Method and system for realizing anomaly identification of mine data based on artificial intelligence

Abstract

The invention provides a mine data anomaly identification method and system based on artificial intelligence, and relates to the technical field of mine data anomaly identification. Historical mine data to be subjected to anomaly identification and data related to the data to be subjected to anomaly identification are acquired, and the mine data to be subjected to anomaly identification are calculated, so that the anomaly of the mine data to be subjected to anomaly identification is identified; the method comprises the steps of selecting a big data type of a correlation coefficient to form a main analysis data set, performing Laplace transformation on data in the main analysis data set, establishing a mine data anomaly detection model, and training the model to obtain a mine data anomaly detection result. Taking the frequency domain features in the main analysis data set at the current time point as the input of the trained mine data anomaly detection model, outputting the frequency domain features of the mine data to be subjected to anomaly recognition at the next time point, comparing the frequency domain features with the actual frequency domain features at the next time point, calculating a feature difference value, and comparing the feature difference value with a set threshold value to obtain the anomaly recognition result of the mine data. And judging whether the data is abnormal or not.

Description

Technical Field

[0001] The present invention relates to the technical field of mine data anomaly recognition, and in particular to a method and system for realizing mine data anomaly recognition based on artificial intelligence. Background Art

[0002] Mine data is crucial to mining production. It involves geological exploration, equipment status, environmental monitoring and other aspects, providing accurate decision support for mine operations. Accurate data can help companies optimize mining plans, improve resource utilization efficiency, and ensure operational safety. Data analysis can also predict equipment failures, reduce operation and maintenance costs, and enhance the overall stability of the production system.

[0003] However, anomalies in mine data can seriously threaten production safety and efficiency. Data anomalies may be caused by sensor errors, operational errors, or system failures. If these problems are not handled properly, they may lead to wrong mining decisions, premature equipment damage, and even safety accidents. Therefore, establishing an effective data monitoring and anomaly detection system is of great significance for maintaining the smooth progress of mine production and preventing production risks.

[0004] In the prior art, the publication number CN117953313B discloses an abnormality recognition method and system for realizing mine data based on artificial intelligence, which obtains the image data and text data of the mine through a preset acquisition unit, extracts features from the acquired image data and text data, obtains mine image features and mine periodic features, performs multimodal fusion on the mine image features and mine periodic features to obtain mine feature data, uses a neural network model to train an abnormality recognition model based on the public standard mine data, uses the abnormality recognition model to perform abnormality recognition on the obtained mine feature data, and judges whether there are safety hazards in the mine based on the abnormality recognition results. However, this method is seriously affected by image quality. Under harsh environmental conditions, such as fog, dust, low light, etc., the image data may become blurred or distorted, resulting in a decrease in recognition accuracy, and image processing usually requires high-performance computing resources.

[0005] The above information disclosed in this Background section is only for enhancement of understanding of the background of the present disclosure and therefore it may contain information that does not constitute the prior art that is already known to one of ordinary skill in the art. Summary of the invention

[0006] The purpose of the present invention is to provide a method and system for realizing anomaly recognition of mining data based on artificial intelligence, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] The method for realizing abnormal identification of mining data based on artificial intelligence includes the following specific steps:

[0009] Step 1: Obtain historical mining data to be identified at the same time interval, as well as historical data of types of data related to the data to be identified, pre-process the acquired historical data to ensure data quality, and stamp all types of data with the timestamp of collection to form a mining data set;

[0010] Step 2: By calculating the Pearson correlation coefficient between the mine data to be identified and the data of the types related to the mine data, the correlation between the types of data related to the mine data to be identified is analyzed, and the data types with large correlation are selected according to the size of the correlation coefficient to form the main analysis data set;

[0011] Step 3: Perform Laplace transform on the data in the main analysis data set, convert the time domain features of the data in the main analysis data set into frequency domain features, and extract key frequency domain features;

[0012] Step 4: Establish a mine data anomaly detection model, take the frequency domain features in the main analysis data set at a time point before the mine data anomaly as input data, and the frequency domain features of the abnormal mine data as output, and train the model;

[0013] Step 5: Use the frequency domain features in the main analysis data set at the current time point as the input of the trained mine data anomaly detection model, output the frequency domain features of the mine data to be identified as anomaly at the next time point, and calculate the feature difference by comparing it with the frequency domain features at the next time point. The feature difference is compared with the set threshold to determine whether the data is abnormal.

[0014] Furthermore, the method for preprocessing the historical data is normalization processing, and the specific processing formula is:

[0015]

[0016] Among them, X chu is historical data, X norm is the preprocessed data, X min , X max They are the minimum and maximum values ​​in the historical data respectively.

[0017] Furthermore, the specific logic of calculating the Pearson correlation coefficient between the mine data to be identified as abnormal and the data of the types related to the mine data is as follows;

[0018]

[0019] Among them, r x,yis the Pearson correlation coefficient between the mine data to be identified and the related data, x i is the i-th data of the mine data for abnormal identification, y i is the i-th data of the y-th related type of data, n is the number of each type of data, i and n are positive integers;

[0020] When | r x,y |≥α r When , it is considered that there is a correlation between the two data types;

[0021] Among them, α r is the correlation threshold.

[0022] Furthermore, the calculation formula for performing Laplace transform on the data in the main analysis data set is:

[0023]

[0024] Among them, X(f) is the frequency domain feature of the data in the main analysis data set, X(t m ) is the time point t m The value at, l is the total number of data points, Δt is the sampling time interval, and f is the frequency, which is usually taken as k is a positive integer, 0 <k<n―1。

[0025] Furthermore, the mine data anomaly detection model is a long short-term memory network model, including an input gate, a forget gate and an output gate;

[0026] The equation of the forget gate is:

[0027] f t =δ(W f ·[h t―1 ,x t ]+b f )

[0028] Among them, f t is the current switch state of the forget gate, δ represents the sigmoid function, W f is the forget gate weight, b f is the forget gate bias parameter, h t―1 is the hidden state at the previous moment, x t Input switch for current data;

[0029] Determine how much of the unit state at the previous moment needs to be retained to the current moment;

[0030] The equation for the input gate is:

[0031] i t =δ(W i ·[ht―1 ,x t ]+b i )

[0032] Among them, i t is the current switch state of the input gate, δ represents the sigmoid function, W i is the input gate weight, b i is the input gate bias parameter, h t―1 is the output hidden state of the previous moment, x t The frequency domain features in the current main analysis data set;

[0033]

[0034] in, Candidate cell state, tanh represents the hyperbolic tangent function, W c is the candidate cell weight, b c is the candidate cell state bias parameter, h t―1 is the hidden state at the previous moment, x t Input switch for current data;

[0035] Determine how much of the network's input data needs to be saved to the unit state at the current moment;

[0036] The update equation is;

[0037]

[0038] Among them, C t Current cell state, C t―1 is the cell state at the previous moment;

[0039] The equation for the output gate is:

[0040] o t =(W o ·[h t―1 ,x t ]+b o )

[0041] h t =o t *tanh(C t )

[0042] Among them, t is the current switch state of the output gate, δ represents the sigmoid function, W o is the output gate weight, b o is the output gate bias parameter, h t―1 is the hidden state at the previous moment, x t is the current data input switch, ht The frequency domain characteristics of the abnormal mining data at the next time point;

[0043] Controls how much of the current unit state needs to be output to the current output value;

[0044] Among them, the mathematical expressions of the sigmoid function and the hyperbolic tangent function are;

[0045]

[0046] Among them, δ(x γ ) is the sigmoid function, x γ is the input of sigmoid function, tanh(C t ) is the hyperbolic tangent function, C t Input to the function.

[0047] Furthermore, the calculation formula for calculating the characteristic difference is:

[0048] p=h t ―α p

[0049] Among them, p is the characteristic difference, α p To set the threshold, h t It is the frequency domain characteristics of mining data.

[0050] The present invention further provides a system for identifying anomalies in mining data based on artificial intelligence, wherein the system for identifying anomalies in mining data based on artificial intelligence is used to execute the above-mentioned method for identifying anomalies in mining data based on artificial intelligence, including:

[0051] A data acquisition module, which is used to collect mining data to be identified as abnormal, and data of related types to the data to be identified as abnormal, and send them to the data analysis module after pre-processing;

[0052] A data analysis module, which is used to calculate the Pearson correlation coefficient between the mine data to be identified and the data of the related types of the mine data, select the data type with high correlation according to the size of the correlation coefficient, form a main analysis data set, and perform Laplace transform on the data in the main analysis data set to extract key frequency domain features;

[0053] A data prediction module, which is used to establish a mine data anomaly detection model, taking the frequency domain features in the main analysis data set at the current time point as the input of the trained mine data anomaly detection model, and predicting the frequency domain features of the mine data to be anomaly identified at the next time point;

[0054] The abnormality identification module compares the frequency domain characteristics of the predicted mining data with the actual frequency domain characteristics at the next time point, calculates the characteristic difference, compares the characteristic difference with the set threshold, and determines whether the data is abnormal.

[0055] Compared with the prior art, the present invention has the following beneficial effects:

[0056] The present invention obtains historical mining data to be identified as abnormal and data of related types with the data to be identified as abnormal, calculates the correlation between the mining data to be identified as abnormal and data of related types with the mining data, selects the big data type of the correlation coefficient, forms a main analysis data set, performs Laplace transform on the data in the main analysis data set, extracts key frequency domain features, then establishes a mining data abnormality detection model, trains the model, uses the frequency domain features in the main analysis data set at the current time point as the input of the trained mining data abnormality detection model, outputs the frequency domain features of the mining data to be identified as abnormal at the next time point, calculates the feature difference by comparing with the actual frequency domain features at the next time point, compares the feature difference with the set threshold value, and judges whether the data is abnormal.

[0057] The present invention analyzes data that is highly correlated with the mine data to be identified as abnormal, performs Fourier transform on it, extracts frequency domain features, and uses a mine data anomaly detection model to identify abnormalities. This can more accurately identify abnormal situations in mine data and reduce the impact of accidental events on the identification of mine data anomalies. By converting to the frequency domain, the data structure can be simplified, important frequency components can be highlighted, and unimportant changes can be ignored, thereby reducing data complexity. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0059] Figure 2 It is a schematic diagram of the overall system structure of the present invention. DETAILED DESCRIPTION

[0060] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with specific embodiments.

[0061] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0062] Example:

[0063] See also Figure 1 , the present invention provides a technical solution:

[0064] The method for realizing abnormal identification of mining data based on artificial intelligence includes the following specific steps:

[0065] Step 1: Obtain historical mining data to be identified with abnormalities at the same time interval, as well as historical data of types of data related to the data to be identified with abnormalities, pre-process the acquired historical data to ensure data quality, and stamp all types of data with the timestamp of collection to form a mining data set.

[0066] Obtaining data at the same time interval ensures the continuity and consistency of the data. This is especially important for time series analysis, because the analysis of time series relies on the relationship between time points. If the data intervals are inconsistent, it may lead to deviations in the analysis results.

[0067] Related types of data refer to data that are related to the data to be identified as abnormal, such as dust and humidity. When the humidity increases, the effect of humidity on dust will decrease. Combining different but related data sets for analysis can provide a more comprehensive perspective and help explore potential connections and patterns between data.

[0068] Normalization is a data preprocessing method that aims to adjust data attributes to a scale between 0 and 1. This process usually involves adjusting the numerical range so that the data can be comparable between different features, thereby avoiding data interpretation difficulties and modeling biases caused by different scales. When the distribution of data is more standardized, the calculated correlation coefficient is more reliable and the statistical inference is more accurate.

[0069] In this embodiment, the method for preprocessing the historical data is normalization processing, and the specific processing formula is:

[0070]

[0071] Among them, X chu is historical data, X norm is the preprocessed data, X min , X max They are the minimum and maximum values ​​in the historical data respectively.

[0072] Step 2: By calculating the Pearson correlation coefficient between the mining data to be identified and the data of related types of mining data, the correlation between the data of related types of mining data to be identified is analyzed, and the data type with high correlation is selected according to the size of the correlation coefficient to form the main analysis data set.

[0073] By calculating the Pearson correlation coefficient between the mine data to be identified and other related types of data, we can understand the strength of association between different data types, so as to select the most influential data for in-depth analysis. Analyzing the correlation can help identify the most critical variables for anomaly identification. By focusing on these variables that are highly correlated with anomalies, abnormal events in mines can be more effectively predicted and detected, thereby improving the accuracy and efficiency of monitoring.

[0074] In this embodiment, the specific logic of calculating the Pearson correlation coefficient between the mine data to be identified as abnormal and the data of the types related to the mine data is as follows:

[0075]

[0076] Among them, r x,y is the Pearson correlation coefficient between the mine data to be identified and the related data, x i is the i-th data of the mine data for abnormal identification, y i is the i-th data of the y-th related type of data, n is the number of each type of data, i and n are positive integers;

[0077] When | r x,y |≥α r When , it is considered that there is a correlation between the two data types;

[0078] Among them, α r is the correlation threshold.

[0079] The larger the absolute value of the correlation coefficient, the stronger the linear relationship between the two variables. When the correlation coefficient is 1, it indicates that there is a perfect positive linear relationship between the two variables, that is, when one variable increases, the other variable also increases at a fixed ratio. When the correlation coefficient is -1, it indicates that there is a perfect negative linear relationship between the two variables, that is, when one variable increases, the other variable decreases at a fixed ratio. When the correlation coefficient is 0, it indicates that there is no linear relationship between the two variables.

[0080] Step 3: Perform Laplace transform on the data in the main analysis data set, convert the time domain features of the data in the main analysis data set into frequency domain features, and extract key frequency domain features.

[0081] In this embodiment, the calculation formula for performing Laplace transform on the data in the main analysis data set is:

[0082]

[0083] Among them, X(f) is the frequency domain feature of the data in the main analysis data set, X(t m ) is the time point t m The value at, l is the total number of data points, Δt is the sampling time interval, and f is the frequency, which is usually taken as k is a positive integer, 0 <k<n―1。

[0084] Fourier transform provides a method to convert time series data from the time domain to the frequency domain. This transformation can reveal the frequency components in the data, which helps to identify anomalies in mining data. In the frequency domain, abnormal behavior (such as sudden frequency changes, significant increases or decreases in energy) is often easier to identify. This is because the frequency distribution under normal operating conditions is usually known, and any significant deviation may indicate the presence of anomalies.

[0085] Step 4: Establish a mine data anomaly detection model, take the frequency domain features in the main analysis data set at the time point before the mine data anomaly as input data, and the frequency domain features of the abnormal mine data as output, and train the model.

[0086] Mining operations involve complex processes, and the monitoring of equipment and operations generates a large amount of time series data, in which many signals and events have long-term temporal dependencies. Anomalies in mining data are often closely related to historical data, and the excellent long-term memory capacity of the long short-term memory network model enables it to effectively capture these historical dependencies, thereby improving the accuracy of anomaly detection. Its gating mechanism can effectively manage information flow and decide when to update, forget or output information, thereby better handling time delays and dependencies.

[0087] In this embodiment, the mine data anomaly detection model is a long short-term memory network model, including an input gate, a forget gate and an output gate;

[0088] The equation of the forget gate is:

[0089] f t =δ(W f ·[h t―1 ,x t ]+b f )

[0090] Among them, f t is the current switch state of the forget gate, δ represents the sigmoid function, W f is the forget gate weight, b f is the forget gate bias parameter, h t―1 is the hidden state at the previous moment, x t Input switch for current data;

[0091] Determine how much of the unit state from the previous moment needs to be retained to the current moment;

[0092] The equation for the input gate is:

[0093] i t =δ(W i ·[h t―1 ,x t ]+b i )

[0094] Among them, i t is the current switch state of the input gate, δ represents the sigmoid function, W i is the input gate weight, b i is the input gate bias parameter, h t―1 is the output hidden state of the previous moment, x t The frequency domain features in the current main analysis data set;

[0095]

[0096] in, Candidate cell state, tanh represents the hyperbolic tangent function, W c is the candidate cell weight, b c is the candidate cell state bias parameter, h t―1 is the hidden state at the previous moment, x t Input switch for current data;

[0097] Determine how much of the network's input data needs to be saved to the unit state at the current moment;

[0098] The update equation is;

[0099]

[0100] Among them, C t Current cell state, C t―1 is the cell state at the previous moment;

[0101] The equation for the output gate is:

[0102] o t =(W o ·[h t―1 ,x t ]+b o )

[0103] g t =o t *tanh(C t )

[0104] Among them, t is the current switch state of the output gate, δ represents the sigmoid function, W o is the output gate weight, b o is the output gate bias parameter, h t―1 is the hidden state at the previous moment, x t is the current data input switch, h t The frequency domain characteristics of the abnormal mining data at the next time point;

[0105] Controls how much of the current unit state needs to be output to the current output value;

[0106] Among them, the mathematical expressions of the sigmoid function and the hyperbolic tangent function are;

[0107]

[0108] Among them, δ(x γ ) is the sigmoid function, x γ is the input of sigmoid function, tanh(C t ) is the hyperbolic tangent function, C t Input to the function.

[0109] Step 5: Use the frequency domain features in the main analysis data set at the current time point as the input of the trained mine data anomaly detection model, output the frequency domain features of the mine data to be identified as anomaly at the next time point, and calculate the feature difference by comparing it with the frequency domain features at the next time point. The feature difference is compared with the set threshold to determine whether the data is abnormal.

[0110] Frequency domain analysis can reveal features such as periodicity, mutation, and noise in the data, which may not be easily observed in the time domain. The comparison between the frequency domain features predicted by the LSTM model and the actual frequency domain features can more sensitively capture small changes, thereby improving the sensitivity of detecting anomalies. By accurately simulating and predicting frequency domain features, the LSTM model can help distinguish between normal operational fluctuations and true abnormal behavior. This precise judgment can reduce false positives and false negatives, and improve the reliability and effectiveness of the entire monitoring system.

[0111] In this embodiment, the calculation formula for calculating the characteristic difference is:

[0112] p=h t ―α p

[0113] Among them, p is the characteristic difference, α p To set the threshold, h t It is the frequency domain characteristics of mining data.

[0114] The characteristic difference directly indicates the deviation between the predicted data and the actual data. When this difference exceeds the normal range or threshold, it may indicate that there is an anomaly in the data. This anomaly may be caused by equipment failure, operating errors, changes in the external environment or other abnormal conditions. By continuously tracking these differences, we can better understand and manage the complex systems in mines.

[0115] See also Figure 2 The present invention further provides a system for identifying anomalies in mining data based on artificial intelligence, wherein the system for identifying anomalies in mining data based on artificial intelligence is used to execute the above-mentioned method for identifying anomalies in mining data based on artificial intelligence, including:

[0116] A data acquisition module, which is used to collect mining data to be identified as abnormal, and data of related types to the data to be identified as abnormal, and send them to the data analysis module after pre-processing;

[0117] A data analysis module, which is used to calculate the Pearson correlation coefficient between the mine data to be identified and the data of the related types of the mine data, select the data type with high correlation according to the size of the correlation coefficient, form a main analysis data set, and perform Laplace transform on the data in the main analysis data set to extract key frequency domain features;

[0118] A data prediction module, which is used to establish a mine data anomaly detection model, taking the frequency domain features in the main analysis data set at the current time point as the input of the trained mine data anomaly detection model, and predicting the frequency domain features of the mine data to be anomaly identified at the next time point;

[0119] The abnormality identification module predicts the frequency domain characteristics of the mining data and compares them with the actual frequency domain characteristics at the next time point, calculates the characteristic difference, compares the characteristic difference with the set threshold, and determines whether the data is abnormal.

[0120] The above formulas are all dimensionless and numerical calculations. The formula is a formula for the most recent real situation obtained by collecting a large amount of data and performing software simulation. The preset parameters in the formula are set by technicians in this field according to actual conditions.

[0121] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product. Those skilled in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0122] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, and may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0123] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for identifying abnormalities in mining data based on artificial intelligence, characterized in that: The specific steps include: Step 1: Obtain historical mining data to be identified at the same time interval, as well as historical data of types of data related to the data to be identified, pre-process the acquired historical data to ensure data quality, and stamp all types of data with the timestamp of collection to form a mining data set; Step 2: By calculating the Pearson correlation coefficient between the mine data to be identified and the data of the types related to the mine data, the correlation between the types of data related to the mine data to be identified is analyzed, and the data types with large correlation are selected according to the size of the correlation coefficient to form the main analysis data set; Step 3: Perform Laplace transform on the data in the main analysis data set, convert the time domain features of the data in the main analysis data set into frequency domain features, and extract key frequency domain features; Step 4: Establish a mine data anomaly detection model, take the frequency domain features in the main analysis data set at a time point before the mine data anomaly as input data, and the frequency domain features of the abnormal mine data as output, and train the model; Step 5: Use the frequency domain features in the main analysis data set at the current time point as the input of the trained mine data anomaly detection model, output the frequency domain features of the mine data to be identified as anomaly at the next time point, and calculate the feature difference by comparing it with the frequency domain features at the next time point. The feature difference is compared with the set threshold to determine whether the data is abnormal.

2. The method for realizing abnormal identification of mine data based on artificial intelligence according to claim 1 is characterized in that: The method for preprocessing the historical data is normalization processing, and the specific processing formula is: Among them, X chu is historical data, X norm is the preprocessed data, X min , X max They are the minimum and maximum values ​​in the historical data respectively.

3. The method for realizing abnormal identification of mine data based on artificial intelligence according to claim 1 is characterized in that: The specific logic of calculating the Pearson correlation coefficient between the mine data to be identified as abnormal and the data of the related types of the mine data is as follows: Among them, r x,y is the Pearson correlation coefficient between the mine data to be identified and the related data, x i is the i-th data of the mine data for abnormal identification, y i is the i-th data of the y-th related type of data, n is the number of each type of data, i and n are positive integers; When | r x,y |≥α r When , it is considered that there is a correlation between the two data types; Among them, α r is the correlation threshold.

4. The method for realizing abnormal identification of mining data based on artificial intelligence according to claim 1 is characterized in that: The calculation formula for performing Laplace transform on the data in the main analysis data set is: Among them, X(f) is the frequency domain feature of the data in the main analysis data set, X(t m ) is the time point t m The value at, l is the total number of data points, Δt is the sampling time interval, and f is the frequency, which is usually taken as k is a positive integer, 0 <k<n―1。 5. The method for realizing abnormal identification of mine data based on artificial intelligence according to claim 1 is characterized in that: The mine data anomaly detection model is a long short-term memory network model, including an input gate, a forget gate and an output gate; The equation of the forget gate is: f t =δ(W f ·[h t―1 ,x t ]+b f ) Among them, f t is the current switch state of the forget gate, δ represents the sigmoid function, W f is the forget gate weight, b f is the forget gate bias parameter, h t―1 is the hidden state at the previous moment, x t Input switch for current data; Determine how much of the unit state at the previous moment needs to be retained to the current moment; The equation for the input gate is: i t =δ(W i ·[h t―1 ,x t ]+b i ) Among them, i t is the current switch state of the input gate, δ represents the sigmoid function, W i is the input gate weight, b i is the input gate bias parameter, h t―1 is the output hidden state of the previous moment, x t The frequency domain features in the current main analysis data set; in, Candidate cell state, tanh represents the hyperbolic tangent function, W c is the candidate cell weight, b c is the candidate cell state bias parameter, h t―1 is the hidden state at the previous moment, x t Input switch for current data; Determine how much of the network's input data needs to be saved to the unit state at the current moment; The update equation is; Among them, C t Current cell state, C t―1 is the cell state at the previous moment; The equation for the output gate is: o t =(W o ·[h t―1 ,x t ]+b o ) h t =o t *tanh(C t ) Among them, t is the current switch state of the output gate, δ represents the sigmoid function, W o is the output gate weight, b o is the output gate bias parameter, h t―1 is the hidden state at the previous moment, x t is the current data input switch, h t The frequency domain characteristics of the abnormal mining data at the next time point; Controls how much of the current unit state needs to be output to the current output value; Among them, the mathematical expressions of the sigmoid function and the hyperbolic tangent function are; Among them, δ(x γ ) is the sigmoid function, x γ is the input of sigmoid function, tanh(C t ) is the hyperbolic tangent function, C t Input to the function.

6. The method for realizing abnormal identification of mining data based on artificial intelligence according to claim 1 is characterized in that: The calculation formula for calculating the characteristic difference is: p=h t -a p Among them, p is the characteristic difference, α p To set the threshold, h t It is the frequency domain characteristics of mining data.

7. An abnormal identification system for mining data based on artificial intelligence, characterized by: The mine data anomaly identification system is used to perform the artificial intelligence-based mine data anomaly identification as described in any one of claims 1 to 6, including: A data acquisition module, which is used to collect mining data to be identified as abnormal, and data of related types to the data to be identified as abnormal, and send them to the data analysis module after pre-processing; A data analysis module, which is used to calculate the Pearson correlation coefficient between the mine data to be identified and the data of the related types of the mine data, select the data type with high correlation according to the size of the correlation coefficient, form a main analysis data set, and perform Laplace transform on the data in the main analysis data set to extract key frequency domain features; A data prediction module, which is used to establish a mine data anomaly detection model, taking the frequency domain features in the main analysis data set at the current time point as the input of the trained mine data anomaly detection model, and predicting the frequency domain features of the mine data to be anomaly identified at the next time point; The abnormality identification module predicts the frequency domain characteristics of the mining data and compares them with the actual frequency domain characteristics at the next time point, calculates the characteristic difference, compares the characteristic difference with the set threshold, and determines whether the data is abnormal.

Images