A data management method and system for resource utilization of mine waste

Abstract

The present application relates to the technical field of data management of mine waste resource utilization, and discloses a data management method and system for mine waste resource utilization, wherein the method comprises the following steps: identifying steady-state sections and jump points based on Euclidean distance sequences, and adaptively dividing material window sequences of each mining area; calculating a spatiotemporal weighted composition fingerprint, performing secret sharing processing on a composition weighted vector, transmitting random shares to a resource utilization data coordination node; performing safe complementary pairing determination, safe mixed proportion verification, time sequence alignment mapping, material conservation verification and desensitization abnormality attribution analysis in the ciphertext domain; and finally encapsulating the whole-process data into a data management report in the form of a hash pointer string.

Description

Technical Field

[0001] This invention relates to the field of data management technology for the resource utilization of mine waste, and more specifically, to a data management method and system for the resource utilization of mine waste. Background Technology

[0002] Resource utilization of mine waste is an important way to reduce the stockpile of solid waste and improve the comprehensive utilization rate of mineral resources. In the scenario of joint resource utilization in multi-ownership mining areas, waste from different mining companies is mixed and processed in the same continuous flow processing facility, involving the collaborative management of data from multiple parties.

[0003] Existing data management methods for waste resource utilization typically rely on centralized platforms to aggregate composition data from various mining areas and record continuous flow materials using batch segmentation at fixed time intervals or with fixed weights.

[0004] The aforementioned existing technologies have the following drawbacks: First, the composition data of waste from each mining area can be used to infer the ore grade and mining and beneficiation process efficiency, which is commercially sensitive information. Mining companies refuse to submit the plain text of the composition to a third-party centralized platform, resulting in the inability to centrally manage cross-mining area data, and there is a fundamental conflict between the data sovereignty of each party and the need for collaborative management. Second, in continuous flow processing, materials undergo a gradient transition, and the fixed segmentation method results in excessive heterogeneity of composition within the data recording window. Furthermore, the physical diffusion and mixing lag effects of adjacent material segments make it impossible to accurately establish the correspondence between input and output data, making it difficult to guarantee the accuracy of material conservation verification and responsibility attribution accounting. Summary of the Invention

[0005] This invention provides a data management method and system for the resource utilization of mine waste, which solves the technical problems in related technologies such as the difficulty in balancing the protection of privacy of component data in various mining areas with multi-party collaborative verification and the lack of credible evidence for material conservation verification.

[0006] This invention discloses a data management method for the resource utilization of mine waste, comprising:

[0007] Each mining area acquires local continuous component data streams, identifies steady-state segments and transition points based on the Euclidean distance sequence of component vectors at sampling points, and adaptively delineates material window sequences;

[0008] For each material window, attenuation weights are assigned according to the time distance of the sampling point from the center of the window to calculate the component weight vector. Secret sharing processing is performed to generate random shares and transmit them to the data coordination node.

[0009] Based on the secret share, the satisfaction or gap status of each material window component in the encrypted domain is determined, candidate complementary pairings are generated, and the safe mixing ratio verification is performed on the candidate complementary pairings.

[0010] Based on the time delay offset, the input windows of each party are aligned with the output windows of the production line to establish an encrypted mapping relationship;

[0011] In the encrypted domain, a set of multi-party material conservation equations is constructed, and a safe comparison is performed between the theoretical mixed output value and the actual output components to output a conservation verification status flag.

[0012] The conservation verification status flag and the calculation results of each party's contribution are encapsulated into data management records, and a full-process data management report is generated by concatenating them according to hash pointers.

[0013] Furthermore, the identification of steady-state segments and transition points based on the Euclidean distance sequence of the component vectors of the sampling points includes: comparing the component vector of the current sampling point with its preceding first to second... The Euclidean distance is calculated for the component vectors of each sampling point, forming a length of... The distance value sequence, where The preset backtracking window length; when continuous When the Euclidean distance of all sampling points is lower than the steady-state threshold, the segment is marked as a steady-state segment. The preset number of continuous judgment points; when the Euclidean distance exceeds the mutation threshold, it is marked as a jump point; the jump point is used as the dividing position of adjacent material windows, and continuous steady-state segments are assigned to the same material window; when the number of sampling points between two adjacent jump points is less than the minimum window length threshold, the segment is merged into the previous material window and marked as containing a transition segment.

[0014] Furthermore, the allocation of attenuation weights based on the time distance between the sampling point and the center of the window includes: assuming the time of the material window center is... The sampling point time is The weight of the sampling point Calculate using the following formula:

[0015] ;

[0016] in, For the sampling point time, At the center of the material window, The Gaussian attenuation bandwidth parameter is taken as a preset proportion of the material window time span; the weights of each sampling point are normalized and used to calculate the weighted mean and weighted standard deviation; the weighted mean is truncated and sorted according to its value and then concatenated, with the weighted standard deviation feature value added, and a hash operation is performed to generate the spatiotemporal weighted component fingerprint.

[0017] Furthermore, the secret sharing process employs an additive secret sharing scheme: let the component weighting vector be... The number of participants is ,generate random vectors The dimension of each random vector and the component weighting vector The dimensions are the same, calculate the th Each share is ,in For the first Each share vector For the summation index variable of shares, each share satisfies Each mining area retains its own share and distributes the remaining shares to the corresponding party or data coordination node.

[0018] Furthermore, the generation of candidate complementary pairs also includes: calculating a complementary matching degree score for the Boolean flag vector of each material window. :

[0019] ;

[0020] in, The complementarity and matching degree between the two mining areas is scored. The total number of component dimensions. For component dimension index variables, and The first Boolean vectors of the two mining areas are respectively the second Boolean vectors of the two mining areas. One portion, This is an indicator function; it filters material window pairs with a complementary match degree below the minimum pairing threshold, and only performs candidate complementary labeling on material window pairs with a complementary match degree exceeding the minimum pairing threshold.

[0021] Furthermore, the verification of the secure mixing ratio for candidate complementary pairings includes: each party using its local component weighting vector as a private input, and determining whether a mixing ratio exists through a confusion circuit. Make all All dimensions are satisfied:

[0022] ;

[0023] in, and The weighted vectors of the two components are respectively the first and second parts. One portion, The first vector required for the target component One portion, For the first Tolerance for each component dimension The component dimension index variable is used; the Boolean determination result of the effective mixing ratio and the boundary value of the feasible ratio range after desensitization are output. The desensitization process only outputs the quantization level of the boundary value of the feasible ratio range.

[0024] Furthermore, the step of aligning the input windows of each party with the output windows of the production line based on the time delay offset includes: obtaining the material window timestamps and mixing confluence point location parameters of each mining area's confirmed material feeding. Let the mining area... The length of the pipeline from the mixing point to the output detection point is The average flow velocity of the material is Then the time delay offset ,in The output window sequence is generated using the same adaptive material window partitioning method as the input window for the component data stream at the output end of the production line. The timestamps of each input window are added with the corresponding time delay offsets and then matched with the time range of the output window. When an input window crosses two adjacent output windows after a time delay offset, the share data of the input window is split and mapped to the two output windows according to the time overlap ratio.

[0025] Furthermore, the construction of a multi-party material conservation equation system in the encrypted domain includes: assuming the number of mining areas participating in a certain output window is... mining area The confirmed feed amount is mining area The component weighted vector of the first The elements are Then the first Theoretical mixed output value of each element satisfy:

[0026] ;

[0027] in, The number of mining areas participating in this output window. For the mining area index variable, For the mining area Confirmed feed amount The component dimension index variable is used; for segments with overlapping windows, a mixing hysteresis effect correction coefficient is applied and the segment is re-evaluated. The correction coefficient is determined based on the ratio of the mixing chamber volume of the equipment to the material throughput per unit time.

[0028] Furthermore, it also includes anonymized anomaly attribution analysis: for output windows whose conservation verification status is marked as abnormal, each party calculates the difference between the current input window component share and the component share of the material window that has passed the conservation verification in the encrypted domain, and collaboratively executes a security comparison protocol to determine whether the modulus of the difference exceeds the component variability threshold; the anomaly attribution pointer is output in the form of a combination of mining area identifier and time window segment identifier, and does not contain plaintext data of component absolute value or difference; each data management record in the full-process data management report is appended with the digital signature timestamp of each participant, and each party's signature only covers its own relevant share data and Boolean judgment result field.

[0029] This invention discloses a data management system for the resource utilization of mine waste, used to execute the aforementioned data management method for the resource utilization of mine waste, comprising:

[0030] The adaptive window partitioning module is used to acquire continuous component data streams from online component detection equipment in each mining area, identify steady-state segments and transition points based on Euclidean distance sequences, and adaptively delineate material window sequences.

[0031] The component fingerprint generation and secret sharing module is used to calculate the component weighting vector by allocating attenuation weights to each material window according to the time distance of the sampling point from the center of the window, perform secret sharing processing to generate random shares, and transmit the random share data and the material window timestamp to the data coordination node.

[0032] The secure complementary pairing determination module is used to execute a secure comparison protocol in the encrypted domain based on secret share data, determine the satisfaction or gap status of each material window component, and generate candidate complementary pairs.

[0033] The secure mixing ratio verification module is used to perform a two-party secure mixing ratio verification protocol on candidate complementary pairings, and outputs the Boolean determination result of the effective mixing ratio and the boundary value of the desensitized feasible ratio range;

[0034] The timing alignment mapping module is used to align the input window share sequence of each party with the production line output window based on the time delay offset, and establish an encrypted mapping relationship.

[0035] The encrypted domain conservation verification module is used to construct a set of conservation equations for multiple materials in the encrypted domain, perform a safe comparison between the theoretical mixed output value and the actual output component data, and output a conservation verification status flag.

[0036] The desensitization anomaly attribution module is used to perform safe differential calculation on output windows whose conservation verification status flag is abnormal, and output the anomaly attribution direction without exposing the absolute component value;

[0037] The full-process recording and encapsulation module is used to encapsulate the conservation verification status flags and the calculation results of each party's contribution into data management records, and generate a full-process data management report by concatenating them according to hash pointers.

[0038] The present invention has the following beneficial effects:

[0039] Firstly, during the component data acquisition stage, the weighted vectors of components from each mining area are secretly shared. Subsequent complementary pairing determination, mixing ratio verification, material conservation equation solving, and yield deviation comparison are all completed in the encrypted domain. None of the participating parties can restore the plaintext of the components of other parties. Thus, collaborative management of cross-mining area data is achieved without requiring all parties to submit the plaintext of the components, solving the problem of data non-centralized management caused by parties refusing to share data due to commercial sensitivity.

[0040] Secondly, the adaptive material window partitioning method based on Euclidean distance sequence for identifying steady-state segments and transition points ensures that the heterogeneity of components within each material window is constrained by a steady-state threshold, avoiding the problem of drastic changes in components within the window caused by a fixed partitioning method. The spatiotemporal weighted component fingerprint reduces the interference of material window edge sampling points on window features. The time delay offset calculation and time series alignment method establish an accurate correspondence between input and output windows, overcoming the problem of inaccurate correspondence between input and output data caused by physical diffusion and mixing hysteresis effects under continuous flow conditions, enabling material conservation verification and responsibility attribution accounting to be effectively performed under continuous flow conditions. Attached Figure Description

[0041] Figure 1 This is a flowchart of a data management method for the resource utilization of mine waste provided in an embodiment of the present invention;

[0042] Figure 2 This is a schematic diagram of the distribution of the number of sampling points for the adaptive material window in each mining area provided in an embodiment of the present invention;

[0043] Figure 3 This is a schematic diagram comparing the time delay offset of pipelines in various mining areas according to an embodiment of the present invention;

[0044] Figure 4 This is a schematic diagram illustrating the relationship between pipeline length and time delay offset provided in an embodiment of the present invention;

[0045] Figure 5 This is a schematic diagram of the candidate complementary pairing complementary matching degree scoring provided in an embodiment of the present invention;

[0046] Figure 6 This is a schematic diagram of the material input distribution of each participant in WO-01 provided in an embodiment of the present invention;

[0047] Figure 7 This is a schematic diagram illustrating the proportion of the W-01 split mapping of the C mine area to the output window provided in an embodiment of the present invention;

[0048] Figure 8 This is a schematic diagram of the conservation verification status and conservation deviation determination results of each output window provided in the embodiments of the present invention;

[0049] Figure 9This is a schematic diagram of the chain structure of the full-process data management record provided in the embodiment of the present invention. Detailed Implementation

[0050] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, some features described in the examples may be combined in other examples.

[0051] At least one embodiment of the present invention discloses a data management method for the resource utilization of mine waste, see [link to relevant documentation]. Figure 1 This includes the following steps:

[0052] Step 1: Adaptively define the material window sequence based on the composition change characteristics.

[0053] Each mining area's data management system acquires continuous component data streams output by local online component detection equipment at a fixed sampling frequency. For the component vector of each sampling point, its relationship with the previous sample is calculated. The Euclidean distance sequence of component vectors of sampling points, where This is the preset backtracking window length. When continuous... When the Euclidean distance of all sampling points is lower than the steady-state threshold, the segment is marked as a steady-state segment. This represents the preset number of continuous judgment points. When the Euclidean distance exceeds the mutation threshold, it is marked as a compositional jump point. Based on the distribution of steady-state segments and jump points, the start and end boundaries of the waste material window are adaptively defined, generating a material window sequence for each mining area.

[0054] It should be noted that the calculation of the above Euclidean distance sequence refers to the calculation of the component vector of the current sampling point relative to its preceding first to second components. The Euclidean distance is calculated for the component vectors of each sampling point, forming a length of... The distance value sequence. The steady-state threshold and the abrupt change threshold are determined based on the historical fluctuation range of the waste composition in each mining area. The steady-state threshold is taken as a preset multiple of the average Euclidean distance between adjacent sampling points under historical stable operating conditions, and the abrupt change threshold is taken as a preset proportion of the average Euclidean distance between adjacent sampling points during historical batch switching. The rule for defining the start and end boundaries of the material window is: the component jump point is used as the dividing position of adjacent material windows, and continuous steady-state segments are grouped into the same material window.

[0055] In this embodiment of the application, in order to avoid frequent fragmentation of the material window due to short-term disturbances, when the number of sampling points between two adjacent transition points is less than the minimum window length threshold, the segment is merged into the previous material window, and the transition segment is marked in the material window attributes.

[0056] Step 2: Generate spatiotemporal weighted component fingerprints and perform secret sharing processing.

[0057] For each adaptively defined material window, the component vector sequence of all sampling points within the window is extracted. Gaussian decay weights are assigned to each sampling point based on its temporal distance from the center of the material window, and the weighted mean and weighted standard deviation of each element are calculated. The weighted mean is truncated, sorted, and concatenated, with the weighted standard deviation eigenvalue appended. A hash operation is then performed to generate the spatiotemporal weighted component fingerprint for that material window. Simultaneously, the component weighted vectors of each material window are secretly shared locally, splitting them into random shares. Only the random share data and the material window timestamp are transmitted to the resource utilization data coordination node.

[0058] It should be noted that the above-mentioned Gaussian decay weight allocation method is as follows: Let the center time of the material window be... The sampling point time is Then the weight of that sampling point Calculate using the following formula:

[0059] ;

[0060] in, For the sampling point time, At the center of the material window, The time difference between the sampling point and the center of the material window. The Gaussian attenuation bandwidth parameter is a preset proportion of the material window time span. and They have the same dimensions, both being units of time. This represents an exponential function, where the exponent is a dimensionless number. The weights of each sampling point are normalized and used to calculate the weighted mean and weighted standard deviation. The aforementioned truncation and sorting refers to truncating elements in the weighted mean whose values ​​are below the detection limit to zero, and then arranging them in descending order of value to eliminate the influence of differences in the order of element arrangement on the spatiotemporal weighted component fingerprint comparison.

[0061] It should be noted that the above secret sharing process uses an additive secret sharing scheme. Let the component weighting vector be... The number of participants is Then generate random vectors ,in For the first random vector, For the second random vector, For the first Given a set of random vectors, the dimension of each random vector and the weighted vector of its components. The dimensions are the same, calculate the th Each share is ,in For the first Each share vector For the summation index variable of shares, each share satisfies ,in For the first Each mining area retains its own share and distributes the remaining shares to the corresponding party or the resource utilization data coordination node.

[0062] Step 3: Perform secure complementary pairing determination based on secret share.

[0063] At the resource utilization data coordination node, based on the secret share data transmitted from each mining area, a secure comparison protocol is used to calculate the component requirement values ​​for the target resource utilization path. Without restoring the original plaintext components of each mining area, the satisfaction or gap status of each material window in each component dimension is determined. When a material window in a mining area has a gap in a certain dimension, and a material window in another mining area has a satisfaction flag in the same dimension, this pairing is marked as a candidate complementary pair, and a candidate complementary pair encrypted index set is generated.

[0064] Step 3 may specifically include:

[0065] Step 301: The resource utilization data coordination node receives the share data transmitted from each mining area and establishes a time-series index of the share of each mining area according to the material window timestamp.

[0066] Step 302: For the preset requirements of each component dimension of the target resource utilization path, each mining area and resource utilization data coordination node jointly execute the security comparison protocol, and determine in the encrypted domain whether the component weighting vector of each material window meets the requirements in each dimension, and output a Boolean flag indicating whether it meets or falls short.

[0067] Step 303: Traverse all Boolean flag combinations of material window pairs in the mining area, mark material window pairs that present a complementary relationship in the same dimension as candidate complementary pairs, and encapsulate them into the candidate complementary pairing encrypted index set.

[0068] In this embodiment of the application, in order to reduce the computational overhead of invalid pairings, the following steps are also included in step 302: calculate the complementarity matching score of the Boolean flag vector of each material window, filter material window pairs with a complementarity matching score lower than the minimum pairing threshold before step 303, and perform candidate complementarity marking only on material window pairs that exceed the minimum pairing threshold.

[0069] It should be noted that the above-mentioned complementary matching score is calculated as follows: Let the mining area... The Boolean flag vector of a certain material window is mining area The Boolean flag vector of a certain material window is Both have dimensions that are the total number of component dimensions. The Boolean flag 1 indicates satisfaction and 0 indicates a gap. For the first... Each component dimension, when and At that time, or when and At that time, this dimension contributes a complementary count, where For the mining area Boolean flag vector The One portion, For the mining area Boolean flag vector The Each component. Complementary matching score. The sum of the complement counts across all dimensions divided by the total number of component dimensions. :

[0070] ;

[0071] in, For the mining area With the mining area Complementary matching score between them The total number of component dimensions. For component dimension index variables, This is an indicator function; it takes the value 1 if the condition within the parentheses is true, and 0 otherwise. If the material window pair is below the minimum pairing threshold, filter the material window pair.

[0072] Step 4: Perform a safe mixing ratio verification.

[0073] For material window pairs involved in the candidate complementary pairing encrypted index set, a two-party secure mixing ratio verification protocol is executed between the relevant mining areas. Each party uses its local component weighting vector as private input and calculates a system of linear mixing equations through a confusion circuit. The output is a Boolean determination result indicating whether the pair has a valid mixing ratio and the desensitized boundary values ​​of the feasible ratio interval, generating a valid complementary pairing data record.

[0074] It should be noted that the above linear mixed equation system has a total number of component dimensions. When the value is greater than 1, it is an overdetermined system of equations, a single scalar mixed ratio. Must meet simultaneously These equality constraints generally do not have exact solutions. The objective of determining a confused circuit is to determine whether it exists. Make all All dimensions are satisfied:

[0075] ;

[0076] in, and These are the weighted vectors of the two components, respectively. and They are respectively and The Each component represents the mass fraction of the corresponding element, and the dimensions are consistent. Require vectors for target components The One portion, For the first Tolerance for each component dimension This is a dimensionless mixing ratio. This is the index variable for the component dimensions. Each dimension corresponds to... The constraint interval is ,in

[0077] ;

[0078] in, For the first Each component dimension affects the mixing ratio The lower bound of the constraint, For the first Each component dimension affects the mixing ratio The upper bound of the constraint. Feasible. The interval boundaries are determined by taking the intersection of constraints across all dimensions. At that time, the dimensional constraint degenerates into a constraint on Unconstrained (if) ) or no feasible solution (if The obfuscation circuit makes a separate judgment on this situation. Desensitization processing refers to outputting only the quantization level of the boundary value of the feasible proportion range, rather than the exact value.

[0079] Step 5: Establish timing alignment mapping for input and output windows based on time delay offset.

[0080] Based on the material window timestamps of each mining area confirmed in the effective complementary pairing data records and the location parameters of the mixing and confluence points of the continuous flow processing equipment, the time delay offset from each party's material window to the production line output detection point is calculated. The share sequence of each party's material window is time-aligned with the output window adaptively divided at the production line output end according to the time delay offset, establishing an encrypted mapping relationship between the set of multiple input windows and the output window.

[0081] Step 5 may specifically include:

[0082] Step 501: Obtain the timestamp of the material window for each mining area to confirm feeding and the location parameters of the mixing and confluence point. Combine the production line length and material flow rate parameters of the processing equipment to calculate the transmission time delay offset of each material window from the mixing and confluence point to the production line output detection point.

[0083] It should be noted that the above-mentioned time delay offset is calculated as follows: assuming the mining area... The length of the pipeline from the mixing point to the output detection point is The unit is meters, and the average flow velocity of the material in the pipeline is... If the unit is meters per second, then the time delay offset is... The unit is seconds, where This is the index variable for the mining area. When the processing equipment contains multiple processing units, the residence time of each processing unit is accumulated to the total time delay offset.

[0084] Step 502: For the output component data stream collected by the online detection equipment at the production line output end, use the same adaptive material window partitioning method as in Step 1 to generate an output window sequence.

[0085] Step 503: Add the corresponding time delay offset to the timestamp of each party's input window, and match it with the time range of the output window to establish an encrypted mapping relationship between the set of multiple input windows and each output window.

[0086] In this embodiment of the application, in order to handle the partial overlap between the input window and the output window, when the time range of an input window after time delay offset spans two adjacent output windows, the share data of the input window is split and mapped to the two output windows according to the time overlap ratio, and they participate in the subsequent conservation verification calculation respectively.

[0087] Step 6: Perform material conservation verification in the encrypted field.

[0088] For each set of encrypted mapping relationships, a secure multi-party linear equation solving protocol is executed based on the secret share data of each party. Using the material input of each party as known coefficients and the component share of each party as private input, a multi-party material conservation equation system is constructed in the encrypted domain to calculate the secret share of the theoretical mixed output value of each element. The actual output window component data collected by the online detection equipment at the production line output end is acquired. A secure comparison protocol is executed between the actual output window component data and the secret share of the theoretical mixed output value, generating a Boolean judgment result in the encrypted domain indicating whether the yield deviation of each element exceeds the allowable threshold. A correction coefficient based on the mixing hysteresis effect is applied to the overlapping window segments, and a re-judgment is performed, outputting the encrypted conservation verification status flag for each output window.

[0089] Step 6 may specifically include:

[0090] Step 601: Obtain the material input data confirmed by all parties, use the material input as a known coefficient, use the secret share of each party's components as a private input, and have all participating parties and resource utilization data coordination nodes collaboratively execute a secure multi-party linear equation system solution protocol to calculate the secret share of the theoretical mixed output value of each element.

[0091] It should be noted that the above multi-party material conservation equations are constructed as follows: Let the number of mining areas participating in a certain output window be... mining area The confirmed feed amount in the corresponding input window is: The unit is kilogram, and the mining area is... The component weighted vector of the first The true value of each element is , representing the mass fraction of the corresponding element, is dimensionless, then the th Theoretical mixed output value of each element satisfy:

[0092] ;

[0093] in, For the first The theoretical mixed output quality score of each element. The number of mining areas participating in this output window. For the mining area index variable, For the mining area Confirmed feed amount For mining area The component weighted vector of the first One portion, This is an index variable for the component dimension. (Molecule) The unit is kilogram, and the denominator is... The unit is kilogram, and both have the same dimension. This represents a dimensionless mass fraction. Each participating party and resource utilization data coordination node performs a weighted summation protocol on the component share of each mining area within the encrypted domain to calculate... The secret share does not require the explicit disclosure of the parties' contributions.

[0094] Step 602: Obtain the actual component data collected in the output window by the online detection equipment at the production line output end, and perform secret sharing processing on the actual component data to generate the corresponding share.

[0095] Step 603: Perform a secure comparison protocol between the secret share of the theoretical mixing output value and the secret share of the actual component data, calculate the yield deviation of each element, determine whether the yield deviation exceeds the allowable threshold, and output the Boolean judgment result.

[0096] Step 604: For segments in the encrypted state mapping relationship where windows overlap, apply a mixed hysteresis effect correction coefficient and re-execute the comparison and judgment in step 603, and output the encrypted state conservation verification status flag of each output window.

[0097] It should be noted that the above-mentioned mixed lag effect correction coefficient is determined as follows: Let the output window... From the previous output window The proportion of residual materials is The corrected theoretical mixed output value is Multiply by the theoretical value of the current output window and add Multiply by the theoretical value of the previous output window, where Determined based on the ratio of the mixing chamber volume to the material throughput per unit time. This is the output window index variable.

[0098] Step 7: Perform desensitization anomaly attribution analysis.

[0099] For output windows whose encrypted conservation verification status is flagged as abnormal, each party performs a secure differential calculation on the component share of its corresponding input window and the component share of the previously verified material window. Without exposing absolute component values, the variability of the component inputs from each party is determined to be abnormal, and the anonymized anomaly attribution is output and associated with the corresponding time window segment.

[0100] It should be noted that the aforementioned security differential calculation refers to each party calculating the difference between the current input window component share and the component share of a previously verified material window in the encrypted domain. The parties collaboratively execute a security comparison protocol to determine whether the magnitude of this difference exceeds the component variability threshold. Anomaly attribution is output as a combination of mine area identifier and time window segment identifier, and does not contain plaintext data of component absolute values ​​or differences.

[0101] Step 8: Generate full-process data management records.

[0102] The spatiotemporal weighted component fingerprints, encrypted state mapping relationship identifiers, encrypted state conservation verification status flags, anomaly attribution pointers, and contribution calculation results of each party from each output window are encapsulated into resource utilization data management records. A full-process data management report for mine waste resource utilization is generated by concatenating these records in chronological order using hash pointers. Each party can independently verify the integrity of its own records but cannot deduce the original component data of other parties.

[0103] It should be noted that the aforementioned hash pointer chaining refers to each resource utilization data management record including the hash value of the previous resource utilization data management record as a pointer field, forming a unidirectional chain structure. The contribution calculation results of each party are obtained based on the boundary values ​​of the feasible proportion interval in the effective complementary pairing data records and the material input confirmed by each party. When verifying the integrity of their own records, each party compares the locally retained share data with the hash value of the corresponding share in the chained records.

[0104] For the anomaly attribution analysis data that passed step 7, step 8, record encapsulation, can be further performed. The anomaly attribution pointer field in step 8 only records the attribution direction identifier and does not record the intermediate results of the differential calculation that triggered the attribution determination.

[0105] In this embodiment of the application, to support subsequent auditing and dispute arbitration, in addition to step 8, the following steps are also included: Each participating party's digital signature timestamp is attached to each resource utilization data management record. Each party's signature only covers its own relevant share data and Boolean judgment result fields, and does not cover the data fields of other parties. When any party raises a dispute, the resource utilization data coordination node can retrieve the signature data of the corresponding record to verify the confirmation status of each party.

[0106] On the other hand, the present invention also proposes a data management system for the resource utilization of mine waste, including:

[0107] The adaptive window partitioning module is used to acquire continuous component data streams from online component detection equipment in each mining area, identify steady-state segments and transition points based on Euclidean distance sequences, and adaptively delineate material window sequences.

[0108] The component fingerprint generation and secret sharing module is used to calculate the spatiotemporally weighted component fingerprint for each material window, perform secret sharing processing to generate random shares, and transmit the random share data and material window timestamps to the resource utilization data coordination node.

[0109] The secure complementary pairing determination module is used to execute a secure comparison protocol based on secret share data at the resource utilization data coordination node, determine the satisfaction or gap status of each material window, and generate a candidate complementary pairing encrypted index set.

[0110] The secure mixing ratio verification module is used to perform a two-party secure mixing ratio verification protocol on the material window pairs involved in the candidate complementary pairing encrypted index set, and output the Boolean judgment result of the effective mixing ratio and the boundary value of the desensitized feasible ratio range.

[0111] The timing alignment mapping module is used to align the input window share sequences of each party with the output window based on the timing delay offset, and establish a cryptographic mapping relationship.

[0112] The encrypted domain conservation verification module is used to construct a set of multi-party material conservation equations in the encrypted domain, calculate the secret share of the theoretical mixed output value and perform a secure comparison protocol with the actual output component data, and output an encrypted conservation verification status flag.

[0113] The anomaly attribution module is used to perform secure differential calculations on output windows whose encrypted conservation verification status flag is anomaly, and output anomaly attribution directions without exposing absolute component values.

[0114] The full-process recording and encapsulation module is used to encapsulate the spatiotemporal weighted component fingerprints, encrypted state mapping relationship identifiers, encrypted state conservation verification status flags, anomaly attribution pointers, and contribution calculation results of each party in each output window into resource utilization data management records, and generate a full-process data management report for the resource utilization of mine waste by concatenating them according to hash pointers.

[0115] By way of example, the system proposed in this invention may also include various features and combinations thereof in the method embodiments, which will not be elaborated here.

[0116] This invention performs secret sharing processing during the component data acquisition stage, splitting the original component weighted vectors of each mining area into random shares and transmitting them to the resource utilization data coordination node. Subsequent complementary pairing determination, mixing ratio verification, material conservation equation solving, and yield deviation comparison are all completed in the encrypted domain. None of the participating parties or the resource utilization data coordination node can restore the plaintext components of other parties. Therefore, it achieves collaborative management of cross-mining area data without relying on the submission of plaintext data by each party, overcoming the problem of data non-centralized management caused by parties refusing to share data due to commercial sensitivity. Meanwhile, by adaptively defining material window boundaries using steady-state segments and abrupt change points based on Euclidean distance sequences, replacing batch segmentation methods with fixed time intervals or fixed weights, the heterogeneity of components within each material window is controlled by steady-state threshold constraints, avoiding drastic changes in composition within the material window caused by fixed segmentation. Furthermore, spatiotemporal weighted component fingerprinting reduces the interference of material window edge sampling points on material window features, and precise correspondence between input and output windows is established through time delay offset calculation and time-series alignment, overcoming the problem of inaccurate input-output data correspondence caused by physical diffusion and mixing hysteresis effects under continuous flow conditions. Based on the synergistic effect of these two technical approaches, the data management goal of verifying material conservation and assigning responsibility is achieved while ensuring data sovereignty for all parties.

[0117] See Figures 2-9 The following application examples are proposed:

[0118] There are three independent mining companies, A, B, and C, in a certain area. The tailings waste generated by each mining area contains three main elements: iron, sulfur, and silicon (total composition dimensions). Three companies feed their waste into the same continuous flow resource recovery production line for the production of building material fillers. Each mine is equipped with online X-ray fluorescence component detection equipment and local data management terminals. The production line has a mixing point and online detection equipment at the output end. All parties refuse to submit plain text composition data to the production line operator, relying on secure multi-party computation to complete complementary pairing, conservation verification, and contribution calculation. The rated material flow rate of the production line is 0.35 meters per second. The pipeline length from the mixing point to the output detection point is 126 meters in mine A, 105 meters in mine B, and 84 meters in mine C.

[0119] Step 1: Adaptively define the material window sequence based on the composition change characteristics.

[0120] Each mining area's local data management terminal acquires a continuous sampling data stream from online component analysis equipment, with a sampling frequency of once per minute. Taking mining area A as an example, the backtracking window length... Set to 5 sampling points, number of consecutive judgment points is set to 3. The system sequentially calculates the Euclidean distance sequence between the component vector of each sampling point and the previous 5 sampling points, and identifies the steady-state segment and the jump point. For Mining Area A, the Euclidean distance detected at the 23rd sampling point exceeds the mutation threshold (0.18), so it is determined as a jump point, which triggers material window splitting. The preceding steady-state segment is delimited as WA-01, and the subsequent segment is delimited as WA-02. Mining Area B and Mining Area C complete their respective material window sequence division synchronously.

[0121] Table 1 Adaptive material window delimitation results of each mining area

[0122]

[0123] Step 2: Generate spatio-temporal weighted component fingerprints and perform secret sharing processing.

[0124] Taking the material window WA-01 of Mining Area A as an example, the central moment of the window is 08:11 on a certain day, and the Gaussian attenuation bandwidth parameter is taken as one quarter of the window time span (about 5.5 minutes). After each sampling point calculates the weight and performs normalization according to the formula , the weighted mean and weighted standard deviation are calculated for three elements, namely iron, sulfur and silicon, respectively. After the weighted mean is truncated and sorted, the standard deviation feature values are spliced, and SHA-256 hash operation is performed to generate the spatio-temporal weighted component fingerprint FPA-01. Subsequently, Mining Area A locally performs processing on the weighted component vector additive secret sharing processing ( ), generates random shares , , calculates the third share , and transmits the non-own shares together with the window timestamp to the resource utilization data coordination node. Mining Area B and Mining Area C perform the same operation on their respective material windows.

[0125] Table 2 Spatio-temporal weighted component fingerprints and secret sharing processing results of material windows

[0126]

[0127] Step 3: Perform secure complementary matching judgment based on secret shares.

[0128] The resource utilization data coordination node receives the share data from three parties and establishes a time-series index according to the material window timestamps. The component requirement values of iron, sulfur and silicon for the target resource utilization path are preset as 0.42, 0.15 and 0.35 in sequence. The coordination node and each mining area cooperatively execute a secure comparison protocol, and output the satisfied (1) or missing (0) Boolean flag for each dimension of each window in the ciphertext domain.

[0129] Taking the matching of WA-01 and WB-01 as an example, according to the complementary matching degree scoring formula Calculation, wherein , and are respectively the boolean flags of Mining Area A and Mining Area in the th dimension. Among the three dimensions, 2 dimensions are complementary, and the score is higher than the minimum pairing threshold of 0.50, so it enters the candidate complementary pair. The score of WJia-01 and WBing-01 is 0.33, which is lower than the threshold and is filtered out.

[0130] Table 3 Boolean flag vectors and complementary matching scores of each material window

[0131]

[0132] Step 4: Perform secure mixing ratio verification.

[0133] For the candidate complementary pair WJia-01 and WYi-01, the two mining areas Jia and Yi respectively take local composition weighted vectors as private inputs, execute a two-party garbled circuit protocol, and calculate the respective mixing ratio of the three dimensions of iron, sulfur and silicon constraint interval , take the intersection of the constraints of the three dimensions, and determine whether there exists such that the mixing errors of all three dimensions are within the allowable deviation . The protocol outputs a Boolean determination result as "valid", outputs the desensitized feasible ratio interval in the form of a quantization level ("Medium leaning toward Jia"), and generates a valid complementary pairing record EM-01. WYi-01 and WBing-01 similarly generate a valid complementary pairing record EM-02.

[0134] Table 4 Output results of secure mixing ratio verification

[0135]

[0136] Step 5: Establish time-series alignment mapping of input and output windows based on time lag offset.

[0137] Calculate the time lag offset of each party based on the pipeline length and average material flow velocity of each mining area meters per second: the time lag offset of Jia Mining Area is seconds (i.e., 6 minutes), that of Yi Mining Area is seconds (i.e., 5 minutes), and that of Bing Mining Area is seconds (i.e., 4 minutes).

[0138] Adaptive window division is synchronously performed on the component data stream at the output end of the production line to generate an output window sequence. The input window timestamps of each party are respectively added with the corresponding time lag offsets, and then matched with the time range of the output window. After WA-01 (08:00 to 08:22) is offset by 6 minutes, it corresponds to the output time period 08:06 to 08:28, which overlaps with the output window WO-01 (08:05 to 08:32). After WB-01 is offset by 5 minutes, it also falls into the range of WO-01, and an encrypted mapping relationship MAP-01 is established between the two. After WC-01 is offset by 4 minutes, its time range spans WO-01 and WO-02, and is split and mapped to the two output windows according to the time overlapping proportion.

[0139] Table 5 Calculation of time lag offsets and timing alignment mapping results

[0140]

[0141] Step 6: Perform material conservation verification in the ciphertext domain.

[0142] For the mapping relationship MAP-01, Party A, Party B, Party C and the coordination node collaboratively execute the secure multi-party linear equation system solving protocol, and each party confirms that the feeding amount of Mining Area A is kg, Mining Area B is kg, and Mining Area C inputs the corresponding share of WO-01 kg (split from the total 2000 kg of WC-01 according to a proportion of 78%). According to the formula calculate the theoretical mixed output value share of the three elements in the ciphertext domain, wherein is the theoretical mixed output value of the dimension, is the feeding amount of the party, is the weighted component value of the party in the dimension. Subsequently, secret sharing processing is performed on the actual component data of the output window WO-01 to generate shares, and a secure comparison protocol is executed between the theoretical shares and the actual shares. The yield deviations of iron and silicon are within the allowable threshold, while the deviation of sulfur exceeds the threshold, and the preliminary conservation status of WO-01 is output as "abnormal pending correction". A mixed lag correction coefficient ( ) is applied to the overlapping section split and mapped from Mining Area C. After correction, the deviation of sulfur recovers to within the threshold, the final conservation verification status flag of WO-01 is updated to "passed", and WO-02 will be subject to the same type of verification after subsequent window data is supplemented.

[0143] Table 6 Status record of ciphertext domain conservation verification process

[0144]

[0145] Step 7: Perform desensitized abnormal attribution analysis.

[0146] After correction, WO-01 passes the conservation verification, and there is no need to trigger abnormal attribution. In the subsequent period, the conservation verification status flag of WO-03 is "abnormal". All parties perform secure difference calculation on the component shares of the corresponding input window and the previous passed window, and collaboratively determine whether the variation degree of each party's feeding components exceeds the threshold. The protocol result shows that the absolute variation degree difference of the input window corresponding to Mining Area A exceeds the set threshold, while Mining Area B and Mining Area C do not. The system outputs the desensitized abnormal attribution direction: mine identifier "Mining Area A", associated time window segment "W_A-03", which does not contain any plaintext of absolute component values or difference values.

[0147] Step 8: Generate the whole-process data management record.

[0148] The space-time weighted component fingerprint of WO-01 (such as FP: 3A7C9F2B), encrypted mapping relationship identifier MAP-01, conservation verification status flag "passed (encrypted state)", abnormal attribution direction (WO-03 corresponds to Mining Area A W_A-03), and the accounting result of each party's contribution (feasible proportion interval level and confirmed feeding amount based on EM-01 and EM-02) are packaged into a resource utilization data management record REC-01. REC-01 contains the hash pointer of the previous record, forming a chain structure. Each participant attaches a digital signature timestamp to their own relevant fields respectively, and finally generates the whole-process data management report for mine waste resource utilization in chronological order.

[0149] Table 7 Encapsulation structure of the whole-process data management record

[0150]

[0151] Throughout the whole implementation process, data starts from the component sampling data stream of each mining area, is converted into a structured material window sequence through the adaptive window division in step 1, and then converted into fingerprint and share data through Gaussian weighting calculation and secret sharing processing in step 2. In steps 3 and 4, the share data drives the security protocol to complete candidate pairing screening and mixing ratio verification, and the valid pairing records are combined with the time lag offset in step 5 to establish an accurate input-output timing correspondence. Step 6 completes the conservation equation solving and yield deviation determination in the ciphertext domain, step 7 traces and attributes the source of the abnormal window, and step 8 encapsulates all intermediate state flags and fingerprints into a hash chain record. Throughout the whole process, the original plaintext of each party's components does not leave the local area, the output data of each step is used as the input basis for subsequent steps, the data flow logic is coherent, and the responsibility attribution is traceable.

[0152] The embodiments of the present invention have been described above. However, the embodiments are not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make more equivalent embodiments under the guidance of the present embodiments, and all of them are within the protection scope of the present embodiments.

Claims

1. A data management method for the resource utilization of mine waste, characterized in that, include: Each mining area acquires local continuous component data streams, identifies steady-state segments and transition points based on the Euclidean distance sequence of component vectors at sampling points, and adaptively delineates material window sequences; For each material window, attenuation weights are assigned according to the time distance of the sampling point from the center of the window to calculate the component weight vector. Secret sharing processing is performed to generate random shares and transmit them to the data coordination node. Based on the secret share, the satisfaction or gap status of each material window component in the encrypted domain is determined, candidate complementary pairings are generated, and the safe mixing ratio verification is performed on the candidate complementary pairings. Based on the time delay offset, the input windows of each party are aligned with the output windows of the production line to establish an encrypted mapping relationship; In the encrypted domain, a set of multi-party material conservation equations is constructed. The theoretical mixed output value is compared with the actual output components to perform a safety comparison and output a conservation verification status flag. The conservation verification status flag and the calculation results of each party's contribution are encapsulated into data management records, and a full-process data management report is generated by concatenating them according to hash pointers.

2. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, The identification of steady-state segments and transition points based on the Euclidean distance sequence of sample point component vectors includes: The component vector of the current sampling point and its preceding first to second components... The Euclidean distance is calculated for the component vectors of each sampling point, forming a length of... The distance value sequence, where The preset backtracking window length; when continuous When the Euclidean distance of all sampling points is lower than the steady-state threshold, the segment is marked as a steady-state segment. The preset number of continuous judgment points; when the Euclidean distance exceeds the mutation threshold, it is marked as a jump point; the jump point is used as the dividing position of adjacent material windows, and continuous steady-state segments are assigned to the same material window; when the number of sampling points between two adjacent jump points is less than the minimum window length threshold, the segment is merged into the previous material window and marked as containing a transition segment.

3. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, The attenuation weight allocation based on the time distance of the sampling point from the center of the window includes: Let the center time of the material window be... The sampling point time is The weight of the sampling point Calculate using the following formula: ； in, For the sampling point time, At the center of the material window, The Gaussian attenuation bandwidth parameter is taken as a preset proportion of the material window time span; the weights of each sampling point are normalized and used to calculate the weighted mean and weighted standard deviation; the weighted mean is truncated and sorted according to its value and then concatenated, with the weighted standard deviation feature value added, and a hash operation is performed to generate the spatiotemporal weighted component fingerprint.

4. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, The secret sharing process employs an additive secret sharing scheme: Let the component weighting vector be The number of participants is ,generate random vectors The dimension of each random vector and the component weighting vector The dimensions are the same, calculate the th Each share is ,in For the first Each share vector For the summation index variable of shares, each share satisfies Each mining area retains its own share and distributes the remaining shares to the corresponding party or data coordination node.

5. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, The generation of candidate complementary pairs also includes: Calculate the complementarity matching score for the Boolean flag vectors of each material window. : ； in, The complementarity and matching degree between the two mining areas is scored. The total number of component dimensions. For component dimension index variables, and The first Boolean vectors of the two mining areas are respectively the second Boolean vectors of the two mining areas. One portion, This is an indicator function; it filters material window pairs with a complementary match degree below the minimum pairing threshold, and only performs candidate complementary labeling on material window pairs with a complementary match degree exceeding the minimum pairing threshold.

6. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, The process of performing secure mixing ratio verification on candidate complementary pairings includes: Each party uses its local component weighted vector as a private input, and a confusion circuit is used to determine whether a mixing ratio exists. Make all All dimensions are satisfied: ； in, and The weighted vectors of the two components are respectively the first and second parts. One portion, The first vector required for the target component One portion, For the first Tolerance for each component dimension The component dimension index variable is used; the Boolean determination result of the effective mixing ratio and the boundary value of the feasible ratio range after desensitization are output. The desensitization process only outputs the quantization level of the boundary value of the feasible ratio range.

7. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, The timing alignment of the input windows of each party with the production line output window based on the time delay offset includes: Obtain the material window timestamp and mixing confluence point location parameters for each mining area to confirm material input. Let the mining area be... The length of the pipeline from the mixing point to the output detection point is The average flow velocity of the material is Then the time delay offset ,in The output window sequence is generated using the same adaptive material window partitioning method as the input window for the component data stream at the output end of the production line. The timestamps of each input window are added with the corresponding time delay offsets and then matched with the time range of the output window. When an input window crosses two adjacent output windows after a time delay offset, the share data of the input window is split and mapped to the two output windows according to the time overlap ratio.

8. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, The construction of a multi-party material conservation equation set in the encrypted domain includes: Let the number of mining areas participating in a certain output window be . mining area The confirmed feed amount is mining area The component weighted vector of the first The elements are Then the first Theoretical mixed output value of each element satisfy: ； in, The number of mining areas participating in this output window. For the mining area index variable, For mining area Confirmed feed amount The component dimension index variable is used; for segments with overlapping windows, a mixing hysteresis effect correction coefficient is applied and the segment is re-evaluated. The correction coefficient is determined based on the ratio of the mixing chamber volume of the equipment to the material throughput per unit time.

9. The data management method for the resource utilization of mine waste according to claim 1, characterized in that, This also includes attribution analysis of desensitization abnormalities: For output windows whose conservation verification status is marked as abnormal, each party calculates the difference between the current input window component share and the component share of the material window that has passed the conservation verification in the encrypted field, and collaboratively executes a security comparison protocol to determine whether the modulus of the difference exceeds the component variability threshold; the anomaly attribution pointer is output in the form of a combination of mining area identifier and time window segment identifier, and does not contain plaintext data of component absolute value or difference; each data management record in the full-process data management report is appended with the digital signature timestamp of each participant, and each party's signature only covers its own relevant share data and Boolean judgment result field.

10. A data management system for the resource utilization of mine waste, used to execute the data management method for the resource utilization of mine waste according to any one of claims 1 to 9, characterized in that, include: The adaptive window partitioning module is used to acquire continuous component data streams from online component detection equipment in each mining area, identify steady-state segments and transition points based on Euclidean distance sequences, and adaptively delineate material window sequences. The component fingerprint generation and secret sharing module is used to calculate the component weighting vector by allocating attenuation weights to each material window according to the time distance of the sampling point from the center of the window, perform secret sharing processing to generate random shares, and transmit the random share data and the material window timestamp to the data coordination node. The secure complementary pairing determination module is used to execute a secure comparison protocol in the encrypted domain based on secret share data, determine the satisfaction or gap status of each material window component, and generate candidate complementary pairs. The secure mixing ratio verification module is used to perform a two-party secure mixing ratio verification protocol on candidate complementary pairings, and outputs the Boolean determination result of the effective mixing ratio and the boundary value of the desensitized feasible ratio range; The timing alignment mapping module is used to align the input window share sequence of each party with the production line output window based on the time delay offset, and establish an encrypted mapping relationship. The encrypted domain conservation verification module is used to construct a set of conservation equations for multiple materials in the encrypted domain, perform a safe comparison between the theoretical mixed output value and the actual output component data, and output a conservation verification status flag. The desensitization anomaly attribution module is used to perform safe differential calculation on output windows whose conservation verification status flag is abnormal, and output the anomaly attribution direction without exposing the absolute component value; The full-process recording and encapsulation module is used to encapsulate the conservation verification status flags and the calculation results of each party's contribution into data management records, and generate a full-process data management report by concatenating them according to hash pointers.

Images