An intelligent monitoring method and system for the safe disposal of coke tar oil

Abstract

The present invention relates to the technical field of safety monitoring of coke tar oil, and discloses an intelligent monitoring method and system for safe disposal of coke tar oil. The method includes: collecting environmental parameter data of the coke tar oil storage environment and performing preprocessing; using a parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar oil environmental parameter data; correcting the preprocessed coke tar oil environmental parameter data according to the parameter data offset estimation result; extracting features from the corrected coke tar oil environmental parameter data; and constructing a coke tar oil monitoring anomaly recognition model for coke tar oil environmental monitoring. The present invention divides the data sequence into multiple batches, calculates the offset estimation coefficient according to the extreme value difference and abnormal data value difference between adjacent batches to correct the data sequence, improves the coherence and stability of the data sequence, and uses the method of feature decomposition combined with the genetic algorithm to extract feature vectors, realizing coke tar oil environmental monitoring and safe disposal.

Description

Technical Field

[0001] The present invention relates to the field of safety monitoring of coke tar, and particularly to an intelligent monitoring method and system for the safe disposal of coke tar. Background Art

[0002] Coke tar is an important industrial raw material, mainly produced during the high-temperature carbonization of coal, and is a by-product in the coal chemical industry. The chemical composition of coke tar is complex, containing toxic and harmful substances such as benzene, naphthalene, phenol, and pyridine, and has certain corrosiveness, flammability, and toxicity. Therefore, in coal chemical enterprises, the safe disposal of coke tar has become the focus of attention. If not properly treated, the leakage, volatilization, or combustion of coke tar will pose a great threat to the environment and personnel. Especially during storage and transportation, it may trigger major safety accidents. With the development of industrial intelligence, more and more intelligent technologies are being applied to the monitoring and management of hazardous chemicals. For example, Internet of Things technology can achieve real-time monitoring of the storage environment of coke tar, and artificial intelligence technology can predict potential safety hazards through data analysis, improving the safety management level of coke tar. In the storage and transportation environment of coke tar, key parameters such as temperature, humidity, and pressure have a direct impact on the safety of coke tar. As an important part of the intelligent monitoring system, the accuracy and reliability of sensors are crucial. At present, many sensor technologies are greatly affected by the environment in practical applications. For example, in high-temperature, high-humidity, and corrosive environments, the sensitivity of sensors may decrease, and even measurement data deviation may occur, resulting in the failure or false alarm of the monitoring system, thereby affecting the safety monitoring effect of coke tar. Summary of the Invention

[0003] In view of this, the present invention provides an intelligent monitoring method for the safe disposal of coke tar, which ensures stable and accurate monitoring data in various complex environments through multi-level sensing data fusion analysis, realizes real-time monitoring of the safety status of coke tar, early warning of potential hazards, and rapid response to accidents.

[0004] To achieve the above object, an intelligent monitoring method for the safe disposal of coke tar provided by the present invention includes the following steps:

[0005] S1: Collect environmental parameter data of the coke tar storage environment and perform preprocessing to obtain preprocessed coke tar environmental parameter data;

[0006] S2: Construct a parameter data offset detection model, and use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar environmental parameter data to obtain a parameter data offset estimation result, where data offset detection uses a multi-batch data statistical analysis method;

[0007] S3: Modify the preprocessed environmental parameter data of coke tar oil according to the parameter data offset estimation result to obtain the modified environmental parameter data of coke tar oil;

[0008] S4: Extract features from the modified environmental parameter data of coke tar oil to obtain the environmental feature vector of coke tar oil;

[0009] S5: Construct a coke tar oil monitoring anomaly recognition model for coke tar oil environmental monitoring. The coke tar oil monitoring anomaly recognition model takes the environmental feature vector of coke tar oil as input and the monitoring anomaly recognition result as output, and performs safety disposal according to the monitoring anomaly recognition result. Among them, the coke tar oil monitoring anomaly recognition model uses a clustering algorithm to judge anomalies.

[0010] As a further improvement method of the present invention:

[0011] Optionally, the S1 step includes:

[0012] Collect environmental parameter data of the coke tar oil storage environment using a variety of sensors. The environmental parameters include environmental temperature, environmental humidity, the static liquid pressure at the bottom of the coke tar oil storage container, the liquid level height of the coke tar oil in the storage container, and the concentration of benzene substances in the coke tar oil. In the embodiment of the present invention, the concentration of benzene substances is measured by gas chromatography. The form of the collected environmental parameter data is:

[0013] x = {x u =(x u (1), x u (2),..., x u (n),..., x u (N))|u ∈ [1, 5]}

[0014] Where:

[0015] x represents the environmental parameter data;

[0016] x u represents the data sequence of the u-th environmental parameter, u ∈ [1, 5]. The 1st - 5th environmental parameters are environmental temperature, environmental humidity, the static liquid pressure at the bottom of the coke tar oil storage container, the liquid level height of the coke tar oil in the storage container, and the concentration of benzene substances in the coke tar oil respectively;

[0017] x u (1), x u (2),..., x u (n),..., x u (N) represents the data value of the u-th environmental parameter at N monitoring times, n ∈ [1, N];

[0018] Preprocess the environmental parameter data x to obtain the preprocessed environmental parameter data of coking coal tar, where the preprocessing process is as follows:

[0019] S11: Filter the data sequence in the environmental parameter data x to obtain the filtered data sequence, where the filtered result of the data sequence x u is x' u , and the data value x u (n) The filtering formula is:

[0020]

[0021] where:

[0022] x' u (n) represents the filtered result of the data value x u (n);

[0023] exp(·) represents the exponential function with the natural constant as the base;

[0024] ε represents the position coefficient;

[0025] x' u =(x' u (1), x' u (2),..., x' u (n),..., x' u (N));

[0026] S12: Calculate the distance between any two data values in the filtered data sequence;

[0027] S13: Calculate the abnormal information coefficient of any data value in the filtered data sequence, mark the data value with the abnormal information coefficient higher than the preset threshold as an abnormal data value, and mark other data values as non-abnormal data values;

[0028] S14: Select the maximum and minimum non-abnormal data values from the filtered data sequence, and perform normalization processing on all data values in the filtered data sequence to obtain the normalized data sequence, where the normalized data sequence of the u-th environmental parameter is X u =(X u (1), X u (2),..., X u (n),..., X u (N)), X u (n) represents the processing result of the data value x u (n) after filtering and normalization processing;

[0029] Take all the normalized data sequences as the preprocessed environmental parameter data of coking coal tar.

[0030] Optionally, building a parameter data offset detection model in step S2 includes:

[0031] Build a parameter data offset detection model, which takes the preprocessed coke tar environment parameter data as input and the parameter data offset estimation result as output. The parameter data offset detection model includes a sequence division layer, a difference calculation layer, and an offset estimation layer;

[0032] The sequence division layer is used to divide the normalized data sequence into continuous data batches;

[0033] The difference calculation layer is used to calculate the statistical difference between continuous data batches and convert the statistical difference into an offset probability;

[0034] The offset estimation layer is used to calculate the offset estimation result of each data batch by combining the offset probability.

[0035] Optionally, using the parameter data offset detection model to perform parameter data offset estimation on the preprocessed coke tar environment parameter data to obtain a parameter data offset estimation result, including:

[0036] Use the parameter data offset detection model to perform parameter data offset estimation on the preprocessed coke tar environment parameter data to obtain the offset estimation result of any normalized data sequence, and integrate the offset estimation results of all normalized data sequences into a parameter data offset estimation result, where the normalized data sequence X u The calculation process of the offset estimation result is as follows:

[0037] S21: The sequence division layer divides the normalized data sequence X u into continuous data batches:

[0038]

[0039] Among them:

[0040] represents the m-th data batch obtained by division, and each data batch contains 3 consecutive normalized data values;

[0041] X u (3m) represents the 3m-th data value in the data sequence X u ;

[0042] S22: The difference calculation layer calculates the statistical difference between continuous data batches, where the statistical difference between the data batch and is

[0043]

[0044] Among them:

[0045] represents the extreme value difference between data batches, max(·) represents obtaining the maximum value through traversal, and min(·) represents obtaining the minimum value through traversal; and The extreme value difference between them, max(·) represents obtaining the maximum value through traversal, and min(·) represents obtaining the minimum value through traversal;

[0046] represents the outlier difference between data batches and ; represents the mean value of the abnormal data values after normalization processing in the calculated data batch;

[0047] S23: Convert the statistical difference of the data batch into an offset probability, where the statistical difference The conversion formula is:

[0048]

[0049] Among them:

[0050] represents the statistical difference The corresponding offset probability;

[0051] S24: Combine the offset probability to calculate the offset estimation result of each data batch; among them, the offset estimation result of the data batch is:

[0052]

[0053] Among them:

[0054] represents the offset estimation result of the data batch ;

[0055] δ(X u (3m)) represents the offset estimation coefficient of the data value X u in the normalized data sequence X u (3m);

[0056] represents the standard deviation of the data batch ; represents the mean value of the data batch ;

[0057] S25: Integrate the offset estimation results of all data batches as the offset estimation result of the normalized data sequence X u ;

[0058] Optionally, the step S3 includes:

[0059] Correct the preprocessed coking kerosene environment parameter data according to the parameter data offset estimation result to obtain the correction result of any normalized data sequence, and form the corrected coking kerosene environment parameter data with the correction results, where the normalized data sequence X u The correction process is as follows:

[0060] S31: Obtain the normalized data sequence X u ;

[0061] S32: Correct any data value in the normalized data sequence X u where the correction formula for the data value X u (n) is:

[0062] Y u (n) = X u (n)·δ(X u (n))

[0063] where:

[0064] Y u (n) represents the correction result of the data value X u (n);

[0065] δ(X u (n)) represents the offset estimation coefficient corresponding to the data value X u (n);

[0066] S33: Form the correction result Y u of the normalized data sequence X u = (Y u (1), Y u (2),..., Y u (n),..., Y u (N)).

[0067] Optionally, the step S4 includes:

[0068] Extract features from the corrected coking kerosene environment parameter data to obtain the coking kerosene environment feature vector F, where the feature extraction process is as follows:

[0069] S41: Form the data matrix C with the corrected coking kerosene environment parameter data:

[0070] C = [Y1, Y2, Y3, Y4, Y5] T

[0071] where:

[0072] Y1, Y2, Y3, Y4, Y5 represent the corrected data sequences corresponding to five environmental parameters, and T represents the transpose;

[0073] S42: Use the genetic algorithm to solve the influence weights of environmental parameters, and perform weighted processing on the data matrix C to form a data weighted matrix C':

[0074] C' = [w1Y1, w2Y2, w3Y3, w4Y4, w5Y5] T

[0075] Where:

[0076] w1, w2, w3, w4, w5 are the influence weights of the five environmental parameters in sequence;

[0077] S43: Perform eigen decomposition on the data weighted matrix C' to obtain the five largest eigenvalues of the data weighted matrix C' and the corresponding eigenvectors, and form the coke tar environment eigenvector F with the five eigenvectors:

[0078] F = [α1, α2, α3, α4, α5]C

[0079] Where:

[0080] α1, α2, α3, α4, α5 represent the eigenvectors corresponding to the five largest eigenvalues of the data weighted matrix C'.

[0081] Optionally, in step S42, using the genetic algorithm to solve the influence weights of environmental parameters includes:

[0082] S421: Initialize D groups of chromosomes, where each group of chromosomes has 5 segments of genes, respectively representing the influence weights of the five environmental parameters. The d-th group of chromosomes generated by initialization is:

[0083]

[0084] Where:

[0085] represents the d-th group of chromosomes generated by initialization, represents the chromosome The 5 segments of genes in;

[0086] S422: Set the current iteration number of the chromosome to t, the initial value of t is 0, and the maximum value is Max. Then the t-th iteration result of the d-th group of chromosomes is

[0087] S423: Set the chromosome genetic objective function G1(·), substitute the iterated chromosome into the chromosome genetic objective function G1(·) to obtain the chromosome genetic objective function value, where the chromosome The corresponding chromosome genetic objective function value is

[0088]

[0089] Where:

[0090] represents the weighted and eigen-decomposed data matrix C using chromosome to form the characteristic vector of the coking coal environment;

[0091] ||·|| represents the L1 norm, and ||·||2 represents the L2 norm; in the embodiments of the present invention, represents the richness of the characteristic information in

[0092] S424: Set the chromosome mutation objective function G2(·), substitute the iteratively obtained chromosome into the chromosome mutation objective function G1(·), and obtain the chromosome mutation objective function value of the chromosome, where the chromosome The corresponding chromosome mutation objective function value is:

[0093]

[0094] Where:

[0095] represents the sequence composed of 5 groups of eigenvectors obtained by weighting and eigen-decomposing the data matrix C using chromosome ;

[0096] S425: Calculate the mutation probability of the chromosome mutation according to the chromosome mutation objective function value of the chromosome; where the chromosome The mutation probability of the mutation is:

[0097]

[0098] Where:

[0099] represents the mutation probability of the chromosome mutating;

[0100] S426: Extract the mutated chromosome as the mutant chromosome, and the chromosome with a chromosome genetic objective function value higher than that of the mutant chromosome as the genetic chromosome, and perform crossover processing on some gene segments in the mutant chromosome and the corresponding gene segments of the genetic chromosome;

[0101] S427: Let d = d + 1, return to step S425 until the maximum number of iterations is reached, and use the chromosome with the highest genetic objective function value as the solution result, and extract the influence weights of the environmental parameters.

[0102] Optionally, in the S5 step, constructing a monitoring anomaly recognition model for coking coal tar to conduct coking coal tar environmental monitoring includes:

[0103] Construct a monitoring anomaly recognition model for coking coal tar. The monitoring anomaly recognition model for coking coal tar takes the extracted coking coal tar environmental feature vector as the input and the monitoring anomaly recognition result as the output. The monitoring anomaly recognition model for coking coal tar includes an input layer, a feature convolution layer, a feature weight calculation layer, and a mapping layer;

[0104] The input layer is used to receive the coking coal tar environmental feature vector;

[0105] The feature convolution layer is a convolutional neural network model structure for performing multi-level convolution processing on the coking coal tar environmental feature vector;

[0106] The feature weight calculation layer includes an attention unit and a residual unit in the residual network for calculating the weights of the convolutional features;

[0107] The mapping layer performs clustering and fusion processing on the convolutional features based on the weights and maps the clustered and fused features to the monitoring anomaly recognition result;

[0108] Using the monitoring anomaly recognition model for coking coal tar to conduct coking coal tar environmental monitoring, where the coking coal tar environmental monitoring process is as follows:

[0109] S51: The input layer receives the coking coal tar environmental feature vector F;

[0110] S52: The feature convolution layer performs multi-level convolution processing on the coking coal tar environmental feature vector F to obtain multi-level convolutional features:

[0111] F(k) = W k×k *F, k ∈ [1, K]

[0112] Where:

[0113] F(k) represents the convolutional feature of the coking coal tar environmental feature vector F at the k-scale level, and K represents the preset maximum scale level; W k×k represents a k-row and k-column convolutional weight matrix, and * represents the convolution operator;

[0114] S53: The feature weight calculation layer calculates the weights of the convolutional features, where the weight of F(k) is:

[0115]

[0116] Where:

[0117] tanh(F(k - 1)) represents the residual information at the (k - 1)-scale level;

[0118] weight(k) represents the weight of F(k);

[0119] S54: The mapping layer weights the convolutional features based on the weights, clusters the weighted convolutional features to obtain a number of clustering clusters, extracts the clustering centers of each clustering cluster for splicing to obtain the clustering fusion feature G, and maps the clustering fusion feature G to the monitoring anomaly recognition result:

[0120]

[0121] Where:

[0122] W represents the mapping matrix;

[0123] V represents the monitoring anomaly recognition result. The higher the V value, the higher the leakage concentration of the tar oil and the more serious the environmental pollution;

[0124] Perform safety disposal according to the monitoring anomaly recognition result.

[0125] Optionally, calculating the anomaly information coefficient of any data value in the filtered data sequence in the S13 step includes:

[0126] The data value x′ u (n)'s anomaly information coefficient calculation process is as follows:

[0127] S131: Traverse to obtain the s-th data value closest to the data value x′ u in x′ and calculate the distance between the two as the nearest neighbor distance of the data value x′ u (n). The data values in x′ whose distance from the data value x′ u (n) is less than the nearest neighbor distance form the nearest neighbor data set Near(x′ u of the data value x′ u (n); u (n)); u (n));

[0128] S132: Calculate the density of the data value x′ u (n):

[0129]

[0130] Where:

[0131] ρ(x′ u (n)) represents the density of the data value x′ u (n);

[0132] σ u represents the standard deviation of the data sequence x′ u ;

[0133] Near(x′ u (n), e) represents the e-th data value in the nearest neighbor data set Near(x′ u (n));

[0134] S133: Calculate the anomaly information coefficient S(x′ u (n)) of the data value x′ u (n):

[0135]

[0136] Wherein:

[0137] ρ(Near(x′ u (n), e)) represents the density of the e-th data value Near(x′ u (n), e) in the nearest neighbor data set Near(x′ u (n));

[0138] To solve the above problems, the present invention provides an intelligent monitoring system for the safe disposal of coke tar oil, and the system includes:

[0139] A data acquisition module, configured to acquire environmental parameter data of the coke tar oil storage environment, perform preprocessing on the acquired data to obtain preprocessed coke tar oil environmental parameter data, construct a parameter data offset detection model, use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar oil environmental parameter data, and correct the preprocessed coke tar oil environmental parameter data according to the estimated parameter data offset estimation result;

[0140] A feature extraction module, configured to extract features from the corrected coke tar oil environmental parameter data to obtain a coke tar oil environmental feature vector;

[0141] An environmental monitoring device, configured to construct a coke tar oil monitoring anomaly recognition model for coke tar oil environmental monitoring.

[0142] To solve the above problems, the present invention further provides an electronic device, and the electronic device includes:

[0143] A memory, storing at least one instruction;

[0144] A communication interface, configured to implement communication of the electronic device; and

[0145] A processor, configured to execute the instructions stored in the memory to implement the above-mentioned intelligent monitoring method for the safe disposal of coke tar oil.

[0146] To solve the above problems, the present invention further provides a computer-readable storage medium, in which at least one instruction is stored, and the at least one instruction is executed by a processor in an electronic device to implement the above-mentioned intelligent monitoring method for safe disposal of coke tar oil.

[0147] Compared with the prior art, the present invention proposes an intelligent monitoring method for safe disposal of coke tar oil, and this technology has the following advantages:

[0148] First of all, the solution of the present invention uses a variety of sensors to collect environmental parameter data of the coke tar oil storage environment, combines the density differences between data values and other adjacent data values in different data sequences, calculates the abnormal information coefficient of the data value, identifies the abnormal data values in the data sequence, divides the data sequence into batches, calculates the extreme value difference and abnormal data value difference between adjacent batches, takes them as the statistical differences of data batches and converts them into offset probabilities. The greater the statistical difference between adjacent batches, the greater the probability of offset of this data batch. Furthermore, an offset estimation coefficient of different data values is generated to realize data offset estimation and correction processing under multiple batches, and improve the coherence and stability of the data sequence.

[0149] At the same time, this solution uses the method of eigenvalue decomposition to extract features from the corrected coke tar oil environmental parameter data. During the eigenvalue decomposition process, according to the amplitude of iterative update of the influence weight and the iterative update effect, a chromosome mutation objective function and a chromosome genetic objective function are respectively constructed. The iterative update effect includes the similarity degree between the final coke tar oil environmental feature vector and the original data matrix and the richness of the feature information of the coke tar oil environmental feature vector. Taking the chromosome as the influence weight of different environmental parameters, the mutation probability and gene inheritance mode of different chromosomes are calculated according to the chromosome mutation objective function value and the chromosome genetic objective function value, reducing the randomness of chromosome selection in the genetic algorithm, improving the iterative efficiency, and using the optimal influence weight obtained by multiple iterations to extract a coke tar oil environmental feature vector with richer feature information, realizing coke tar oil environmental monitoring, and performing safe disposal according to the monitoring results. BRIEF DESCRIPTION OF THE DRAWINGS

[0150] Figure 1 It is a schematic flowchart of an intelligent monitoring method for safe disposal of coke tar oil provided by an embodiment of the present invention;

[0151] Figure 2 It is a functional module diagram of an intelligent monitoring system for safe disposal of coke tar oil provided by an embodiment of the present invention;

[0152] Figure 2Chinese: 100 - Jiao kerosene safety disposal intelligent monitoring system, 101 data acquisition module, 102 feature extraction module, 103 environmental monitoring device;

[0153] The realization of the purpose of the present invention, its functional characteristics and advantages will be further described in conjunction with embodiments with reference to the accompanying drawings. Detailed implementation manners

[0154] It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0155] The embodiments of the present application provide an intelligent monitoring method for the safe disposal of jiao kerosene. The execution subject of the intelligent monitoring method for the safe disposal of jiao kerosene includes but is not limited to at least one of electronic devices such as a server, a terminal, etc. that can be configured to execute the method provided by the embodiments of the present application. In other words, the intelligent monitoring method for the safe disposal of jiao kerosene can be executed by software or hardware installed on a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to: a single server, a server cluster, a cloud server, or a cloud server cluster, etc.

[0156] Embodiment 1:

[0157] An intelligent monitoring method for the safe disposal of jiao kerosene includes the following steps:

[0158] S1: Collect the environmental parameter data of the jiao kerosene storage environment and perform pre - processing to obtain the pre - processed jiao kerosene environmental parameter data.

[0159] The S1 step includes:

[0160] Use a variety of sensors to collect the environmental parameter data of the jiao kerosene storage environment, where the environmental parameters include environmental temperature, environmental humidity, the static liquid pressure at the bottom of the jiao kerosene storage container, the liquid level height of the jiao kerosene in the storage container, and the concentration of benzene substances in the jiao kerosene. The form of the collected environmental parameter data is:

[0161] x = {x u =(x u (1),x u (2),...,x u (n),...,x u (N))|u∈[1,5]}

[0162] Where:

[0163] x represents the environmental parameter data;

[0164] x uThe data sequence representing the u-th environmental parameter, where u ∈ [1, 5], and the 1st - 5th environmental parameters are environmental temperature, environmental humidity, static liquid pressure at the bottom of the coke tar storage container, liquid level height of coke tar in the storage container, and concentration of benzene substances in coke tar in sequence;

[0165] x u (1), x u (2),..., x u (n),..., x u (N) represents the data value of the u-th environmental parameter at N monitoring moments, where n ∈ [1, N];

[0166] Preprocess the environmental parameter data x to obtain the preprocessed coke tar environmental parameter data, and the preprocessing process is as follows:

[0167] S11: Perform filtering processing on the data sequence in the environmental parameter data x to obtain the filtered data sequence, where the filtering processing result of the data sequence x u is x′ u , and the filtering processing formula for the data value x u (n) is:

[0168]

[0169] Where:

[0170] x′ u (n) represents the filtering processing result of the data value x u (n);

[0171] exp(·) represents the exponential function with the natural constant as the base;

[0172] ε represents the position coefficient;

[0173] x′ u =(x′ u (1), x′ u (2),..., x′ u (n),..., x′ u (N));

[0174] S12: Calculate the distance between any two data values in the filtered data sequence;

[0175] S13: Calculate the abnormal information coefficient of any data value in the filtered data sequence, mark the data values with abnormal information coefficients higher than the preset threshold as abnormal data values, and mark other data values as non-abnormal data values;

[0176] S14: Select the maximum and minimum non-abnormal data values from the filtered data sequence, and normalize all the data values in the filtered data sequence to obtain a normalized data sequence. The normalized data sequence of the u-th environmental parameter is X u =(X u (1), X u (2),..., X u (n),..., X u (N)), where X u (n) represents the processing result of the data value x u (n) after filtering and normalization;

[0177] Take all the normalized data sequences as the preprocessed coke tar environmental parameter data.

[0178] S2: Construct a parameter data offset detection model, and use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar environmental parameter data to obtain a parameter data offset estimation result.

[0179] In the S2 step, constructing the parameter data offset detection model includes:

[0180] Construct a parameter data offset detection model. The parameter data offset detection model takes the preprocessed coke tar environmental parameter data as input and the parameter data offset estimation result as output. The parameter data offset detection model includes a sequence division layer, a difference calculation layer, and an offset estimation layer;

[0181] The sequence division layer is used to divide the normalized data sequence into continuous data batches;

[0182] The difference calculation layer is used to calculate the statistical difference between continuous data batches and convert the statistical difference into an offset probability;

[0183] The offset estimation layer is used to calculate the offset estimation result of each data batch in combination with the offset probability.

[0184] The using the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar environmental parameter data to obtain a parameter data offset estimation result includes:

[0185] Use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar environmental parameter data to obtain the offset estimation result of any normalized data sequence, and integrate the offset estimation results of all normalized data sequences into a parameter data offset estimation result. The calculation process of the offset estimation result of the normalized data sequence X u is as follows:

[0186] S21: The sequence partitioning layer partitions the normalized data sequence X u into consecutive data batches:

[0187]

[0188] where:

[0189] represents the m-th data batch obtained by partitioning, and each data batch contains 3 consecutive normalized data values;

[0190] X u (3m) represents the 3m-th data value in the data sequence X u ;

[0191] S22: The difference calculation layer calculates the statistical differences between consecutive data batches, where the statistical difference between data batch and is

[0192]

[0193] where:

[0194] represents the extreme value difference between data batch and , max(·) represents obtaining the maximum value by traversal, and min(·) represents obtaining the minimum value by traversal;

[0195] represents the outlier difference between data batch and , represents the mean value of the abnormal data values after normalization calculation in the data batch;

[0196] S23: Convert the statistical differences of the data batches into offset probabilities, where the conversion formula for the statistical difference is:

[0197]

[0198] where:

[0199] represents the offset probability corresponding to the statistical difference ;

[0200] S24: Combine the offset probabilities to calculate the offset estimation results for each data batch; where the offset estimation result for the data batch X u m is:

[0201]

[0202] Among them:

[0203] represents the offset estimation result of the data batch ;

[0204] δ(X u (3m)) represents the offset estimation coefficient of the data value X u in the normalized data sequence X u (3m);

[0205] represents the standard deviation of the data batch ; represents the mean value of the data batch ;

[0206] S25: Integrate the offset estimation results of all data batches as the offset estimation result of the normalized data sequence X u ;

[0207] S3: Correct the preprocessed coke tar environmental parameter data according to the parameter data offset estimation result to obtain the corrected coke tar environmental parameter data.

[0208] The step S3 includes:

[0209] Correct the preprocessed coke tar environmental parameter data according to the parameter data offset estimation result to obtain the correction result of any normalized data sequence, and form the corrected coke tar environmental parameter data with the correction results, where the correction process of the normalized data sequence X u is as follows:

[0210] S31: Obtain the normalized data sequence X u ;

[0211] S32: Correct any data value in the normalized data sequence X u , where the correction formula for the data value X u (n) is:

[0212] Y u (n) = X u (n) · δ(X u (n))

[0213] Among them:

[0214] Y u (n) represents the correction result of the data value X u (n);

[0215] δ(X u(n)) represents the data value X u The offset estimation coefficient corresponding to (n);

[0216] S33: Construct the normalized data sequence X u The correction result Y u = (Y u (1), Y u (2),..., Y u (n),..., Y u (N)).

[0217] S4: Extract features from the corrected coke tar environmental parameter data to obtain the coke tar environmental feature vector.

[0218] The S4 step includes:

[0219] Extract features from the corrected coke tar environmental parameter data to obtain the coke tar environmental feature vector F, where the process of feature extraction is:

[0220] S41: Construct the data matrix C from the corrected coke tar environmental parameter data:

[0221] C = [Y1, Y2, Y3, Y4, Y5] T

[0222] Where:

[0223] Y1, Y2, Y3, Y4, Y5 represent the corrected data sequences corresponding to 5 environmental parameters, and T represents the transpose;

[0224] S42: Use the genetic algorithm to solve the influence weights of the environmental parameters, perform weighted processing on the data matrix C, and construct the data weighted matrix C':

[0225] C' = [w1Y1, w2Y2, w3Y3, w4Y4, w5Y5] T

[0226] Where:

[0227] w1, w2, w3, w4, w5 are the influence weights of 5 environmental parameters in turn;

[0228] S43: Perform eigen decomposition on the data weighted matrix C' to obtain the 5 largest eigenvalues of the data weighted matrix C' and the corresponding eigenvectors, and construct the 5 eigenvectors into the coke tar environmental feature vector F:

[0229] F = [α1, α2, α3, α4, α5]C

[0230] Where:

[0231] α1, α2, α3, α4, α5 represent the eigenvectors corresponding to the 5 largest eigenvalues of the data weighting matrix C'.

[0232] In step S42, the genetic algorithm is used to solve the influence weights of environmental parameters, including:

[0233] S421: Initialize D groups of chromosomes. Each group of chromosomes has 5 segments of genes, respectively representing the influence weights of 5 environmental parameters. The initialized d-th group of chromosomes is:

[0234]

[0235] Where:

[0236] represents the initialized d-th group of chromosomes, represents the chromosome in 5 segments of genes;

[0237] S422: Set the current iteration number of the chromosome as t. The initial value of t is 0, and the maximum value is Max. Then the t-th iteration result of the d-th group of chromosomes is

[0238] S423: Set the chromosome genetic objective function G1(·). Substitute the iterated chromosome into the chromosome genetic objective function G1(·) to obtain the chromosome genetic objective function value of the chromosome. Among them, the chromosome corresponding chromosome genetic objective function value is

[0239]

[0240] Where:

[0241] represents using the chromosome to weight and eigen-decompose the data matrix C, and the formed coke tar environmental eigenvector;

[0242] ||·|| represents the L1 norm, and ||·||2 represents the L2 norm; in the embodiments of the present invention, represents the richness of the feature information in;

[0243] S424: Set the chromosome mutation objective function G2(·). Substitute the iterated chromosome into the chromosome mutation objective function G1(·) to obtain the chromosome mutation objective function value of the chromosome. Among them, the chromosome corresponding chromosome mutation objective function value is:

[0244]

[0245] Among them:

[0246] represents a sequence composed of 5 groups of eigenvectors obtained by weighting and eigen-decomposing the data matrix C using chromosomes ;

[0247] S425: Calculate the mutation probability of a chromosome according to the chromosome mutation objective function value of the chromosome; among them, the chromosome The mutation probability of undergoing mutation is:

[0248]

[0249] Among them:

[0250] represents the chromosome The mutation probability of undergoing mutation;

[0251] S426: Extract the mutated chromosome as the mutant chromosome, and the chromosome whose chromosome genetic objective function value is higher than that of the mutant chromosome as the genetic chromosome, and perform crossover processing on some gene segments in the mutant chromosome and the corresponding gene segments of the genetic chromosome;

[0252] S427: Let d = d + 1, return to step S425 until the maximum number of iterations is reached, and use the chromosome with the highest chromosome genetic objective function value as the solution result, and extract the influence weight of the environmental parameters.

[0253] S5: Construct a tar oil monitoring anomaly recognition model for tar oil environmental monitoring. The tar oil monitoring anomaly recognition model takes the tar oil environmental feature vector as the input and the monitoring anomaly recognition result as the output, and performs safety disposal according to the monitoring anomaly recognition result.

[0254] In the S5 step, constructing a tar oil monitoring anomaly recognition model for tar oil environmental monitoring includes:

[0255] Construct a tar oil monitoring anomaly recognition model. The tar oil monitoring anomaly recognition model takes the extracted tar oil environmental feature vector as the input and the monitoring anomaly recognition result as the output. The tar oil monitoring anomaly recognition model includes an input layer, a feature convolution layer, a feature weight calculation layer, and a mapping layer;

[0256] The input layer is used to receive the tar oil environmental feature vector;

[0257] The feature convolution layer is a convolutional neural network model structure for performing multi-level convolution processing on the tar oil environmental feature vector;

[0258] The feature weight calculation layer includes an attention unit and a residual unit in the residual network, and is used to calculate the weights of convolutional features;

[0259] The mapping layer performs clustering fusion processing on the convolutional features based on the weights, and maps the clustering fusion features to the monitoring anomaly recognition results;

[0260] The kerosene monitoring anomaly recognition model is used to monitor the kerosene environment feature vector, and the kerosene environment monitoring process is as follows:

[0261] S51: The input layer receives the kerosene environment feature vector F;

[0262] S52: The feature convolution layer performs multi-level convolution processing on the kerosene environment feature vector F to obtain multi-level convolution features:

[0263] F(k) = W k×k *F, k ∈ [1, K]

[0264] Where:

[0265] F(k) represents the convolution feature of the kerosene environment feature vector F at the k-scale level, and K represents the preset maximum scale level; W k×k represents the convolution weight matrix of k rows and k columns, and * represents the convolution operator;

[0266] S53: The feature weight calculation layer calculates the weights of the convolution features, and the weight of F(k) is:

[0267]

[0268] Where:

[0269] tanh(F(k - 1)) represents the residual information at the k - 1 scale level;

[0270] weight(k) represents the weight of F(k);

[0271] S54: The mapping layer performs weighted processing on the convolution features based on the weights, clusters the weighted convolution features, obtains several clustering clusters, extracts the clustering centers of each clustering cluster for splicing to obtain the clustering fusion feature G, and maps the clustering fusion feature G to the monitoring anomaly recognition results:

[0272]

[0273] Where:

[0274] W represents the mapping matrix;

[0275] V represents the monitoring anomaly recognition result. The higher the V value, the higher the leakage concentration of kerosene and the more serious the environmental pollution;

[0276] Perform safety disposal according to the monitoring anomaly recognition result.

[0277] Embodiment 2:

[0278] As Figure 2 shown, it is a functional module diagram of the intelligent monitoring system for the safety disposal of coke tar oil provided by an embodiment of the present invention, which can implement the intelligent monitoring method for the safety disposal of coke tar oil in Embodiment 1.

[0279] The intelligent monitoring system 100 for the safety disposal of coke tar oil according to the present invention can be installed in an electronic device. According to the functions achieved, the intelligent monitoring system for the safety disposal of coke tar oil can include a data acquisition module 101, a feature extraction module 102, and an environmental monitoring device 103. The modules described in the present invention can also be referred to as units, which refer to a series of computer program segments that can be executed by the processor of the electronic device and can complete fixed functions, and are stored in the memory of the electronic device.

[0280] The data acquisition module 101 is used to collect the environmental parameter data of the coke tar oil storage environment and perform preprocessing to obtain the preprocessed coke tar oil environmental parameter data, construct a parameter data offset detection model, use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar oil environmental parameter data, and correct the preprocessed coke tar oil environmental parameter data according to the estimated parameter data offset estimation result;

[0281] The feature extraction module 102 is used to extract features from the corrected coke tar oil environmental parameter data to obtain a coke tar oil environmental feature vector;

[0282] The environmental monitoring device 103 is used to construct a coke tar oil monitoring anomaly recognition model for coke tar oil environmental monitoring.

[0283] Specifically, each module in the intelligent monitoring system 100 for the safety disposal of coke tar oil in the embodiment of the present invention uses the same technical means as those described in the above Figure 1 intelligent monitoring method for the safety disposal of coke tar oil and can produce the same technical effects, which will not be elaborated here.

[0284] It should be understood that the above embodiments are only for illustration purposes and are not limited by this structure in the scope of the patent application.

[0285] It should be noted that the serial numbers of the above embodiments of the present invention are only for description and do not represent the superiority or inferiority of the embodiments. And the term "comprising" or "including" or any other variant thereof in this text is intended to cover non-exclusive inclusion, so that a process, device, article or method comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such a process, device, article or method. Without further limitation, an element defined by the statement "comprising one..." does not exclude the existence of another identical element in the process, device, article or method comprising such an element.

[0286] Through the description of the above embodiments, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation. Based on such an understanding, the technical solution of the present invention, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium as described above (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions for causing a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0287] The above are only the preferred embodiments of the present invention, and do not limit the patent scope of the present invention accordingly. Any equivalent structure or equivalent process transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in other related technical fields, shall be equally included in the patent protection scope of the present invention.

Claims

1. An intelligent monitoring method for the safe disposal of coking kerosene, characterized in that, The method includes: S1: Collect environmental parameter data of the coke tar storage environment and perform preprocessing to obtain preprocessed coke tar environmental parameter data, where the environmental parameters include temperature, humidity, pressure, liquid level, and benzene compound concentration; S2: Construct a parameter data offset detection model, and use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar environmental parameter data; The parameter data offset detection model is used to estimate the parameter data offset of the preprocessed coking coal tar environmental parameter data, obtain the offset estimation results of any normalized data sequence, and integrate the offset estimation results of all normalized data sequences into the parameter data offset estimation results, where the normalized data sequence is X u The calculation process of the offset estimation result is as follows: S21: The sequence partitioning layer partitions the normalized data sequence X u into consecutive data batches: Wherein: Denote the m-th data batch obtained by partitioning, where each data batch contains 3 consecutive normalized data values; X u (3m) represents the 3m-th data value in the data sequence X u ; S22: The difference calculation layer calculates the statistical differences between consecutive data batches, where the data batches and The statistical difference between them is S23: Convert the statistical difference of the data batch into an offset probability, where the statistical difference The corresponding offset probability is S24: Calculate the offset estimation result for each data batch in combination with the offset probability; wherein the offset estimation result of the data batch is as follows: Wherein: Indicates the data batch of the offset estimation result; δ(X u (3m)) represents the normalized data sequence X u in the data value X u (3m) offset estimation coefficient; Represents the standard deviation of the data batch , and represents the mean of the data batch . S25: Integrate the offset estimation results of all data batches as the offset estimation result of the normalized data sequence X u ; S3: Correct the preprocessed coke tar environmental parameter data according to the estimated parameter data offset estimation result to obtain corrected coke tar environmental parameter data; S4: Extract features from the corrected coke tar environmental parameter data to obtain a coke tar environmental feature vector; S5: Construct a coke tar monitoring anomaly recognition model for coke tar environmental monitoring. The coke tar monitoring anomaly recognition model takes the extracted coke tar environmental feature vector as input and the monitoring anomaly recognition result as output, and performs safety disposal according to the monitoring recognition result.

2. The intelligent monitoring method for the safe disposal of coke tar oil according to claim 1, wherein, In the step S1, collecting environmental parameter data of the coke tar storage environment and performing preprocessing includes: Collect environmental parameter data of the coke tar storage environment, where the environmental parameters include temperature, humidity, pressure, liquid level, and benzene compound concentration. The liquid level represents the liquid surface height of the coke tar in the container, the pressure represents the static liquid pressure at the bottom of the coke tar storage container, and the benzene compound represents the main chemical components of the coke tar. The concentration of benzene compounds is measured by gas chromatography. The form of the collected environmental parameter data is: x = {x u = (x u (1), x u (2),..., x u (n),..., x u (N)) | u ∈ [1, 5]} Wherein: x represents environmental parameter data; x u The data sequence representing the u-th environmental parameter, where u ∈ [1, 5], and the 1st - 5th environmental parameters are temperature, humidity, pressure, liquid level, and benzene concentration in sequence; x u (1), x u (2),..., x u (n),..., x u (N) represents the data value of the u-th environmental parameter at N monitoring times, where n ∈ [1, N]; Perform preprocessing on the environmental parameter data x to obtain preprocessed coke tar environmental parameter data, where the preprocessing process is: S11: Filter the data sequence in the environmental parameter data x to obtain the filtered data sequence, where the data sequence is x u The filtering result of which is x ' u ; S12: Calculate the distance between any two data values in the filtered data sequence; S13: Calculate the anomaly information coefficient of any data value in the filtered data sequence, and mark the data value with an anomaly information coefficient higher than the preset threshold as an abnormal data value; S14: Select the maximum and minimum non-abnormal data values from the filtered data sequence, and perform normalization processing on all data values in the filtered data sequence to obtain the normalized data sequence. The normalized data sequence of the u-th environmental parameter is X u =(X u (1), X u (2),..., X u (n),..., X u (N)), where X u (n) represents the processing result of the data value x u (n) after filtering and normalization processing; Take all the normalized data sequences as the preprocessed coke tar environmental parameter data.

3. The intelligent monitoring method for safe disposal of coke tar oil according to claim 1, characterized in that In the step S2, constructing a parameter data offset detection model includes: Construct a parameter data offset detection model, and use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar environmental parameter data. The parameter data offset detection model takes the preprocessed coke tar environmental parameter data as input and the parameter data offset estimation result as output. The parameter data offset detection model includes a sequence division layer, a difference calculation layer, and an offset estimation layer; The sequence division layer is used to divide the normalized data sequence into continuous data batches; The difference calculation layer is used to calculate the statistical difference between continuous data batches and convert the statistical difference into an offset probability; The offset estimation layer is used to calculate the offset estimation result of each data batch in combination with the offset probability.

4. The intelligent monitoring method for safe disposal of coke tar oil according to claim 1, characterized in that, In the step S3, correcting the preprocessed coke tar environmental parameter data according to the estimated parameter data offset estimation result includes: Based on the parameter data offset estimation result obtained from the estimation, the preprocessed coking coal tar environment parameter data is corrected to obtain the correction result of any normalized data sequence, and the correction results are used to form the corrected coking coal tar environment parameter data, where the normalized data sequence is X u The correction process is as follows: S31: Obtain the normalized data sequence X u ; S32: Correct any data value in the normalized data sequence X u where the data value X u (n) has the following correction formula: Y u y(n) = X u (n) · δ(X u (n)) Wherein: Y u (n) represents the data value X u (n)'s correction result; δ(X u (n)) represents the offset estimation coefficient corresponding to the data value X u (n); S33: Construct the normalized data sequence X u The corrected result Y u =(Y u (1), Y u (2),..., Y u (n),..., Y u (N)).

5. The intelligent monitoring method for safe disposal of coke tar oil according to claim 4, characterized in that, In the step S4, extracting features from the corrected coke tar environmental parameter data includes: Extract the features of the corrected environmental parameter data of tar oil to obtain the tar oil environmental feature vector F, where the feature extraction process is as follows: S41: Construct a data matrix C from the corrected environmental parameter data of tar oil; S42: Use the genetic algorithm to solve the influence weights of environmental parameters to form the data weighting matrix C ' ; S43: Weight the data matrix C ' Perform eigen decomposition on it to obtain the five largest eigenvalues of the data weighting matrix C ' and the corresponding eigenvectors, and form the feature vector F of the kerosene environment with the five eigenvectors.

6. The intelligent monitoring method for safe disposal of coke tar oil according to claim 5, characterized in that, In step S42, the genetic algorithm is used to solve the influence weights of environmental parameters, including: S421: Initialize the chromosomes in Group D. Each group of chromosomes contains 5 segments of genes, which respectively represent the influence weights of 5 environmental parameters. The d-th group of chromosomes generated by initialization is S422: Set the current iteration number of the chromosome to t, where the initial value of t is 0 and the maximum value is Max. Then, the result of the t-th iteration of the d-th group of chromosomes is S423: Set the chromosomal genetic objective function G1(·), substitute the iteratively obtained chromosome into the chromosomal genetic objective function G1(·), and obtain the chromosomal genetic objective function value of the chromosome, where the chromosome The corresponding chromosomal genetic objective function value is S424: Set the chromosome variation objective function G2(·), substitute the iteratively obtained chromosome into the chromosome variation objective function G1(·) to obtain the chromosome variation objective function value of the chromosome, where the chromosome The corresponding chromosome variation objective function value is S425: Calculate the mutation probability of a chromosome based on the objective function value of the chromosomal variation of the chromosome; wherein the chromosome The mutation probability of undergoing mutation is as follows: Among them: Indicates a chromosome The mutation probability for mutation S426: Extract the mutated chromosomes as mutant chromosomes, and the chromosomes with chromosome genetic objective function values higher than those of the mutant chromosomes as genetic chromosomes, and perform crossover processing on some gene segments in the mutant chromosomes and the corresponding gene segments of the genetic chromosomes; S427: Let d = d + 1, return to step S425 until the maximum number of iterations is reached, and take the chromosome with the highest chromosome genetic objective function value as the solution result to extract the influence weights of environmental parameters.

7. The intelligent monitoring method for safe disposal of coke tar oil according to claim 1, wherein In step S5, a tar oil monitoring anomaly recognition model is constructed for tar oil environmental monitoring, including: Construct a tar oil monitoring anomaly recognition model, which takes the extracted tar oil environmental feature vector as input and the monitoring anomaly recognition result as output. The tar oil monitoring anomaly recognition model includes an input layer, a feature convolution layer, a feature weight calculation layer, and a mapping layer; The input layer is used to receive the tar oil environmental feature vector; The feature convolution layer is a convolutional neural network model structure for performing multi-level convolution processing on the tar oil environmental feature vector; The feature weight calculation layer includes an attention unit and a residual unit in the residual network for calculating the weights of convolutional features; The mapping layer performs clustering fusion processing on the convolutional features based on the weights and maps the clustering fusion features to the monitoring anomaly recognition result; Use the tar oil monitoring anomaly recognition model to perform tar oil environmental monitoring on the tar oil environmental feature vector, where the tar oil environmental monitoring process is as follows: S51: The input layer receives the tar oil environmental feature vector F; S52: The feature convolution layer performs multi-level convolution processing on the tar oil environmental feature vector F to obtain multi-level convolution features: F(k) = W k×k *F, k ∈ [1, K] Among them: F(k) represents the convolution feature of the tar oil environmental feature vector F at the k-scale level, and K represents the preset maximum scale level; W k×k represents a k-by-k convolutional weight matrix, and * represents the convolution operator; S53: The feature weight calculation layer calculates the weights of the convolutional features, where the weight of F(k) is weight(k); S54: The mapping layer performs weighted processing on the convolutional features based on the weights, clusters the weighted convolutional features to obtain several clustering clusters, extracts and splices the clustering centers of each clustering cluster to obtain the clustering fusion feature G, and maps the clustering fusion feature G to the monitoring anomaly recognition result: Among them: W represents the mapping matrix; V represents the monitoring anomaly recognition result. The higher the V value, the higher the leakage concentration of tar oil and the more serious the environmental pollution; Perform safety disposal according to the monitoring and recognition results.

8. The intelligent monitoring method for safe disposal of coke tar oil according to claim 2, wherein, In step S13, calculate the anomaly information coefficient of any data value in the filtered data sequence, including: Data value x ' u (n) The calculation process of the abnormal information coefficient is as follows: S131: Traverse to get x ' u Middle distance data value x ' u (n) The nearest s-th data value, and calculate the distance between the two as the data value x ' u (n) is the nearest neighbor distance, and x ' u The data value x ' u (n) The data value whose distance is less than the nearest neighbor distance constitutes the data value x ' u (n)’s nearest neighbor data set Near(x ' u (n)); S132: Calculate the data value x ' u The density ρ(x ' u (n)); S133: Calculate the data value x ' u (n) anomaly information coefficient S(x ' u (n)): Among them: ρ(Near(x ' u (n),e)) represents the density of the e-th data value Near(x ' u (n)) in the nearest neighbor data set Near(x ' u (n),e).

9. An intelligent monitoring system for the safe disposal of coke tar oil, characterized in that, The system includes: A data acquisition module, which is used to collect the environmental parameter data of the coke tar storage environment, preprocess it to obtain the preprocessed coke tar environmental parameter data, construct a parameter data offset detection model, use the parameter data offset detection model to estimate the parameter data offset of the preprocessed coke tar environmental parameter data, and correct the preprocessed coke tar environmental parameter data according to the estimated parameter data offset estimation result; A feature extraction module, which is used to extract features from the corrected coke tar environmental parameter data to obtain a coke tar environmental feature vector; An environmental monitoring device, which is used to construct a coke tar monitoring anomaly recognition model for coke tar environmental monitoring, and perform safety disposal on the coke tar according to the monitoring and recognition result, so as to implement an intelligent monitoring method for coke tar safety disposal as described in any one of claims 1-8.

Images