For example, if the initial cumulative value parameter of the belt scale is 15000.0 kg and the final cumulative value parameter is 15122.9166 kg, the difference calculation submodule subtracts 15000.0 kg from 15122.9166 kg to obtain a difference of 122.9166 kg. The final generated material feeding weight difference parameter is 122.9166 kg. The execution node of this calculation logic is that the difference calculation submodule performs a subtraction operation based on the first and last count values ​​to obtain the material feeding weight difference result, which then triggers the accuracy verification and closed-loop archiving subsequent operations of the voucher locking submodule.

[0041] The voucher locking submodule performs numerical matching based on the difference in feeding weight and the planned feeding weight. When the difference in feeding weight is equal to the planned feeding weight, the ore blending operation order number is extracted. An interval lock is performed on the ore blending operation order number, and the associated attribute is changed to the locked state. The voucher closing record is generated by combining the locked state with the ore blending operation order number. The voucher locking submodule calls the equivalence matching logic interface to compare the measured material weight difference parameter with the originally issued planned material weight parameter. When the equivalence matching logic interface determines that the material weight difference parameter is completely equal to the planned material weight parameter, i.e., meeting the ideal delivery state without deviation, the voucher locking submodule extracts the corresponding ore blending operation order number parameter from the context and performs a memory-level range locking operation on that ore blending operation order number parameter in the system's global scheduling lock manager, changing its associated attribute from the executing state to the locked state. During the generation of the voucher closure record, the voucher locking submodule synchronously triggers process control signals, instructing the on-site execution agency to stop its intervention actions on this batch of materials. The voucher locking submodule combines the final changed lock status identifier parameter with the unique ore blending operation order number parameter, encrypts and encapsulates it using a digital signature hash algorithm, and generates an immutable voucher closure record parameter stored in the historical audit ledger. For example, if the difference in feed weight parameter 122.9166 kg is completely identical to the planned feed weight parameter 122.9166 kg, the voucher locking submodule extracts the order number ORE-2026-999, sets its identifier to a locked state in the lock manager, and triggers the process control logic to terminate the on-site feed action. Finally, it packages this value with the status and encrypts it using a hash algorithm to generate the voucher closure record parameters. The execution node of this operation logic is that the voucher locking submodule triggers a memory locking mechanism based on consistency verification to obtain the voucher closure record result, thereby completing the system closed-loop management task for the entire lifecycle of this ore blending operation.

[0042] Please see Figure 8 The closed-loop management method for concentrate blending operations is based on the aforementioned closed-loop management system for concentrate blending operations and includes the following steps: S1: Obtain the cross-sectional area of ​​the sample groove, effective sampling length, wet sample weight, moisture content and target depth number, convert the wet sample weight to dry basis weight, calculate the theoretical weight range corresponding to the target depth number and construct the sampling verification parameter set; S2: Based on the sampling verification parameter set, call the support vector machine model to compare the dry basis weight with the theoretical weight range, output the boundary judgment result, and when the dry basis weight is within the theoretical weight range, establish the mapping relationship between the test value and the target depth number, extract the impurity content measurement value to generate data binding samples; S3: Calculate the remaining available impurity quota based on the measured impurity content, the contract impurity limit, the impurity safety retention amount, and the impurity quota occupancy value. Calculate the theoretical configuration amount when the measured impurity content is greater than the contract impurity limit and generate the feeding capping parameters. S4: Based on the feeding capping parameters, compare the candidate feeding share with the theoretical configuration amount to determine the planned feeding weight, and construct a weight mapping set for each batching operation number in conjunction with the batching operation number; S5: Based on the single-number weight mapping set, calculate the difference in feeding weight according to the starting cumulative value and the ending cumulative value of the belt scale. When matching the planned feeding weight, perform range locking on the ore blending operation single number and generate a voucher closed record.

[0043] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A closed-loop management system for concentrate blending operations, characterized in that, The system includes: The sampling parameter acquisition module acquires the cross-sectional area of ​​the sample groove, the effective sampling length, the wet sample weight, the moisture content and the target depth number, converts the wet sample weight into the dry basis weight, calculates the theoretical weight range corresponding to the target depth number and constructs the sampling verification parameter set. The depth consistency verification module, based on the sampling verification parameter set, calls the support vector machine model to compare the dry basis weight with the theoretical weight range, outputs the boundary judgment result, and when the dry basis weight is within the theoretical weight range, establishes a mapping relationship between the test value and the target depth number, extracts the impurity content measurement value to generate data-bound samples; The depth consistency verification module includes: The boundary calculation submodule parses the sampling verification parameter set, extracts the dry basis weight, theoretical weight range and target depth number, calls the support vector machine model to calculate the numerical deviation between the dry basis weight and the boundary endpoint of the theoretical weight range, compares the numerical deviation with the preset classification hyperplane distance limit value, determines the data point distribution status, and generates the boundary determination result. The process of comparing the numerical deviation with the preset classification hyperplane distance limit specifically involves: extracting historical weight features and historical moisture features to construct a training matrix; mapping the training matrix to a high-dimensional space based on a kernel function; solving for the maximum class margin to generate weight coefficients and bias coefficients; establishing the support vector machine model using the weight coefficients and bias coefficients; inputting the dry-basis weight and the theoretical weight range into the support vector machine model; obtaining the spatial vectors after mapping the boundary endpoints of the dry-basis weight and the theoretical weight range to the high-dimensional space; calculating the vertical distance between the spatial vectors and the classification hyperplane of the support vector machine model; using the vertical distance as the numerical deviation; obtaining the weighing error parameter and the ore blending accuracy parameter; multiplying the weighing error parameter and the ore blending accuracy parameter to generate a basic product term; adding the basic product term to the historical discrete variance to generate a boundary benchmark value; configuring the boundary benchmark value as the preset classification hyperplane distance limit value; and comparing the numerical deviation with the preset classification hyperplane distance limit value. The impurity limit calculation module calculates the remaining available impurity quota based on the measured impurity content, the contract impurity limit, the impurity safety retention amount, and the impurity quota occupancy value. When the measured impurity content is greater than the contract impurity limit, it calculates the theoretical configuration amount and generates the feeding capping parameters. The batching share locking module compares the candidate batching share with the theoretical configuration amount based on the feeding capping parameters to determine the planned feeding weight, and constructs a batching weight mapping set by combining the batching operation order number; The weight range verification module calculates the difference in feeding weight based on the weight mapping set of the order number, according to the starting cumulative value and the ending cumulative value of the belt scale. When matching the planned feeding weight, it performs range locking on the ore blending operation order number and generates a voucher closure record.

2. The closed-loop management system for concentrate blending operations according to claim 1, characterized in that, The sampling parameter acquisition module includes: The dry basis conversion submodule obtains the wet sample weight and moisture content, multiplies the wet sample weight by the moisture content to calculate the moisture content value, and subtracts the moisture content value from the wet sample weight to generate the dry basis weight. The interval calculation submodule obtains the cross-sectional area of ​​the sample groove, the effective sampling length and the target depth number, multiplies the cross-sectional area of ​​the sample groove by the effective sampling length to obtain the volume value, calculates the fluctuation limit of the volume value in combination with the preset allowable error ratio, performs matching and alignment processing between the fluctuation limit and the target depth number to generate the theoretical weight interval. The parameter aggregation submodule calls the dry basis weight, the theoretical weight range, and the target depth number to extract the dry basis weight and the theoretical weight range as evaluation reference values. The evaluation reference values ​​are written into the layer mapping node associated with the target depth number, and the data structure items of the layer mapping node are integrated to construct a sampling verification parameter set.

3. The closed-loop management system for concentrate blending operations according to claim 2, characterized in that, The process of matching and aligning the fluctuation limit with the target depth number specifically involves: acquiring the historical calibration records of the sampling equipment and the ore material type index; extracting the wear allowance parameter from the historical calibration records; determining the corresponding loosening coefficient based on the ore material type index; multiplying the wear allowance parameter by the loosening coefficient to obtain the basic deviation value; adding the preset fixed mechanical tolerance to the basic deviation value; and setting the preset allowable error ratio. Multiply the volume value by the preset allowable error ratio to obtain the volume compensation value, add the volume value and the volume compensation value to obtain the upper limit of the volume, subtract the volume value from the compensation value to obtain the lower limit of the volume, and combine the upper limit of the volume and the lower limit of the volume to form the fluctuation limit; Extract the depth level identifier of the target depth number, establish a data mapping relationship between the depth level identifier and the fluctuation limit, convert the fluctuation limit with the data mapping relationship into a numerical range format, and output the theoretical weight range.

4. The closed-loop management system for concentrate blending operations according to claim 1, characterized in that, The depth consistency verification module also includes: The assay mapping submodule extracts the target depth number from the boundary determination result, collects the assay values ​​attached to the node attributes, mounts the assay values ​​to the main index path of the target depth number, performs key-value alignment and merging processing, and generates an assay association matrix. The sample extraction submodule parses the underlying element fields of the test correlation matrix, reads the impurity content measurement value of the impurity element identifier, merges and encapsulates the impurity content measurement value with the associated target depth number and writes it into the memory of a preset structure to generate a data-bound sample.

5. The closed-loop management system for concentrate blending operations according to claim 1, characterized in that, The impurity limit inverse calculation module includes: The quota calculation submodule parses the data binding sample, extracts the impurity content measurement value, obtains the contract impurity upper limit, the impurity safety retention amount and the impurity quota occupancy value, subtracts the impurity safety retention amount and the impurity quota occupancy value from the contract impurity upper limit, and generates the remaining impurity available quota through arithmetic difference operation; The allocation inverse calculation submodule compares the measured impurity content with the contract impurity upper limit. When the measured impurity content is greater than the contract impurity upper limit, it reads the remaining available impurity quota, obtains the conversion ratio constant, divides the remaining available impurity quota by the conversion ratio constant, performs a quotient operation, and generates the theoretical allocation quantity. The capping construction submodule collects the candidate batch identifiers for the theoretical configuration quantity, combines the theoretical configuration quantity with the candidate batch identifiers to construct a restriction mapping dictionary, writes the restriction mapping dictionary into the business configuration table node, and generates the capping parameters for material feeding.

6. The closed-loop management system for concentrate blending operations according to claim 5, characterized in that, The process of obtaining the conversion ratio constant is as follows: subtract the contract impurity upper limit from the impurity content measurement value to obtain the impurity overflow concentration value; input the impurity overflow concentration value into a preset classification matrix to perform matching, and extract the overflow amplification coefficient associated with the impurity overflow concentration value; multiply the ratio base by the overflow amplification coefficient to set the conversion ratio constant; The process of dividing the remaining available amount of impurities by the conversion ratio constant to perform a quotient operation specifically involves dividing the remaining available amount of impurities by the conversion ratio constant to generate an initial configuration limit; when the initial configuration limit is greater than the lower limit of material input, the initial configuration limit is set as the theoretical configuration amount.

7. The closed-loop management system for concentrate blending operations according to claim 1, characterized in that, The allocation quota locking module includes: The share parsing submodule parses the feeding capping parameters, extracts the theoretical configuration amount, obtains candidate feeding shares, performs memory alignment between the candidate feeding shares and the theoretical configuration amount, and combines the candidate feeding shares and the theoretical configuration amount to generate a feeding comparison parameter group. The weight setting submodule calls the feeding comparison parameter group, reads the candidate feeding share and the theoretical configuration amount, compares the candidate feeding share and the theoretical configuration amount, and when the candidate feeding share is greater than the theoretical configuration amount, sets the theoretical configuration amount as the target feeding value and generates the planned feeding weight. The order number mapping submodule collects the ore blending operation order number for the planned material feeding weight, establishes a mapping relationship between the ore blending operation order number and the planned material feeding weight, uses the ore blending operation order number as the primary key identifier and the planned material feeding weight as the associated value to perform data binding, and constructs an order number weight mapping set.

8. The closed-loop management system for concentrate blending operations according to claim 1, characterized in that, The weight range verification module includes: The mapping parsing submodule parses the order number weight mapping set, obtains the ore blending operation order number and planned feeding weight, collects the belt scale start cumulative value and belt scale end cumulative value, merges the belt scale start cumulative value and belt scale end cumulative value, and generates a cumulative value data packet. The difference calculation submodule calls the cumulative value data package to read the starting cumulative value and the ending cumulative value of the belt scale, subtracts the starting cumulative value from the ending cumulative value of the belt scale, performs an arithmetic difference operation to obtain the numerical difference, and generates the feeding weight difference. The voucher locking submodule performs numerical matching based on the difference in material input weight and the planned material input weight. When the difference in material input weight is equal to the planned material input weight, it extracts the ore blending operation order number, performs interval locking on the ore blending operation order number, changes the associated attribute to the locked state, and generates a voucher closure record by combining the locked state with the ore blending operation order number.

9. A closed-loop management method for concentrate blending operations, characterized in that, The closed-loop management system for concentrate blending operations according to any one of claims 1-8 includes the following steps: S1: Obtain the cross-sectional area of ​​the sample groove, effective sampling length, wet sample weight, moisture content and target depth number, convert the wet sample weight to dry basis weight, calculate the theoretical weight range corresponding to the target depth number and construct the sampling verification parameter set; S2: Based on the sampling verification parameter set, call the support vector machine model to compare the dry basis weight with the theoretical weight range, output the boundary judgment result, and when the dry basis weight is within the theoretical weight range, establish the mapping relationship between the test value and the target depth number, extract the impurity content measurement value to generate data binding samples; S3: Calculate the remaining available impurity quota based on the measured impurity content, the contract impurity limit, the impurity safety retention amount, and the impurity quota occupancy value, and calculate the theoretical configuration amount when the measured impurity content is greater than the contract impurity limit to generate feeding capping parameters; S4: Based on the feeding capping parameters, compare the candidate feeding share with the theoretical configuration amount to determine the planned feeding weight, and construct a weight mapping set by combining the ore blending operation order number; S5: Based on the weight mapping set of the order number, calculate the difference in feeding weight according to the starting cumulative value and the ending cumulative value of the belt scale. When matching the planned feeding weight, perform range locking on the ore blending operation order number and generate a voucher closure record.

Images