Precious metal extraction data query method and system

Abstract

The invention discloses a precious metal refining data query method and system, and relates to the technical field of industrial big data processing, and the method comprises the steps: collecting operation parameters in a precious metal refining process to form original data, carrying out the time sequence division of the original data according to the refining time, and generating time sequence data blocks; constructing a query index tree based on the time sequence data blocks, wherein each node of the query index tree stores a feature value of one time sequence data block; receiving a query request, and extracting a target time period and a target parameter range from the query request; and in the query index tree, by comparing the target parameter range with the node feature value, positioning a time sequence data block meeting a condition, and extracting corresponding original data from the time sequence data block as a query result. According to the method, the performance bottleneck in large-scale time sequence data query is solved, the index structure can be intelligently optimized according to the historical query mode, and efficient, accurate and intelligent retrieval of data in the precious metal extraction process is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial big data processing, and in particular to a method and system for querying precious metal refining data. Background Art

[0002] As one of the core technologies of the modern metallurgical industry, precious metal refining plays a vital role in many high-tech fields such as electronics, new energy, and medical treatment. With the in-depth development of Industry 4.0 and big data technology, data management and analysis in the precious metal refining process have gradually become the key to improving production efficiency and process level. In recent years, domestic and foreign research institutions and enterprises have begun to focus on data collection, storage, and intelligent analysis technologies in the refining process, striving to achieve refined management and process optimization of precious metal refining through data-driven methods. However, traditional data query methods often face technical bottlenecks such as large data volume, low query efficiency, and complex index construction, which seriously restrict the real-time nature and value mining of data.

[0003] The existing data query technology for precious metal refining has limitations in query performance and data organization. Traditional methods usually use relational databases or sequential scanning to query data. This method has a slow retrieval speed when facing large-scale, high-dimensional time series data, and does not provide sufficient support for complex time and parameter range queries. Especially when it is necessary to accurately locate operating parameters within a specific time period, existing technologies often need to traverse a large amount of irrelevant data, which not only increases the computational burden, but also reduces the real-time and accuracy of data utilization. In addition, existing technologies lack flexibility and scalability in data index construction, making it difficult to quickly respond to query requirements of different dimensions and granularity. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention is proposed.

[0005] Therefore, the problem to be solved by the present invention is that the existing methods have slow retrieval speed when facing large-scale, high-dimensional time series data, and lack support for complex time and parameter range queries.

[0006] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, an embodiment of the present invention provides a method for querying precious metal refining data, which comprises:

[0008] Collecting the operating parameters in the precious metal refining process to form raw data, and dividing the raw data into time series according to the refining time to generate time series data blocks;

[0009] Building a query index tree based on the time series data block, each node of the query index tree stores a feature value of a time series data block;

[0010] receiving a query request, and extracting a target time period and a target parameter range from the query request;

[0011] In the query index tree, by comparing the target parameter range with the node feature value, the time series data block that meets the conditions is located, and the corresponding original data is extracted from the time series data block as the query result.

[0012] As a preferred solution of the method for querying precious metal refining data of the present invention, wherein: the operating parameters include temperature parameters, pressure parameters, and purity parameters;

[0013] Generating the time series data block includes:

[0014] Dividing the original data into time series according to the extraction time, and calculating the purity value of the intermediate moment according to the purity parameter in the original data to obtain a preliminary time series data block;

[0015] Performing time series correlation evaluation on the preliminary time series data block and calculating a quality index;

[0016] Optimizing the preliminary time series data block based on the quality indicator to obtain the time series data block;

[0017] Storing the time series data block in a time series table and creating a composite index;

[0018] The optimization is shown in the following formula:

[0019] R′(t k )=R(t i )·w1+R(t j )·w2

[0020] Among them, t k is the time to be interpolated, R′(t k ) is t i and t j The purity value at the middle moment of time, t i and t j For two adjacent purity collection moments, the weight coefficient is calculated as follows:

[0021]

[0022] Where λ is the time decay coefficient.

[0023] As a preferred solution of the method for querying precious metal refining data of the present invention, wherein: the characteristic value includes the maximum temperature parameter, the minimum temperature parameter, the average pressure parameter, the average purity parameter, the parameter fluctuation range value and the timestamp information of the time series data block;

[0024] The maximum temperature parameter is selected from all the temperature sensor data in the time series data block to obtain the maximum value, and the minimum temperature parameter is selected from all the temperature sensor data in the time series data block to obtain the minimum value;

[0025] The average pressure parameter is obtained by calculating the arithmetic mean of the data collected by all pressure sensors in the time series data block within the time period, and the average purity parameter is obtained by calculating the weighted average of all purity data in the time series data block, and the weighting coefficient decays exponentially as the time interval increases;

[0026] The parameter fluctuation range value is obtained by calculating the square root of the sum of squares of the differences between the maximum and minimum values ​​of the temperature parameter, the pressure parameter and the purity parameter.

[0027] As a preferred solution of the query method for precious metal refining data of the present invention, wherein: in the query index tree, each non-leaf node additionally stores subtree range information, including upper and lower bounds of the characteristic values ​​of all nodes in the subtree, for quickly determining whether the query condition is related to the subtree;

[0028] The non-leaf nodes establish neighboring pointers between nodes through the similarity relationship of the eigenvalues; the similarity is calculated by weighted distance of each component of the eigenvalue; the weight coefficient is dynamically adjusted based on the historical query frequency; the leaf node stores the storage address pointing to the actual time series data block, and establishes a bidirectional linked list between adjacent leaf nodes.

[0029] As a preferred solution of the method for querying precious metal refining data of the present invention, the bidirectional linked list established between adjacent leaf nodes is expressed by the following formula:

[0030] L(N)={N i |(F(N),F(N i ))≤δ}

[0031] Among them, δ is the proximity threshold, which is adaptively adjusted based on historical query statistics:

[0032]

[0033] Among them, μ is the mean of historical distance, σ is the standard deviation, and Q is the number of queries.

[0034] As a preferred solution of the method for querying precious metal refining data of the present invention, the query index tree further includes:

[0035] When the number of child nodes in a node exceeds the maximum limit, the node is split according to the distribution of its eigenvalues, and the split point that makes the difference in the eigenvalues ​​of the child nodes after the split is selected is selected; when the number of child nodes in a node is lower than the minimum limit, the adjacent nodes with the closest eigenvalues ​​are merged.

[0036] As a preferred solution of the method for querying precious metal refining data of the present invention, the time series data block that is located to meet the conditions includes:

[0037] The target parameter range is compared with the node characteristic value; the comparison process first checks whether the parameter boundary of the node intersects with the target range, specifically including whether the maximum temperature parameter of the comparison node is greater than the lower limit of the target temperature range and whether the minimum temperature parameter is less than the upper limit of the target temperature range, whether the average pressure parameter of the comparison node is between the upper and lower limits of the target pressure range, and whether the average purity parameter of the comparison node is between the upper and lower limits of the target purity range; when the characteristic value of the node meets any of the above conditions, the node is added to the candidate node set.

[0038] In a second aspect, an embodiment of the present invention provides a query system for precious metal refining data, which includes: a collection module for collecting operating parameters in the precious metal refining process to form raw data, and dividing the raw data into time series according to the refining time to generate time series data blocks;

[0039] A retrieval tree construction module, used to construct a query index tree based on the time series data block, each node of the query index tree stores a feature value of a time series data block;

[0040] An extraction module, used for receiving a query request and extracting a target time period and a target parameter range from the query request;

[0041] The query module is used to locate the time series data block that meets the conditions in the query index tree by comparing the target parameter range with the node feature value, and extract the corresponding original data from the time series data block as the query result.

[0042] In a third aspect, an embodiment of the present invention provides a computer device, including a memory and a processor, wherein the memory stores a computer program, wherein: when the processor executes the computer program, it implements any step of the above-mentioned method for querying precious metal refining data.

[0043] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein: when the computer program is executed by a processor, any step of the above-mentioned method for querying precious metal refining data is implemented.

[0044] The beneficial effects of the present invention are: constructing a multi-dimensional balanced query index tree, significantly improving the efficiency and accuracy of data retrieval. Compared with the traditional sequential scanning method, this solution can quickly locate the target time series data block, and adopt a multi-stage retrieval strategy, through dynamic pruning and priority sorting mechanism, greatly reducing the query response time. At the same time, the present invention not only solves the performance bottleneck in large-scale time series data query, but also intelligently optimizes the index structure according to the historical query pattern, realizing efficient, accurate and intelligent retrieval of precious metal refining process data, and has significant technical innovation and practical application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for describing the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. Among them:

[0046] Figure 1 Flowchart of a query method for precious metals refining data. DETAILED DESCRIPTION

[0047] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0048] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0049] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive with other embodiments.

[0050] The present invention is described in detail with reference to schematic diagrams. When describing the embodiments of the present invention, for the sake of convenience, the cross-sectional diagrams showing the device structure will not be partially enlarged according to the general scale, and the schematic diagrams are only examples, which should not limit the scope of protection of the present invention. In addition, in actual production, the three-dimensional dimensions of length, width and depth should be included.

[0051] At the same time, in the description of the present invention, it should be noted that the directions or positional relationships indicated by the terms "upper, lower, inner and outer" are based on the directions or positional relationships shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific direction, be constructed and operated in a specific direction, and therefore cannot be understood as limiting the present invention. In addition, the terms "first, second or third" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0052] In the present invention, unless otherwise clearly specified and limited, the terms "install, connect, connect" should be understood in a broad sense, for example: it can be a fixed connection, a detachable connection or an integral connection; it can also be a mechanical connection, an electrical connection or a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0053] Example 1

[0054] Reference Figure 1 , which is the first embodiment of the present invention, provides a method for querying precious metal refining data, comprising:

[0055] S1: collecting temperature parameters, pressure parameters, and purity parameters during the precious metal refining process to form raw data, and storing the raw data in a database;

[0056] The temperature parameters, pressure parameters and purity parameters in the precious metal refining process are collected to form raw data, wherein the temperature parameters are collected by multiple temperature sensors arranged in the refining equipment, and the temperature sensors are distributed in the feed end, reaction chamber and discharge end of the refining equipment, and the collection frequency is once per second; the pressure parameters are collected by pressure sensors arranged in the refining equipment, and the pressure sensors are installed at the top and bottom of the reaction chamber, and the collection frequency is once per second; the purity parameters are collected in real time at the discharge end by a spectrometer, and the collection frequency is once every five seconds; the temperature parameters, pressure parameters and purity parameters are marked according to the collection time to form the raw data, and the raw data are transmitted to the database server through the data collection module, and the database server adopts a distributed storage architecture to store the raw data in the corresponding data table in the database.

[0057] S2: Divide the original data into time series according to the extraction time to generate time series data blocks, each of which contains temperature parameters, pressure parameters, and purity parameters within a corresponding time period;

[0058] The raw data is divided into time series according to the extraction time, and the time interval of the time series division is 30 minutes. For the time point t, the data range of the time series data block B(t) is [t, t+1800s]; the temperature parameter set T(t), pressure parameter set P(t) and purity parameter set R(t) in the time series data block B(t) are preprocessed according to the following formula:

[0059] T(t)={T i (t)|i∈[1,n],t∈[t,t+1800s]}

[0060] P(t)={P j (t)|j∈[1,m],t∈[t,t+1800s]}

[0061] Among them, T i (t) represents the collected value of the ith temperature sensor at time t, n is the total number of temperature sensors, P j (t) represents the collected value of the jth pressure sensor at time t, and m is the total number of pressure sensors.

[0062] For purity parameters with lower acquisition frequency, the weighted time decay linear interpolation method is used to calculate the purity value at the intermediate time:

[0063] R′(t k )=R(t i )·w1+R(t j )·w2

[0064] Among them, t k is the time to be interpolated, t i and t j For two adjacent purity collection moments, the weight coefficient is calculated as follows:

[0065]

[0066] Among them, λ is the time attenuation coefficient, which is 0.001.

[0067] Furthermore, the time series correlation evaluation is performed on the preprocessed parameters, and the quality index Q(t) of the time series data block is calculated:

[0068] Q(t)=α·Corr(T(t),P(t))+β·Corr(P(t),R′(t))+γ·Corr(T(t),R′(t))

[0069] Among them, Corr() is the Pearson correlation coefficient of the parameter sequence, and α, β, and γ are weight coefficients.

[0070] The time series data blocks are optimized based on the quality index Q(t). When Q(t)<threshold θ, the sliding window method is used to correct the abnormal data points:

[0071] X′(t)=median{X(tk),…,X(t),…,X(t+k)}

[0072] Among them, X represents the parameter to be corrected, k is the sliding window radius, and its value is 3.

[0073] The processed time series data block is stored in the time series table of the database, and a composite index I(t) is established:

[0074]

[0075] Among them, hash() is a hash function, timestamp() is a timestamp conversion function, It is a bitwise XOR operation.

[0076] S3: constructing a query index tree based on the time series data block, wherein each node of the query index tree stores a characteristic value of a time series data block, wherein the characteristic value includes a maximum temperature parameter, a minimum temperature parameter, an average pressure parameter, and an average purity parameter of the time series data block;

[0077] A multidimensional balanced query index tree is constructed based on the time series data block, the query index tree adopts a three-layer structure of a root node, an intermediate node and a leaf node, and the maximum number of child nodes of each node is sixteen; each node of the query index tree stores a set of characteristic values, and the characteristic values ​​include the maximum temperature parameter, the minimum temperature parameter, the average pressure parameter, the average purity parameter, the parameter fluctuation range value and the timestamp information of the time series data block; the maximum temperature parameter is filtered from all the temperature sensor data in the time series data block to obtain the maximum value, and the minimum temperature parameter is filtered from all the temperature sensor data in the time series data block to obtain the minimum value; the average pressure parameter is obtained by calculating the arithmetic mean of the data collected by all the pressure sensors in the time series data block within the time period, and the average purity parameter is obtained by calculating the weighted average of all the purity data in the time series data block, and the weighting coefficient decays exponentially with the increase of the time interval; the parameter fluctuation range value is obtained by calculating the square root of the sum of the squares of the differences between the maximum and minimum values ​​of the temperature parameter, the pressure parameter and the purity parameter.

[0078] That is, the depth D of the tree is calculated by the following formula:

[0079]

[0080] Among them, m is the order of the tree, which is 16, and B is the total number of time series data blocks;

[0081] It should be noted that the calculation of the eigenvalue is as follows:

[0082] F(t)=[T max (t),T min (t),P avg (t),R avg (t),V(t)]

[0083] Among them, F(t) is the feature vector of the time series data block corresponding to time t, t is the starting timestamp of the time series data block, T max (t) is the maximum temperature value in the time series data block, T min (t) is the minimum temperature value in the time series data block, P avg (t) is the average pressure value in the time series data block, R avg (t) is the average purity value in the time series data block, and V(t) is the comprehensive value of the parameter fluctuation range.

[0084] in:

[0085] T max (t) = max{T i (t)|i∈[1,n],t∈[t,t+1800]}

[0086] T min (t) = min{T i (t)|i∈[1,n],t∈[t,t+1800]}

[0087]

[0088] Among them, T i (t) is the temperature value of the i-th temperature sensor at time t, n is the total number of temperature sensors, P j (t) is the pressure value of the jth pressure sensor at time t, m is the total number of pressure sensors, R′(t) is the purity value after interpolation processing, and k is the number of purity data points.

[0089] In the query index tree, each non-leaf node additionally stores subtree range information, including the upper and lower bounds of the eigenvalues ​​of all nodes in the subtree, which is used to quickly determine whether the query condition is related to the subtree; the non-leaf nodes establish neighboring pointers between nodes through the similarity relationship of the eigenvalues, and the similarity is calculated by the weighted distance of each component of the eigenvalue, and the weight coefficient is dynamically adjusted based on the historical query frequency; the leaf node stores the storage address pointing to the actual time series data block, and establishes a bidirectional linked list between adjacent leaf nodes to support fast traversal of range queries.

[0090] The non-leaf node N(i) of the query index tree stores multiple index entries E(i,j):

[0091] E(i,j)={F(t),ptr(j)}

[0092] Among them, ptr(j) is a pointer to the next layer of nodes.

[0093] The leaf node L(i) of the query index tree stores the data entry D(i,j):

[0094] D(i,j)={F(t),block ptr (t)}

[0095] Among them, block ptr (t) is a pointer to the actual timing data block.

[0096] The similarity is calculated by weighted distance of each component of the eigenvalue, and the distance is obtained according to the distance metric function between nodes, that is:

[0097] dist(F1,F2)=e1|T max1 -T max2 |e2|T min1 -T min2 |e3|P avg1 -P avg2 |e4|R avg1 -R avg2 |e5|V1-V2|

[0098] Among them, F1 and F2 are two feature vectors to be compared, e1~e5 are the weight coefficients of each feature component, T max1 ,T max2 is the maximum temperature value of the two eigenvectors, T min1 ,T min2 is the minimum temperature value of the two eigenvectors, P avg1 ,P avg2 is the average pressure value of the two eigenvectors, R avg1 ,R avg2 is the average purity value of the two eigenvectors, V1 and V2 are the parameter fluctuation range values ​​of the two eigenvectors.

[0099] It should be noted that a bidirectional linked list L(N) is established between adjacent leaf nodes, as shown below:

[0100] L(N)={N i |(F(N),F(N i ))≤δ}

[0101] Among them, δ is the proximity threshold, which is adaptively adjusted based on historical query statistics:

[0102]

[0103] Among them, μ is the mean of historical distance, σ is the standard deviation, and Q is the number of queries.

[0104] The query index tree maintains the balance of the tree through a dynamic balancing mechanism. When the number of child nodes in a node exceeds the maximum limit, it is split according to the distribution of the node characteristic value, and the split point that makes the difference in the characteristic values ​​of the child nodes after the split is the largest is selected; when the number of child nodes in a node is lower than the minimum limit, the adjacent nodes with the closest characteristic values ​​are merged; the query index tree performs rebalancing operations regularly to adjust the connection relationship between nodes and optimize the overall structure of the tree. The rebalancing operation is triggered when the system load is low to avoid affecting normal query operations.

[0105] More specifically, the index tree is dynamically balanced based on the distance metric function. When a node splits, the split point k that minimizes the following equation is selected:

[0106] Split(k)=∑dist(F i ,C1)+∑dist(F j ,C2)

[0107] Among them, C1 and C2 are the feature centers of the two nodes after splitting.

[0108] It should be noted that the definition of the maximum limit includes the upper limit of node capacity and the upper limit of data volume, wherein the upper limit of node capacity includes that each node can accommodate up to 16 child nodes (m=16), and this limit is based on the following considerations:

[0109] Memory page size: usually 4KB or 8KB. The data structure of a node should be able to fill exactly one or a few memory pages.

[0110] Cache efficiency: Too large a node will reduce the CPU cache hit rate

[0111] Disk I / O efficiency: The amount of data read at a time should be moderate to avoid delays caused by excessive data

[0112] Node splitting overhead: A node that is too large will consume more computing resources when it splits.

[0113] Upper limit of data volume: Each leaf node stores the feature values ​​and pointers of up to 12 time series data blocks (m'=12). The factors to be considered include:

[0114] Storage space for eigenvalues: Each eigenvalue contains multiple components such as temperature, pressure, purity, etc.

[0115] Pointer storage space: including data block pointers and inter-node link pointers

[0116] Reserved space: To support local adjustments during data insertion, about 25% of the space is reserved.

[0117] Definition of the minimum limit (MIN_LIMIT):

[0118] Minimum number of child nodes for non-leaf nodes: That is, half of the maximum limit (rounded up). This is stipulated in order to:

[0119] Ensure the balance of the tree: prevent the tree from degenerating into a linked list

[0120] Maintain reasonable space utilization: at least 50% space utilization

[0121] Reduce rebalancing operations: avoid frequent node merging and splitting

[0122] Minimum number of data entries for leaf nodes: Considerations include:

[0123] Storage space utilization efficiency

[0124] Query performance balance

[0125] Maintenance overhead.

[0126] S4: receiving a query request, and extracting a target time period and a target parameter range from the query request, wherein the target parameter range includes a target temperature range, a target pressure range, and a target purity range;

[0127] A query request is received, the query request is encapsulated in JSON format and transmitted to the query processing module through a secure channel; the query request includes a query identification number, a request timestamp, a query priority, a target time period, a target parameter range and a query tolerance value; the target time period is represented by a start timestamp and an end timestamp, and the timestamp is accurate to milliseconds; the target parameter range includes a target temperature range, a target pressure range and a target purity range, wherein the target temperature range is defined by an upper temperature limit and a lower temperature limit, and the setting of the upper temperature limit and the lower temperature limit must meet the actual process parameter constraints; the target pressure range is defined by an upper pressure limit and a lower pressure limit, and the upper pressure limit shall not exceed ninety percent of the rated pressure of the equipment; the target purity range is defined by an upper purity limit and a lower purity limit, and the setting of the purity range must meet the product quality standard requirements.

[0128] The query request is subjected to a validity check and standardized processing, and a query context environment is further established, wherein the query context environment includes a query identification number, a standardized time interval, a standardized parameter range, a query status flag, an intermediate result cache area, and query statistical information; the query status flag is used to mark various stages of query processing, including states such as parameter extraction completion, index retrieval start, and data acquisition completion; the intermediate result cache area is used to store temporary data during the query process; the query statistical information includes performance indicators such as processing time, number of accessed nodes, and number of returned results.

[0129] The query features are recorded in the query context environment, including time interval features and parameter range features; the time interval features include information such as interval length, interval position, whether it spans multiple time series data blocks, etc.; the parameter range features include information such as range width, range center point, related constraints between parameters, etc.; the query features are used for subsequent query optimization and index tuning.

[0130] S5: In the query index tree, by comparing the target parameter range with the node feature value, locate the time series data block that meets the conditions, and extract the corresponding original data from the time series data block as the query result.

[0131] A multi-stage search is performed in the query index tree, first starting from the root node, and the target parameter range is compared with the node characteristic value; the comparison process first checks whether the parameter boundary of the node intersects with the target range, specifically including whether the maximum temperature parameter of the comparison node is greater than the lower limit of the target temperature range and whether the minimum temperature parameter is less than the upper limit of the target temperature range, whether the average pressure parameter of the comparison node is between the upper and lower limits of the target pressure range, and whether the average purity parameter of the comparison node is between the upper and lower limits of the target purity range; when the characteristic value of the node meets any of the above conditions, the node is added to the candidate node set.

[0132] The nodes in the candidate node set are prioritized, and the priority calculation takes into account factors such as the matching degree between the node characteristic value and the target parameter range, the node's historical access frequency, and the node's data update time. The matching degree is determined by calculating the degree of overlap between the characteristic value and the target range. The greater the overlap, the higher the priority. The historical access frequency is obtained by counting the node's access counter. The higher the access frequency, the higher the priority. The data update time is determined by recording the timestamp of the node's last data change. The newer the data, the higher the priority.

[0133] The child nodes of the candidate node are visited one by one in the order of priority, and the feature value comparison and priority sorting process is repeated on the child nodes until the leaf node is reached; a dynamic pruning threshold is maintained during the visit process, and when the matching degree of the node is lower than the threshold, the node is deleted from the candidate set; the dynamic pruning threshold is dynamically adjusted according to the matching results that have been found, and the threshold is increased when a high matching result is found to speed up the retrieval speed.

[0134] When a leaf node is reached, the time series data block pointed to by the leaf node is obtained, and the original data in the time series data block is accurately matched with the target parameter range; the accurate matching process checks each data point in the time series data block to ensure that the temperature parameter is within the target temperature range, the pressure parameter is within the target pressure range, and the purity parameter is within the target purity range; the data points that meet all conditions are extracted to form a query result set.

[0135] Example 2

[0136] Based on the first embodiment, this embodiment further provides a precious metal refining data query system, including:

[0137] The acquisition module is used to acquire the operating parameters in the precious metal refining process to form raw data, and divide the raw data into time series according to the refining time to generate time series data blocks;

[0138] A retrieval tree construction module, used to construct a query index tree based on the time series data block, each node of the query index tree stores a feature value of a time series data block;

[0139] An extraction module, used for receiving a query request and extracting a target time period and a target parameter range from the query request;

[0140] The query module is used to locate the time series data block that meets the conditions in the query index tree by comparing the target parameter range with the node feature value, and extract the corresponding original data from the time series data block as the query result.

[0141] This embodiment also provides a computer device suitable for the method of querying precious metal refining data, including a memory and a processor; the memory is used to store computer executable instructions, and the processor is used to execute computer executable instructions to implement the method of querying precious metal refining data proposed in the above embodiment.

[0142] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be achieved through WIFI, an operator network, NFC (near field communication) or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a key, trackball or touchpad provided on the housing of the computer device, or an external keyboard, touchpad or mouse, etc.

[0143] This embodiment also provides a storage medium on which a computer program is stored. When the program is executed by a processor, the method for querying precious metal refining data as proposed in the above embodiment is implemented.

[0144] The storage medium proposed in this embodiment and the data storage method proposed in the above embodiment belong to the same inventive concept. The technical details not fully described in this embodiment can be found in the above embodiment, and this embodiment has the same beneficial effects as the above embodiment.

[0145] Example 3

[0146] On the basis of the previous two embodiments, this embodiment provides a method for querying precious metal refining data. In order to verify the beneficial effects of the present invention, scientific demonstration is carried out through economic benefit calculation and simulation experiments.

[0147] In the process of platinum refining, data management and efficient retrieval have always been the key technical bottlenecks restricting production efficiency. In order to verify the innovation and practicality of the method of the present invention, a comprehensive performance comparison experiment was designed in a high-end precious metal smelting laboratory. The experiment selected the continuous production data of platinum refining from January 15 to February 15, 2024 as the research object. Through the independently developed data acquisition and query system, a comprehensive monitoring network including 6 temperature sensors (distributed at the feed end, the middle of the reaction chamber, the top of the reaction chamber, the bottom of the reaction chamber, the front section of the discharge, and the discharge end), 4 pressure sensors (front side of the top of the reaction chamber, the back side of the top, the front side of the bottom, and the back side of the bottom) and 2 spectral purity detection modules were deployed. According to the preset plan, the temperature sensor collects data once per second, the pressure sensor collects data once per second, and the spectral purity analyzer collects data every 5 seconds.

[0148] The data preprocessing process adopts the innovative method proposed in this invention, including: (1) data segmentation based on 30-minute time windows; (2) weighted average of multi-sensor data; (3) time-attenuated linear interpolation for purity parameters; (4) time series correlation evaluation; (5) sliding window abnormal data correction. The query index tree is constructed using a 16-tree structure. Each node stores the characteristic values ​​of temperature, pressure, and purity, and maintains the structural stability of the tree through a dynamic balance mechanism. The query performance test designed multiple sets of parameter range query scenarios with different complexities, focusing on evaluating key indicators such as query response time, number of node visits, and result accuracy. The experimental related data are shown in Table 1 below:

[0149] Table 1 Precious metal refining data query performance comparison experimental data

[0150]

[0151] Through in-depth analysis of experimental data, the method of the present invention shows significant performance advantages. In different query scenarios, compared with traditional sequential scanning and relational database query methods, the index tree query method of the present invention has achieved revolutionary improvements in key indicators such as response time, number of node accesses, and query result accuracy. Specifically, for complex large-scale query scenarios across time series data blocks, the response time of traditional methods is as high as 18236 milliseconds, while the method of the present invention only takes 512 milliseconds, with a performance improvement of more than 35 times; the number of node accesses is reduced from 45123 to 1256, a decrease of up to 97.2%.

[0152] More importantly, the method of the present invention not only significantly improves the query efficiency, but also ensures the high accuracy of the query results through innovative technologies such as multi-dimensional eigenvalue indexing, dynamic pruning and priority sorting. Even in complex query scenarios with large spans of temperature, pressure and purity parameters, the accuracy of the query results remains above 95%, fully demonstrating the robustness and practicality of the method. The dynamically balanced query index tree structure enables the system to adaptively adjust the node distribution, effectively overcoming the performance bottleneck of traditional indexing methods when processing large-scale heterogeneous data.

[0153] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for querying precious metal refining data, characterized in that: include Collecting the operating parameters in the precious metal refining process to form raw data, and dividing the raw data into time series according to the refining time to generate time series data blocks; Building a query index tree based on the time series data block, each node of the query index tree stores a feature value of a time series data block; receiving a query request, and extracting a target time period and a target parameter range from the query request; In the query index tree, by comparing the target parameter range with the node feature value, the time series data block that meets the conditions is located, and the corresponding original data is extracted from the time series data block as the query result.

2. The method for querying precious metal refining data according to claim 1, characterized in that: The operating parameters include temperature parameters, pressure parameters, and purity parameters; Generating the time series data block includes: Dividing the original data into time series according to the extraction time, and calculating the purity value of the intermediate moment according to the purity parameter in the original data to obtain a preliminary time series data block; Performing time series correlation evaluation on the preliminary time series data block and calculating a quality index; Optimizing the preliminary time series data block based on the quality indicator to obtain the time series data block; Storing the time series data block in a time series table and creating a composite index; The optimization is shown in the following formula: R′(t k )=R(t i )·w1+R(t j )·w2 Among them, t k is the time to be interpolated, R′(t k ) is t i and t j The purity value at the middle moment of time, t i and t j For two adjacent purity collection moments, the weight coefficient is calculated as follows: Where λ is the time decay coefficient.

3. The method for querying precious metal refining data according to claim 2, characterized in that: The characteristic values ​​include the maximum temperature parameter, the minimum temperature parameter, the average pressure parameter, the average purity parameter, the parameter fluctuation range value and the timestamp information of the time series data block; The maximum temperature parameter is selected from all the temperature sensor data in the time series data block to obtain the maximum value, and the minimum temperature parameter is selected from all the temperature sensor data in the time series data block to obtain the minimum value; The average pressure parameter is obtained by calculating the arithmetic mean of the data collected by all pressure sensors in the time series data block within the time period, and the average purity parameter is obtained by calculating the weighted average of all purity data in the time series data block, and the weighting coefficient decays exponentially as the time interval increases; The parameter fluctuation range value is obtained by calculating the square root of the sum of squares of the differences between the maximum and minimum values ​​of the temperature parameter, the pressure parameter and the purity parameter.

4. The method for querying precious metal refining data according to claim 3, characterized in that: In the query index tree, each non-leaf node additionally stores subtree range information, including upper and lower bounds of the characteristic values ​​of all nodes in the subtree, for quickly determining whether the query condition is relevant to the subtree; The non-leaf nodes establish neighboring pointers between nodes through the similarity relationship of the eigenvalues; the similarity is calculated by the weighted distance of each component of the eigenvalue; the weight coefficient is dynamically adjusted based on the historical query frequency; The leaf nodes store storage addresses pointing to actual time series data blocks, and a bidirectional linked list is established between adjacent leaf nodes.

5. The method for querying precious metal refining data according to claim 4, characterized in that: The bidirectional linked list established between adjacent leaf nodes is expressed by the following formula: L(N)={N i |(F(N),F(N i ))≤δ} in, δ is the proximity threshold, which is adaptively adjusted based on historical query statistics: Among them, μ is the mean of historical distance, σ is the standard deviation, and Q is the number of queries.

6. The method for querying precious metal refining data according to claim 5, characterized in that: The query index tree also includes: When the number of child nodes in a node exceeds the maximum limit, the node is split according to the distribution of its eigenvalues, and the split point that makes the difference in the eigenvalues ​​of the child nodes after the split is selected is selected; when the number of child nodes in a node is lower than the minimum limit, the adjacent nodes with the closest eigenvalues ​​are merged.

7. The method for querying precious metal refining data according to claim 6, characterized in that: The time series data blocks that meet the positioning conditions include: The target parameter range is compared with the node characteristic value; the comparison process first checks whether the parameter boundary of the node intersects with the target range, specifically including whether the maximum temperature parameter of the comparison node is greater than the lower limit of the target temperature range and whether the minimum temperature parameter is less than the upper limit of the target temperature range, whether the average pressure parameter of the comparison node is between the upper and lower limits of the target pressure range, and whether the average purity parameter of the comparison node is between the upper and lower limits of the target purity range; when the characteristic value of the node meets any of the above conditions, the node is added to the candidate node set.

8. A precious metal refining data query system, based on the precious metal refining data query method according to any one of claims 1 to 7, characterized in that: include, The acquisition module is used to acquire the operating parameters in the precious metal refining process to form raw data, and divide the raw data into time series according to the refining time to generate time series data blocks; A retrieval tree construction module, used to construct a query index tree based on the time series data block, each node of the query index tree stores a feature value of a time series data block; An extraction module, configured to receive a query request and extract a target time period and a target parameter range from the query request; The query module is used to locate the time series data block that meets the conditions in the query index tree by comparing the target parameter range with the node feature value, and extract the corresponding original data from the time series data block as the query result.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for querying precious metal refining data according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for querying precious metal refining data according to any one of claims 1 to 7 are implemented.

Images