Mine ecological problem identification method and system based on artificial intelligence
Through the artificial intelligence-based method, the feature analysis and sharing coefficient calculation of mining ecological problems data is solved, and the problems of low accuracy and high cost of identification of mining ecological problems are achieved, achieving more efficient identification results.
Patent Information
- Application Number
- CN202411159181.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-22
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2044-08-22
AI Technical Summary
Identification of mining ecological problems requires a lot of time and labor costs, and due to the complex terrain and difficult survey, the identification accuracy is not high.
Using an artificial intelligence-based method, key features are analyzed by analyzing the data of the to-be-processed mining ecological problems, splitting them into local data, and sharing coefficient calculations are performed with reference data, determining the pending data, and finally determining the target data, improving identification accuracy and efficiency.
It improves the accuracy and efficiency of mining ecological problems, narrows the scope of identification, can more accurately capture local problem characteristics, and reduces the cost of surveying.
Smart Images

Figure CN118673416B_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the technical field of problem identification, and more specifically, to a method and system for identifying mine ecological problems based on artificial intelligence. Background Art
[0002] At present, before the mine ecology is managed, technical surveys are required, and then management can be carried out. The survey requires a lot of time and manpower costs. In addition, due to the complex terrain of the mine, some places are difficult to reach due to equipment and manpower. The time cost and labor cost of the survey exceed the cost of general projects, and the mine ecological problems obtained are not complete and accurate enough. Therefore, a mine ecological identification technology is urgently needed to improve the accuracy of mine ecological problem identification. Summary of the invention
[0003] In order to improve the technical problems existing in the relevant technologies, this application provides a method and system for identifying mine ecological problems based on artificial intelligence.
[0004] In the first aspect, a method for identifying mine ecological problems based on artificial intelligence is provided, comprising:
[0005] Obtain data on ecological issues in mines to be addressed;
[0006] Performing key feature analysis on the to-be-processed mine ecological problem data to obtain a number of local mine ecological problem data with significant features;
[0007] For each local mine ecological problem data, obtain a plurality of reference mine ecological problem data corresponding to the local mine ecological problem data, and a first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data;
[0008] In combination with the first sharing coefficient, determining at least one pending mine ecological problem data associated with the local mine ecological problem data among the plurality of reference mine ecological problem data, and determining a second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data;
[0009] In combination with the second sharing coefficient, target mine ecological problem data are determined from the pending mine ecological problem data.
[0010] In the present application, the obtaining of a plurality of reference mine ecological problem data corresponding to the local mine ecological problem data includes:
[0011] Extracting features of the local mine ecological problem data to obtain mine ecological problem data features;
[0012] An original feature vector associated with the mine ecological problem data feature is obtained from a plurality of pre-set original feature vectors and determined as a reference feature vector; the plurality of original feature vectors are extracted from a plurality of pre-set original mine ecological problem data;
[0013] Performing mine ecological problem data tracing processing on the reference feature vector to obtain original mine ecological problem data corresponding to the reference feature vector;
[0014] The original mine ecological problem data corresponding to the reference feature vector is determined as the reference mine ecological problem data.
[0015] In the present application, the original feature vector associated with the mine ecological problem data feature obtained from a plurality of pre-set original feature vectors is determined as a reference feature vector, including:
[0016] Obtaining undetermined sharing coefficients between the mine ecological problem data features and the plurality of original feature vectors;
[0017] Numbering the plurality of original feature vectors according to the order in which the to-be-determined sharing coefficients are set to obtain numbered original feature vectors;
[0018] Determine the first M original feature vectors among the numbered original feature vectors as original feature vectors associated with the mine ecological problem data features, where M is a positive integer;
[0019] The original feature vector associated with the mine ecological problem data feature is determined as the reference feature vector.
[0020] In the present application, the obtaining of the undetermined sharing coefficients between the mine ecological problem data features and the several original feature vectors respectively includes:
[0021] For each original feature vector, the original feature vector is numerically described to obtain a local AI vector space corresponding to the original feature vector, wherein the local AI vector space includes a plurality of local vectors having the same direction, and the direction of the local vector is smaller than the direction of the original feature vector;
[0022] The local AI vector space is divided by a preset division network to obtain a division benchmark of the local AI vector space;
[0023] Compressing the local AI vector space according to the division criteria to obtain category compression corresponding to the local AI vector space;
[0024] Through the category compression and the mine ecological problem data characteristics, the undetermined sharing coefficient between the mine ecological problem data characteristics and the original feature vector is determined.
[0025] In the present application, the step of obtaining the first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data includes:
[0026] For the reference mine ecological problem data, identifying the vector number of reference feature vectors corresponding to the reference mine ecological problem data;
[0027] If the number of the vectors is one, a reference feature vector corresponding to the reference mine ecological problem data and a mine ecological problem data feature corresponding to the local mine ecological problem data are used to calculate a sharing coefficient to obtain a first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data;
[0028] If the number of vectors is several, the sharing coefficients of several reference feature vectors corresponding to the reference mine ecological problem data and the mine ecological problem data features corresponding to the local mine ecological problem data are calculated respectively to obtain several original sharing coefficients, and the several original sharing coefficients are spliced to obtain the first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data.
[0029] In the present application, the combining of the first sharing coefficient to determine at least one pending mine ecological problem data associated with the local mine ecological problem data among the plurality of reference mine ecological problem data includes:
[0030] Numbering the plurality of reference mine ecological problem data in combination with the order set by the first sharing coefficient to obtain numbered reference mine ecological problem data;
[0031] The first a reference mine ecological problem data among the numbered reference mine ecological problem data are determined as the pending mine ecological problem data, wherein a is a positive integer.
[0032] In the present application, the determination of the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data in combination with the first sharing coefficient includes:
[0033] For each pending mine ecological problem data, the first sharing coefficients of the pending mine ecological problem data corresponding to each local mine ecological problem data are spliced to obtain the second sharing coefficients of the pending mine ecological problem data corresponding to the pending mine ecological problem data.
[0034] In the present application, the first sharing coefficients of the pending mine ecological problem data corresponding to each local mine ecological problem data are spliced to obtain the second sharing coefficients of the pending mine ecological problem data corresponding to the pending mine ecological problem data, including:
[0035] Obtaining the splicing weights of each local mine ecological problem data corresponding to the pending mine ecological problem data and the number of several local mine ecological problem data;
[0036] By using the splicing weight, a first function is performed on the first sharing coefficient of each local mine ecological problem data corresponding to the pending mine ecological problem data to obtain a splicing sharing coefficient;
[0037] The comparison result of the splicing sharing coefficient and the number of the plurality of local mine ecological problem data is calculated to obtain the second sharing coefficient.
[0038] In this application, the key feature analysis processing is performed on the mine ecological problem data to be processed to obtain several local mine ecological problem data with significant features, including:
[0039] Identify attribute nodes in the mine ecological problem data to be processed, where the attribute nodes are periodic nodes in which characteristics of the mine ecological problem data to be processed change;
[0040] The key feature analysis processing of the mine ecological problem data to be processed is performed through the attribute nodes to obtain a number of local mine ecological problem data with significant features.
[0041] In the present application, the identifying of attribute nodes in the mine ecological problem data to be processed includes:
[0042] Performing feature extraction on the mine ecological problem data to be processed to obtain feature numbers, wherein each feature in the feature number corresponds to a period node in the mine ecological problem data to be processed;
[0043] Dividing the feature number into a first sub-feature number and a second sub-feature number;
[0044] Performing queue calculation on the first sub-feature number, the second sub-feature number, and the feature number respectively to obtain a first queue corresponding to the first sub-feature number, a second queue corresponding to the second sub-feature number, and a third queue corresponding to the feature number;
[0045] A maximum weight value is determined according to the first queue, the second queue, and the third queue; if the maximum weight value meets the preset requirements, the period node corresponding to the last feature in the first sub-feature number is determined as the attribute node.
[0046] In the present application, the step of combining the second sharing coefficient to determine the target mine ecological problem data from the pending mine ecological problem data includes:
[0047] The undetermined mine ecological problem data having the largest second sharing coefficient among all the undetermined mine ecological problem data associated with the plurality of reference mine ecological problem data is determined as the target mine ecological problem data.
[0048] In a second aspect, a mine ecological problem identification system based on artificial intelligence is provided, comprising a processor and a memory communicating with each other, wherein the processor is used to read a computer program from the memory and execute it to implement the above method.
[0049] The artificial intelligence-based mine ecological problem identification method and system provided in the embodiment of the present application can, after obtaining the mine ecological problem data to be processed, split the mine ecological problem data to be processed into several local mine ecological problem data with significant characteristics, and then obtain several reference mine ecological problem data corresponding to the local mine ecological problem data for each local mine ecological problem data, as well as a first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data; then, based on the first sharing coefficient, determine at least one pending mine ecological problem data associated with the local mine ecological problem data among the several reference mine ecological problem data, and determine a second sharing coefficient between the pending mine ecological problem data and the pending mine ecological problem data; finally, based on the second sharing coefficient, determine the target mine ecological problem data from the pending mine ecological problem data. In the embodiment of the present application, for each local mine ecological problem data of the mine ecological problem data to be processed, some similar reference mine ecological problem data can be preliminarily recalled according to the local mine ecological problem data, and then according to the first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data, similar pending mine ecological problem data are recalled from several reference mine ecological problem data, which improves the recall efficiency on the one hand, and on the other hand, because the local mine ecological problem data has significant characteristics, the characteristics of the local mine ecological problem data can be captured more accurately, thereby improving the recall accuracy. Then, according to the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data, the target mine ecological problem data is screened out from the pending mine ecological problem data, which can narrow the recognition range on the one hand, and on the other hand, the target mine ecological problem data that is more similar to the pending mine ecological problem data as a whole can be screened out from the pending mine ecological problem data. Thus, the accuracy of mine ecological problem data identification is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0050] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.
[0051] Figure 1 A flowchart of a method for identifying mine ecological problems based on artificial intelligence provided in an embodiment of the present application. DETAILED DESCRIPTION
[0052] In order to better understand the above technical scheme, the technical scheme of the present application is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical scheme of the present application, rather than limitations on the technical scheme of the present application. In the absence of conflict, the embodiments of the present application and the technical features in the embodiments can be combined with each other.
[0053] See also Figure 1 , shows a method for identifying mine ecological problems based on artificial intelligence, which may include the technical solutions described in the following steps 101-105.
[0054] 101. Obtain data on ecological problems in mines to be addressed.
[0055] For example, before conducting ecological governance of mines, it is necessary to identify the ecological situation of the area and what pollution in the area has an impact on the ecology. For example, in areas with high heavy metal content, less vegetation can grow, there are few species here, and the ecology is seriously unstable.
[0056] 102. The key features of the mine ecological problem data to be processed are analyzed and processed to obtain several local mine ecological problem data with significant characteristics.
[0057] Among them, key feature analysis processing refers to the decomposition and extraction of different feature parts in the mine ecological problem data.
[0058] Among them, the local mine ecological problem data has significant characteristics indicating that the local mine ecological problem data fragments are relatively independent in characteristics, that is, each local mine ecological problem data each contains specific characteristic information or content, and does not need to rely on the entire mine ecological problem data to be processed or other local mine ecological problem data to understand its meaning.
[0059] In a possible implementation, when the mine ecological problem data to be processed do not have characteristics, the mine ecological problem data to be processed can also be analyzed for key features through the characteristics of the mine ecological problem data to obtain several local mine ecological problem data with significant characteristics, that is, different local mine ecological problem data among the several local mine ecological problem data after splitting can have different mine ecological problem data characteristics.
[0060] In a possible implementation example, in step 102, the specific implementation of the step of "performing key feature analysis processing on the mine ecological problem data to be processed to obtain a number of local mine ecological problem data with significant features" may include:
[0061] A1. Identify attribute nodes in the mine ecological problem data to be processed, where the attribute nodes are periodic nodes where characteristics of the mine ecological problem data to be processed change.
[0062] In one possible implementation, the attribute nodes in the mine ecological problem data can be identified by a pre-trained attribute node prediction thread, wherein the attribute node prediction thread can output the corresponding attribute nodes according to the input mine ecological problem data to be processed. Exemplarily, when training the attribute node prediction thread, a large number of mine ecological problem data examples can be prepared first, and then the attribute nodes in the mine ecological problem data examples are marked, and then the mine ecological problem data examples with the attribute nodes marked are input into the original thread for thread training, and the goal is to enable the original thread to accurately identify the attribute nodes in the mine ecological problem data to obtain the attribute node prediction thread. Optionally, the original thread can select a thread suitable for signal processing of mine ecological problem data, such as a recurrent neural network (RNN), a convolutional neural network (CNN), or a deep neural network (DNN). These threads can be used to learn patterns and features in mine ecological problem data. Optionally, a supervised learning method can be used in the training process of the original thread to adjust the thread parameters through training data.
[0063] In an alternative embodiment, in step A1, the specific implementation of the step of "identifying attribute nodes in the mine ecological problem data to be processed" may include:
[0064] A11a. Extract features from the mine ecological problem data to be processed to obtain feature numbers, where each feature in the feature number corresponds to a period node in the mine ecological problem data to be processed.
[0065] A12a. Divide the feature number into a first sub-feature number and a second sub-feature number.
[0066] A13a. Perform queue calculation on the first sub-feature number, the second sub-feature number and the feature number respectively to obtain a first queue corresponding to the first sub-feature number, a second queue corresponding to the second sub-feature number and a third queue corresponding to the feature number.
[0067] A14a. Determine a maximum weight value according to the first queue, the second queue, and the third queue; if the maximum weight value meets the preset requirements, determine the period node corresponding to the last feature in the first sub-feature number as an attribute node.
[0068] In an alternative embodiment, in step A1, the specific implementation of the step of "identifying attribute nodes in the mine ecological problem data to be processed" may include:
[0069] A11b. Extracting mine pollution characteristics from the mine ecological problem data to be processed to obtain a number of mine pollution characteristics, each of which corresponds to a periodic node in the mine ecological problem data to be processed, and the several mine pollution characteristics are arranged in order from the earliest to the latest periodic nodes.
[0070] The number and direction of mine pollution characteristics can be customized according to actual needs and are not limited here.
[0071] A12b. Determine at least one target mine pollution feature from a plurality of mine pollution features. The target mine pollution feature is used to segment a plurality of mine pollution features into a plurality of mine pollution feature sets, and the mine pollution features in each mine pollution feature set conform to the Gaussian distribution.
[0072] In a possible implementation example, taking a number of mine pollution characteristics constituting a mine pollution characteristic number as an example, a specific implementation method of determining at least one target mine pollution characteristic from a number of mine pollution characteristics may include: selecting a mine pollution characteristic from the mine pollution characteristic number and determining it as a reference mine pollution characteristic.
[0073] Then, the reference mine pollution feature and the mine pollution feature numbered before the reference mine pollution feature are determined as a set of mine pollution features to be tested, and it is determined whether the set of mine pollution features to be tested conforms to the Gaussian distribution.
[0074] If it meets the requirements, the reference mine pollution characteristic can be determined as the target mine pollution characteristic, and the reference mine pollution characteristic and the mine pollution characteristic that is located before the reference mine pollution characteristic in the mine pollution characteristic number can be deleted from the mine pollution characteristic number to obtain a new mine pollution characteristic number, and based on the new mine pollution characteristic number, the above-mentioned step of "selecting a mine pollution characteristic from the mine pollution characteristic number and determining it as the reference mine pollution characteristic" is returned to be executed until there is no set of mine pollution characteristics to be tested that conforms to the Gaussian distribution in the new mine pollution characteristic number.
[0075] If it does not meet the requirements, the adjacent mine pollution characteristics in the mine pollution characteristic number that is located after the reference mine pollution characteristics will be determined as new reference mine pollution characteristics, and based on the new reference mine pollution characteristics, the above-mentioned step of "determining the reference mine pollution characteristics and the mine pollution characteristics in the mine pollution characteristic number that are located before the reference mine pollution characteristics as the mine pollution characteristic set to be tested, and judging whether the mine pollution characteristic set to be tested meets the Gaussian distribution" will be returned to be executed until the mine pollution characteristic set to be tested meets the Gaussian distribution.
[0076] In this embodiment, the mine pollution characteristics are screened out one by one from several mine pollution characteristics to determine as reference mine pollution characteristics, and the mine pollution characteristic number to be tested is constructed based on the reference mine pollution characteristics. Then, it is identified whether the mine pollution characteristic number to be tested accurately conforms to the Gaussian distribution to determine whether the reference mine pollution characteristic is the target mine pollution characteristic. In this way, the target mine pollution characteristic can be accurately determined from several mine pollution characteristics, avoiding the omission of the target mine pollution characteristic.
[0077] In an alternative embodiment, in step A12b, the specific implementation of the step of "determining at least one target mine pollution feature from a plurality of mine pollution features" may include:
[0078] A121b. Determine an undetermined mine pollution characteristic from several mine pollution characteristics.
[0079] In a possible implementation, a mine pollution feature may be randomly selected from a number of mine pollution features to be determined as a pending mine pollution feature, wherein when selecting the pending mine pollution feature, the first mine pollution feature and the last mine pollution feature among the several mine pollution features may be ignored. For example, taking a number of mine pollution features as an example of constituting a mine pollution feature number, when selecting the pending mine pollution feature, the mine pollution feature near the middle of the mine pollution feature number may be preferentially selected to be determined as the pending mine pollution feature.
[0080] A122b. According to the pending mine pollution characteristics, several mine pollution characteristics are divided into a first mine pollution characteristic set and a second mine pollution characteristic set; the first mine pollution characteristic set includes the pending mine pollution characteristics and the mine pollution characteristics whose periodic nodes are before the pending mine pollution characteristics; the second mine pollution characteristic set includes the mine pollution characteristics whose periodic nodes are after the pending mine pollution characteristics.
[0081] A123b. Determine an assessment value corresponding to the pending mine pollution feature based on the first mine pollution feature set, the second mine pollution feature set and several mine pollution features; the assessment value represents the possibility that the first mine pollution feature set and the second mine pollution feature set both conform to the Gaussian distribution.
[0082] Specifically, in step A123b, according to the first mine pollution feature set, the second mine pollution feature set and a plurality of mine pollution features, a specific implementation method for determining the evaluation value corresponding to the pending mine pollution feature may include:
[0083] A1231b. Based on the first mine pollution feature set, the second mine pollution feature set and several mine pollution features, determine the maximum weight value corresponding to the undetermined mine pollution feature, and the maximum weight value represents the original possibility that the first mine pollution feature set and the second mine pollution feature set both conform to the Gaussian distribution.
[0084] In a possible implementation, the first mine pollution feature set, the second mine pollution feature set and several mine pollution features can be calculated respectively by the queue calculation formula to obtain the first queue corresponding to the first mine pollution feature set, the second queue corresponding to the second mine pollution feature set, and the third queue corresponding to the several mine pollution features.
[0085] Then, determinants are calculated for the first queue, the second queue, and the third queue respectively to obtain a first determinant corresponding to the first queue, a second determinant corresponding to the second queue, and a third determinant corresponding to the third queue.
[0086] Then, the product of the logarithm of the first determinant and the number of mine pollution features in the first mine pollution feature set is determined as the first calculated value, the product of the logarithm of the second determinant and the number of mine pollution features in the second mine pollution feature set is determined as the second calculated value, and the product of the logarithm of the third determinant and the number of mine pollution features in several mine pollution features is determined as the third calculated value.
[0087] Finally, the maximum weight value corresponding to the pending mine pollution feature can be obtained by subtracting the first calculated value from the third calculated value and then subtracting the second calculated value.
[0088] A1232b. Obtain the feature numbers and feature directions of several mine pollution features, and determine abnormal values based on the feature numbers and feature directions. The abnormal values are used to debug the original possibilities.
[0089] A1233b. Determine the evaluation value based on the maximum weight value and the outlier value.
[0090] In a possible implementation, the weight corresponding to the outlier may be obtained, and then the product of the weight and the outlier may be calculated to obtain a fourth calculated value, and then the fourth calculated value may be subtracted from the maximum weight value to obtain an evaluation value.
[0091] A124b. If the evaluation value is the maximum value in the corresponding value range, the pending mine pollution characteristic is determined as the target mine pollution characteristic.
[0092] Among them, when the evaluation value is the maximum value in the corresponding value range of the evaluation value, it indicates that the possibility that the first mine pollution feature set and the second mine pollution feature set both conform to the Gaussian distribution is the greatest, that is, the possibility that the periodic node corresponding to the undetermined mine pollution feature that divides the first mine pollution feature set and the second mine pollution feature set is an attribute node is the greatest.
[0093] In this embodiment, by converting the process of finding attribute nodes into the process of selecting the following two hypothetical threads, and then using the Bayesian information criterion to select threads, the target mine pollution characteristics corresponding to the attribute nodes can be quickly found among several mine pollution characteristics, thereby improving the efficiency of determining the target mine pollution characteristics.
[0094] A13b. Determine the periodic nodes corresponding to the pollution characteristics of the target mine as attribute nodes.
[0095] A2. Based on the attribute nodes, key feature analysis is performed on the mine ecological problem data to be processed to obtain several local mine ecological problem data with significant characteristics.
[0096] 103. For each local mine ecological problem data, obtain a number of reference mine ecological problem data corresponding to the local mine ecological problem data, and a first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data.
[0097] The reference mine ecological problem data may be benchmark mine ecological problem data for comparison with the mine ecological problem data to be processed. For example, compared with the mine ecological problem data to be processed, the benchmark mine ecological problem data may be known, complete, and noise-free mine ecological problem data. Optionally, the reference mine ecological problem data may be pre-stored in the mine ecological problem database.
[0098] Among them, each local mine ecological problem data can establish a mapping relationship with a number of reference mine ecological problem data in advance, and the reference mine ecological problem data corresponding to the local mine ecological problem data can be mine ecological problem data with similar characteristics to the local mine ecological problem data.
[0099] In a possible implementation example, in step 103, the specific implementation method of obtaining a number of reference mine ecological problem data corresponding to the local mine ecological problem data may include: S1, extracting features of the local mine ecological problem data to obtain mine ecological problem data features.
[0100] In a possible implementation embodiment, the local mine ecological problem data can be feature extracted by a pre-set mine ecological problem data feature extraction unit. Optionally, the extracted mine ecological problem data features can be associated with the above-mentioned reference feature vector in the feature direction to facilitate the calculation of the sharing coefficient between the two. Optionally, the pre-set mine ecological problem data feature extraction unit can be CNN14 (a feature extraction network), which mainly includes 12 convolutional layers and 2 fully connected layers. Among them, the pre-set mine ecological problem data feature extraction unit can also be other feature extraction networks (such as CNN10, CNN6, etc.), and the specific feature extraction unit can be set according to actual needs, which is not limited here.
[0101] S2. An original feature vector associated with the characteristics of the mine ecological problem data is obtained from a number of original feature vectors set in advance and determined as a reference feature vector; the several original feature vectors are extracted from a number of original mine ecological problem data set in advance.
[0102] In a possible implementation example, a specific implementation of step S2 may include: S21, obtaining undetermined sharing coefficients between the mine ecological problem data features and a plurality of original feature vectors.
[0103] In a possible implementation example, the mine ecological problem data feature can be directly compared with the sharing coefficient of each original feature vector in a plurality of original feature vectors to obtain the undetermined sharing coefficient.
[0104] In an alternative embodiment, the reference feature vector may be obtained by a vector recall method (such as brute force enumeration), so the specific implementation of obtaining the undetermined sharing coefficients between the mine ecological problem data features and a plurality of original feature vectors in S21 may include:
[0105] S211. For each original feature vector, the original feature vector is numerically described to obtain a local AI vector space corresponding to the original feature vector. The local AI vector space includes a number of local vectors with the same direction, and the direction of the local vector is smaller than the direction of the original feature vector.
[0106] S212, partitioning the local AI vector space through a preset partitioning network to obtain a partitioning benchmark for the local AI vector space.
[0107] Exemplarily, partitioning methods include: a-MEANS algorithm, a-MEDOIDS algorithm, and CLARANS algorithm.
[0108] S213 . Compress the local AI vector space based on the division criterion to obtain category compression corresponding to the local AI vector space.
[0109] S214. Based on the category compression and the mine ecological problem data characteristics, determine the undetermined sharing coefficient between the mine ecological problem data characteristics and the original feature vector.
[0110] Optionally, category compression can be used to preferentially select the division criteria and mine ecological problem data characteristics in the local AI vector space for sharing coefficient calculation, and then all vectors in the local AI vector space are compared to obtain vectors similar to the mine ecological problem data characteristics, wherein the sharing coefficient can be calculated by means of vector inner product.
[0111] Optionally, the mine ecological problem data features can also be divided into local vectors of mine ecological problem data, and then the sum of the differences between each local vector of mine ecological problem data and the corresponding numerical description vector of each local AI vector space is calculated to obtain the undetermined sharing coefficient.
[0112] S22. Numbering a plurality of original feature vectors according to the order in which the to-be-determined sharing coefficients are set, to obtain numbered original feature vectors.
[0113] S23. Determine the first M original feature vectors among the numbered original feature vectors as original feature vectors associated with the mine ecological problem data features, where M is a positive integer.
[0114] S24. Determine the original feature vector associated with the mine ecological problem data features as a reference feature vector.
[0115] S3. Perform mining ecological problem data tracing processing on the reference feature vector to obtain the original mining ecological problem data corresponding to the reference feature vector.
[0116] Among them, several original feature vectors set in advance can be stored in a preset vector library, and the preset vector library pre-stores several original feature vectors and a vector mapping relationship table. The vector mapping relationship table pre-records the original mine ecological problem data corresponding to each original feature vector in several original feature vectors, and the original feature vector is extracted from the original mine ecological problem data corresponding to it. Therefore, after determining the original mine ecological problem data, the original feature vector corresponding to the original mine ecological problem data can be found from the preset vector library according to the vector mapping relationship table. Among them, one original mine ecological problem data can correspond to one or several original feature vectors.
[0117] S4. Determine the original mine ecological problem data corresponding to the reference feature vector as the reference mine ecological problem data.
[0118] In a possible implementation example, in step 103, the specific implementation method of obtaining the first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data may include: for the reference mine ecological problem data, identifying the vector number of the reference feature vector corresponding to the reference mine ecological problem data.
[0119] If the number of vectors is one, the reference feature vector corresponding to the reference mine ecological problem data and the mine ecological problem data feature corresponding to the local mine ecological problem data are used to calculate the sharing coefficient to obtain the first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data.
[0120] If the number of vectors is several, the sharing coefficients of several reference feature vectors corresponding to the reference mine ecological problem data and the mine ecological problem data features corresponding to the local mine ecological problem data are calculated respectively to obtain several original sharing coefficients, and the several original sharing coefficients are spliced to obtain the first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data.
[0121] Exemplarily, the specific method of obtaining the first sharing coefficient may be as follows:
[0122] B1. Obtain reference feature vectors corresponding to each reference mine ecological problem data from the preset vector library.
[0123] The reference mine ecological problem data may correspond to one or several reference feature vectors.
[0124] B2. Extract features of local mine ecological problem data to obtain the characteristics of mine ecological problem data.
[0125] B3. For each reference mine ecological problem data, the reference mine ecological problem data is used to calculate the sharing coefficient of the reference feature vector and the mine ecological problem data characteristics, so as to obtain the first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data.
[0126] When the number of reference feature vectors corresponding to the reference mine ecological problem data is several, in step B3, the sharing coefficients of the reference feature vectors corresponding to the reference mine ecological problem data and the mine ecological problem data characteristics are calculated to obtain the first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data. The specific implementation method may include: calculating the sharing coefficients of the mine ecological problem data characteristics and several reference feature vectors corresponding to the reference mine ecological problem data respectively to obtain several original sharing coefficients.
[0127] Several original sharing coefficients are concatenated to obtain the first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data.
[0128] 104. According to the first sharing coefficient, determine at least one pending mine ecological problem data associated with the local mine ecological problem data among several reference mine ecological problem data, and determine the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data.
[0129] In a possible implementation, in step 104, based on the first sharing coefficient, the specific implementation method of determining at least one pending mine ecological problem data associated with the local mine ecological problem data among several reference mine ecological problem data may include: determining the reference mine ecological problem data among several reference mine ecological problem data whose first sharing coefficient is greater than or equal to the sharing coefficient threshold as pending mine ecological problem data, for example, the first sharing coefficients between the local mine ecological problem data x and the reference mine ecological problem data Z1, the reference mine ecological problem data Z2, and the reference mine ecological problem data Z3 are 0.5, 0.2, and 0.3 respectively. If the sharing coefficient threshold is 0.3, then the reference mine ecological problem data Z1 and the reference mine ecological problem data Z3 can be determined as the pending mine ecological problem data corresponding to the local mine ecological problem data x.
[0130] In an alternative embodiment, in step 104, based on the first sharing coefficient, the specific implementation method of determining at least one pending mine ecological problem data associated with the local mine ecological problem data among a number of reference mine ecological problem data may include: numbering the several reference mine ecological problem data according to the order set by the first sharing coefficient to obtain the numbered reference mine ecological problem data.
[0131] The first a reference mine ecological problem data among the numbered reference mine ecological problem data are determined as pending mine ecological problem data, where a is a positive integer.
[0132] Exemplarily, for example, a number of local mine ecological problem data include local mine ecological problem data x and local mine ecological problem data b, and the reference mine ecological problem data include reference mine ecological problem data Z1, reference mine ecological problem data Z2, and reference mine ecological problem data Z3. The first sharing coefficients between the local mine ecological problem data x and the reference mine ecological problem data Z1, reference mine ecological problem data Z2, and reference mine ecological problem data Z3 are 0.5, 0.2, and 0.3, respectively. The first sharing coefficients between the local mine ecological problem data b and the reference mine ecological problem data Z1, reference mine ecological problem data Z2, and reference mine ecological problem data Z3 are 0.6, 0.1, and 0.4, respectively. If a is 2, then the reference mine ecological problem data Z1 and the reference mine ecological problem data Z3 can be determined as the pending mine ecological problem data corresponding to the local mine ecological problem data x; the reference mine ecological problem data Z1 and the reference mine ecological problem data Z3 can be determined as the pending mine ecological problem data corresponding to the local mine ecological problem data x.
[0133] In a possible implementation example, the reference mine ecological problem data may also be directly determined as pending mine ecological problem data.
[0134] In a possible implementation example, in step 104, based on the first sharing coefficient, the specific implementation method of determining the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data may include: for each pending mine ecological problem data, the first sharing coefficients of the pending mine ecological problem data corresponding to each local mine ecological problem data are spliced to obtain the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data.
[0135] In a possible implementation example, the specific implementation method of the step of "splicing the first sharing coefficients of the pending mine ecological problem data corresponding to each local mine ecological problem data to obtain the second sharing coefficients of the pending mine ecological problem data corresponding to the pending mine ecological problem data" may include: obtaining the splicing weights of the pending mine ecological problem data corresponding to each local mine ecological problem data and the number of several local mine ecological problem data; based on the splicing weights, performing a first function processing on the first sharing coefficients of the pending mine ecological problem data corresponding to each local mine ecological problem data to obtain the splicing sharing coefficient; calculating the comparison result of the splicing sharing coefficient and the number of several local mine ecological problem data to obtain the second sharing coefficient.
[0136] 105. According to the second sharing coefficient, the target mine ecological problem data is determined from the pending mine ecological problem data.
[0137] This application uses artificial intelligence, which is specifically introduced as follows: Artificial Intelligence (AI) is the theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. In other words, artificial intelligence is a comprehensive technology in computer science that attempts to understand the essence of intelligence and produce a new intelligent machine that can respond in a similar way to human intelligence. Artificial intelligence is to study the design principles and implementation methods of various intelligent machines so that machines have the functions of perception, reasoning and decision-making.
[0138] Artificial intelligence technology is a comprehensive discipline that covers a wide range of fields, including both hardware-level and software-level technologies. The basic technologies of artificial intelligence generally include sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics and other technologies. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, and machine learning / deep learning.
[0139] In the embodiments of the present application, the artificial intelligence software technologies mainly involved include the above-mentioned speech processing technology and machine learning. For example, it may involve speech recognition technology (Automatic Speech Recognition, ASR) in speech technology (Speech Technology), including speech signal preprocessing (Speech signal preprocessing), speech signal frequency domain analysis (Speech signal frequency analyzing), speech signal feature extraction (Speech signal feature extraction), speech signal feature matching / recognition (Speech signal feature matching / recognition), speech training (Speech training), etc.
[0140] Artificial Intelligence (AI) is a technology that uses digital computers to simulate human perception of the environment, acquire knowledge, and use knowledge. This technology can enable machines to have functions similar to human perception, reasoning, and decision-making. The basic technologies of artificial intelligence generally include sensors, dedicated artificial intelligence chips, cloud computing, distributed storage, big data processing technology, operation / interaction systems, mechatronics, and other technologies. Artificial intelligence software technology mainly includes computer vision technology, speech processing technology, natural language processing technology, as well as machine learning / deep learning, autonomous driving, smart transportation, and other major directions.
[0141] In a possible implementation example, in step 105, based on the second sharing coefficient, a specific implementation method for determining the target mine ecological problem data from the pending mine ecological problem data may include: determining the pending mine ecological problem data with the largest second sharing coefficient among all the pending mine ecological problem data associated with a number of reference mine ecological problem data as the target mine ecological problem data.
[0142] In a possible implementation, after determining the target mine ecological problem data, the target mine ecological problem data may also be output. Optionally, in addition to outputting the target mine ecological problem data, mine ecological problem data information corresponding to the target mine ecological problem data may also be output.
[0143] It can be seen that in this embodiment, after obtaining the mine ecological problem data to be processed, the mine ecological problem data to be processed is split into several local mine ecological problem data with significant characteristics, and then for each local mine ecological problem data, several reference mine ecological problem data corresponding to the local mine ecological problem data are obtained, as well as the first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data; then, according to the first sharing coefficient, at least one pending mine ecological problem data associated with the local mine ecological problem data among the several reference mine ecological problem data is determined, and the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data is determined; finally, according to the second sharing coefficient, the target mine ecological problem data is determined from the pending mine ecological problem data. Since for each local mine ecological problem data of the mine ecological problem data to be processed, according to the first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data, similar pending mine ecological problem data are recalled from several reference mine ecological problem data, on the one hand, the recall efficiency is improved, and on the other hand, since the local mine ecological problem data has significant characteristics, the characteristics of the local mine ecological problem data can be captured more accurately, thereby improving the recall accuracy. Then, according to the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data, the target mine ecological problem data is screened out from the pending mine ecological problem data, on the one hand, the recognition scope can be narrowed, and on the other hand, the target mine ecological problem data that is more similar to the pending mine ecological problem data as a whole can be screened out from the pending mine ecological problem data. Thus, the accuracy and efficiency of mine ecological problem data identification are improved.
[0144] On the basis of the above, a mine ecological problem identification device based on artificial intelligence is provided, and the device comprises:
[0145] A data acquisition module is used to obtain data on mine ecological problems to be processed;
[0146] A data acquisition module is used to perform key feature analysis on the mine ecological problem data to be processed to obtain a number of local mine ecological problem data with significant characteristics;
[0147] A first coefficient obtaining module is used to obtain, for each local mine ecological problem data, a plurality of reference mine ecological problem data corresponding to the local mine ecological problem data, and a first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data;
[0148] a second coefficient obtaining module, for determining, in combination with the first sharing coefficient, at least one pending mine ecological problem data associated with the local mine ecological problem data among the plurality of reference mine ecological problem data, and determining a second sharing coefficient of the pending mine ecological problem data corresponding to the mine ecological problem data to be processed;
[0149] A problem data determination module is used to determine target mine ecological problem data from the pending mine ecological problem data in combination with the second sharing coefficient.
[0150] On the basis of the above, a mine ecological problem identification system based on artificial intelligence is shown, which includes a processor and a memory that communicate with each other, and the processor is used to read and execute a computer program from the memory to implement the above method.
[0151] Based on the above, a computer-readable storage medium is also provided, on which a computer program stored implements the above method when running.
[0152] In summary, based on the above scheme, the embodiment of the present application can, after obtaining the mine ecological problem data to be processed, split the mine ecological problem data to be processed into several local mine ecological problem data with significant characteristics, and then, for each local mine ecological problem data, obtain several reference mine ecological problem data corresponding to the local mine ecological problem data, as well as the first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data; then, based on the first sharing coefficient, determine at least one pending mine ecological problem data associated with the local mine ecological problem data among the several reference mine ecological problem data, and determine the second sharing coefficient of the pending mine ecological problem data corresponding to the mine ecological problem data to be processed; finally, based on the second sharing coefficient, determine the target mine ecological problem data from the pending mine ecological problem data.
[0153] In the embodiment of the present application, for each local mine ecological problem data of the mine ecological problem data to be processed, some similar reference mine ecological problem data can be preliminarily recalled according to the local mine ecological problem data, and then according to the first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data, similar pending mine ecological problem data are recalled from several reference mine ecological problem data, which improves the recall efficiency on the one hand, and on the other hand, because the local mine ecological problem data has significant characteristics, the characteristics of the local mine ecological problem data can be captured more accurately, thereby improving the recall accuracy. Then, according to the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data, the target mine ecological problem data is screened out from the pending mine ecological problem data, which can narrow the recognition range on the one hand, and on the other hand, the target mine ecological problem data that is more similar to the pending mine ecological problem data as a whole can be screened out from the pending mine ecological problem data. Thus, the accuracy of mine ecological problem data identification is improved.
[0154] It should be understood that the system and its modules shown above can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or a dedicated design hardware. Those skilled in the art will understand that the above methods and systems can be implemented using computer executable instructions and / or included in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The system and its modules of the present application can not only be implemented by hardware circuits such as ultra-large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., but can also be implemented by software such as executed by various types of processors, and can also be implemented by a combination of the above hardware circuits and software (e.g., firmware).
[0155] It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects that may be produced may be any one or a combination of the above, or any other beneficial effects that may be obtained.
Claims
1. A method for identifying mine ecological problems based on artificial intelligence, characterized in that: include: Obtain data on ecological issues in mines to be addressed; Performing key feature analysis on the to-be-processed mine ecological problem data to obtain a number of local mine ecological problem data with significant features; For each local mine ecological problem data, a plurality of reference mine ecological problem data corresponding to the local mine ecological problem data and a first sharing coefficient between the local mine ecological problem data and each reference mine ecological problem data are obtained; wherein the first sharing coefficient is used to characterize a first similarity, each reference mine ecological problem data is understood as sample data, and the similarity between the local mine ecological problem data and the sample data; In combination with the first sharing coefficient, determining at least one pending mine ecological problem data associated with the local mine ecological problem data among the plurality of reference mine ecological problem data, and determining a second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data; wherein the second sharing coefficient is used to characterize a second similarity, and is used to determine the ecological problem in the pending mine ecological problem data; In combination with the second sharing coefficient, determining target mine ecological problem data from the pending mine ecological problem data; The step of combining the first sharing coefficient to determine at least one pending mine ecological problem data associated with the local mine ecological problem data from among the plurality of reference mine ecological problem data includes: Numbering the plurality of reference mine ecological problem data in combination with the order set by the first sharing coefficient to obtain numbered reference mine ecological problem data; Determine the first a reference mine ecological problem data among the numbered reference mine ecological problem data as the pending mine ecological problem data, where a is a positive integer; Wherein, in combination with the first sharing coefficient, determining the second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data includes: For each pending mine ecological problem data, the first sharing coefficients of each local mine ecological problem data corresponding to the pending mine ecological problem data are spliced to obtain a second sharing coefficient of the pending mine ecological problem data corresponding to the pending mine ecological problem data; Among them, the key feature analysis and processing of the mine ecological problem data to be processed was carried out to obtain several local mine ecological problem data with significant characteristics, including: Identify the attribute nodes in the mine ecological problem data to be processed, where the attribute nodes are periodic nodes where the characteristics of the mine ecological problem data to be processed change; Based on the attribute nodes, the key features of the mine ecological problem data to be processed are analyzed and processed to obtain several local mine ecological problem data with significant features; Among them, identifying the attribute nodes in the mine ecological problem data to be processed includes: Extracting mine pollution features from the mine ecological problem data to be processed to obtain a number of mine pollution features, each of which corresponds to a periodic node in the mine ecological problem data to be processed, and the mine pollution features are arranged in order from the earliest to the latest periodic nodes; Determine at least one target mine pollution feature from a plurality of mine pollution features, the target mine pollution feature is used to segment a plurality of mine pollution features into a plurality of mine pollution feature sets, and the mine pollution features in each mine pollution feature set conform to Gaussian distribution; The periodic nodes corresponding to the pollution characteristics of the target mine are determined as attribute nodes.
2. The method for identifying mine ecological problems based on artificial intelligence according to claim 1 is characterized in that: The obtaining of a plurality of reference mine ecological problem data corresponding to the local mine ecological problem data comprises: Extracting features of the local mine ecological problem data to obtain mine ecological problem data features; An original feature vector associated with the mine ecological problem data feature is obtained from a plurality of pre-set original feature vectors and determined as a reference feature vector; the plurality of original feature vectors are extracted from a plurality of pre-set original mine ecological problem data; Performing mine ecological problem data tracing processing on the reference feature vector to obtain original mine ecological problem data corresponding to the reference feature vector; The original mine ecological problem data corresponding to the reference feature vector is determined as the reference mine ecological problem data.
3. The method for identifying mine ecological problems based on artificial intelligence according to claim 2 is characterized in that: The original feature vector associated with the mine ecological problem data feature is obtained from a plurality of pre-set original feature vectors and determined as a reference feature vector, including: Obtaining undetermined sharing coefficients between the mine ecological problem data features and the plurality of original feature vectors; Numbering the plurality of original feature vectors according to the order in which the to-be-determined sharing coefficients are set to obtain numbered original feature vectors; Determine the first M original feature vectors among the numbered original feature vectors as original feature vectors associated with the mine ecological problem data features, where M is a positive integer; The original feature vector associated with the mine ecological problem data feature is determined as the reference feature vector.
4. The method for identifying mine ecological problems based on artificial intelligence according to claim 3 is characterized in that: The obtaining of the undetermined sharing coefficients between the mine ecological problem data features and the plurality of original feature vectors includes: For each original feature vector, the original feature vector is numerically described to obtain a local AI vector space corresponding to the original feature vector, wherein the local AI vector space includes a plurality of local vectors having the same direction, and the direction of the local vector is smaller than the direction of the original feature vector; The local AI vector space is divided by a preset division network to obtain a division benchmark of the local AI vector space; Compressing the local AI vector space according to the division criteria to obtain category compression corresponding to the local AI vector space; Through the category compression and the mine ecological problem data characteristics, the undetermined sharing coefficient between the mine ecological problem data characteristics and the original feature vector is determined; wherein the undetermined sharing coefficient is used to characterize the similarity between the initial data and the ecological problem data.
5. The method for identifying mine ecological problems based on artificial intelligence according to claim 2 is characterized in that: The obtaining of the first sharing coefficients between the local mine ecological problem data and each reference mine ecological problem data includes: For the reference mine ecological problem data, identifying the vector number of reference feature vectors corresponding to the reference mine ecological problem data; If the number of the vectors is one, a reference feature vector corresponding to the reference mine ecological problem data and a mine ecological problem data feature corresponding to the local mine ecological problem data are used to calculate a sharing coefficient to obtain a first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data; If the number of vectors is several, the sharing coefficients of several reference feature vectors corresponding to the reference mine ecological problem data and the mine ecological problem data features corresponding to the local mine ecological problem data are calculated respectively to obtain several original sharing coefficients, and the several original sharing coefficients are spliced to obtain the first sharing coefficient between the local mine ecological problem data and the reference mine ecological problem data.
6. The method for identifying mine ecological problems based on artificial intelligence according to claim 1 is characterized in that: The step of splicing the first sharing coefficients of the pending mine ecological problem data corresponding to each local mine ecological problem data to obtain the second sharing coefficients of the pending mine ecological problem data corresponding to the pending mine ecological problem data includes: Obtaining the splicing weights of each local mine ecological problem data corresponding to the pending mine ecological problem data and the number of several local mine ecological problem data; By using the splicing weight, a first function is performed on the first sharing coefficient of each local mine ecological problem data corresponding to the pending mine ecological problem data to obtain a splicing sharing coefficient; The comparison result of the splicing sharing coefficient and the number of the plurality of local mine ecological problem data is calculated to obtain the second sharing coefficient.
7. A mine ecological problem identification system based on artificial intelligence, characterized in that: The invention comprises a processor and a memory communicating with each other, wherein the processor is used to read a computer program from the memory and execute the computer program to implement the method according to any one of claims 1 to 6.
