Insulating gas decomposer concentration determination method, apparatus and device, and storage medium

Abstract

The invention relates to the technical field of gas decomposer analysis, in particular to an insulating gas decomposer concentration determination method, device and equipment and a storage medium. Segmenting and collecting the plurality of gas concentration data queues into a plurality of segmented data sets, and constructing a frequency domain matrix by a plurality of frequency domain feature vectors extracted according to each segmented data set; performing convolution and pooling operation on each frequency domain matrix to obtain a feature matrix; and finally, sequentially extracting matrixes from the plurality of characteristic matrixes, inputting the matrixes into a gas concentration prediction model, and when the fluctuation of the output of the model is smaller than a threshold value, determining the concentration of various gases according to the output of the model. According to the method, the characteristic matrix is input for multiple times, and the concentration of the insulating gas decomposer is determined when the output value is stable, so that the stable value of the sensor can be obtained in time, the influence of different gases on the sensor is eliminated, and the concentration result is more accurate.

Description

Technical Field

[0001] The present invention relates to the technical field of gas decomposition product analysis, and in particular to a method, device, equipment and storage medium for determining the concentration of decomposition products of insulating gas. Background Art

[0002] SF6 gas is a colorless, odorless, non-toxic and non-flammable inert gas. Due to its strong electronegativity, unique insulation properties and efficient arc extinguishing performance, it has become one of the most important insulating media in the power industry and has been widely used in gas insulated switchgear (GIS).

[0003] In GIS electrical equipment, the insulation performance of SF6 gas / mixed gas is directly related to the safe operation of the equipment. By detecting the concentration of SF6 gas / mixed gas / mixed gas, potential insulation failures can be discovered in time, thereby avoiding safety accidents caused by insulation problems of the equipment. Among them, the physical indicators such as the density pressure of SF6 gas / mixed gas and the purity, micro-water content, and decomposition products in the gas chamber are important bases for judging the operating status of GIS electrical equipment. By detecting these indicators, it can be determined whether there is leakage inside the GIS electrical equipment, whether the equipment is well sealed, etc., so as to timely discover and deal with potential safety hazards.

[0004] There are many methods for detecting insulating gas and decomposition products of high-voltage electrical equipment, such as electrochemical method, chromatography, and spectroscopy. At present, the mainstream method for detecting insulating gas and decomposition products is to detect different gas components through multiple sensors. The accuracy of the detection results returned by the sensor is mainly affected by two factors: gas diffusion and the effect of other gas components. The effect of gas diffusion means that after the gas composition changes, it takes a period of diffusion for the sensor to return a more accurate detection value; the effect of other gas components means that different gas components will affect the detection value. For example, in an SF6 / N2 mixed gas environment, the presence of N2 may interfere with the detection of certain decomposition products.

[0005] At present, some technologies use noise removal algorithms or data comparison reference methods to correct the detection results of insulating gas and decomposition products. The advantages of these methods are that the algorithms are relatively simple and easy to implement. The disadvantage is that they have very limited effect on improving the accuracy and timeliness of the detection results of insulating gas and decomposition products.

[0006] Based on this, it is necessary to develop and design a method for determining the concentration of insulating gas decomposition products. Summary of the invention

[0007] The embodiments of the present invention provide a method, device, equipment and storage medium for determining the concentration of decomposition products of insulating gas, which are used to solve the problem in the prior art that the detection results of insulating gas and decomposition products are inaccurate due to the influence of other gas components.

[0008] In a first aspect, an embodiment of the present invention provides a method for determining the concentration of decomposition products of an insulating gas, comprising:

[0009] Acquire multiple gas concentration data queues, wherein the gas concentration data queues represent characteristics of gas concentration changing over time;

[0010] The plurality of gas concentration data queues are segmented and grouped into a plurality of segmented data sets, and a plurality of frequency domain feature vectors extracted from each segmented data set are constructed into a frequency domain matrix, wherein each segmented data set includes a data segment extracted from the plurality of gas concentration data queues, the plurality of data segments in the segmented data set correspond to the same segmented time period, and the frequency domain feature vector is extracted according to the data segment;

[0011] Perform convolution and pooling operations on each frequency domain matrix to obtain a feature matrix;

[0012] Matrix inputs are sequentially extracted from a plurality of feature matrices and fed into a gas concentration prediction model. When the fluctuation of the output of the model is less than a threshold, the concentrations of the plurality of gases are determined according to the output of the model.

[0013] In a possible implementation, the segmenting and grouping of the plurality of gas concentration data queues into a plurality of segmented data sets, and constructing a plurality of frequency domain feature vectors extracted from each segmented data set into a frequency domain matrix, includes:

[0014] Get the basic cycle;

[0015] Determine the length of the segmented period according to the basic cycle duration, wherein the length of the segmented period is an integer multiple of the basic cycle duration;

[0016] For each split period, perform the following steps:

[0017] extracting data segments from the plurality of gas concentration data queues according to segmentation time periods to construct segmented data sets;

[0018] According to a preset gas concentration sequence, extracting data segments from the segmented data set as data segments to be processed;

[0019] A plurality of frequency domain features of the data segment to be processed are extracted according to a first formula, and the extracted plurality of frequency domain features are constructed into a frequency domain feature vector, wherein the first formula is:

[0020]

[0021] Wherein, GF(k) is the kth frequency domain feature, qnMax is the total number of data in the data segment to be processed, QCD(qn) is the qnth data in the data segment to be processed, e is a natural constant, j is an imaginary unit, π is the pi, ω0 is the frequency corresponding to the basic cycle, and tN is the number of gas concentration data obtained within the duration of the basic cycle;

[0022] Adding the frequency domain eigenvector to the frequency domain matrix;

[0023] If the traversal of the segmented data set is not completed, the process jumps to the step of extracting data segments from the segmented data set as data segments to be processed according to the preset gas sequence.

[0024] In a possible implementation, performing convolution and pooling operations on each frequency domain matrix to obtain a feature matrix includes:

[0025] For each frequency domain matrix, perform the following steps:

[0026] Get the convolution operator;

[0027] Extracting a data block of the same type as the convolution operator from the frequency domain matrix according to the position index as a data block to be processed;

[0028] Performing a dot product operation on the convolution operator and the data block to be processed to obtain a dot product result;

[0029] Adding the dot product result to the convolution matrix according to the position index;

[0030] If the traversal of the frequency domain matrix is ​​not completed, the position index is offset according to the preset step length, and the process jumps to the step of extracting a data block of the same type as the convolution operator from the frequency domain matrix according to the position index as the data block to be processed;

[0031] Otherwise, perform maximum pooling on the convolution matrix to obtain a feature matrix.

[0032] In a possible implementation, the matrix is ​​extracted from the plurality of feature matrices in sequence and inputted into the gas concentration prediction model. When the fluctuation of the output of the model is less than a threshold, the concentrations of the plurality of gases are determined according to the output of the model, including:

[0033] Extract matrices from multiple feature matrices in sequence as the matrix to be input;

[0034] Inputting the matrix to be input into the gas concentration prediction model to obtain a plurality of second outputs;

[0035] The output deviation rate is determined according to the second formula, the plurality of second outputs and the plurality of first outputs, wherein the plurality of first outputs are the plurality of outputs of the gas concentration prediction model in the previous time, and the second formula is:

[0036]

[0037] Where Bias is the output deviation rate, is the cnth second output, is the cnth first output, cnMax is the total number of output nodes of the gas concentration prediction model;

[0038] Adding the output deviation rate to a deviation rate queue;

[0039] If the values ​​of the last multiple bits of the deviation rate queue are all less than the deviation rate threshold, the multiple second outputs are used as the concentrations of the multiple gases;

[0040] Otherwise, jump to the step of sequentially extracting matrices from a plurality of feature matrices as matrices to be input.

[0041] In a possible implementation manner, the gas concentration prediction model is constructed according to a plurality of historical characteristic matrices, and the construction process of the gas concentration prediction model includes:

[0042] Obtain a prediction basic model, multiple historical feature matrices, and multiple historical gas concentration arrays, wherein each historical feature matrix corresponds to a historical gas concentration array, and the time node corresponding to the historical gas concentration array is later than the time node corresponding to the historical feature matrix;

[0043] Randomly generate a plurality of coefficient arrays, and substitute the plurality of coefficient arrays into the prediction basic model to generate a plurality of intermediate models;

[0044] Taking out a matrix from the plurality of historical feature matrices according to the order of time nodes as the historical feature matrix to be input;

[0045] Inputting the historical feature matrices to be input into the multiple intermediate models respectively to obtain multiple output arrays, wherein each output array corresponds to an intermediate model;

[0046] Determine a plurality of model deviations according to the plurality of output arrays and a target historical gas concentration array, wherein the target historical gas concentration array is a historical gas concentration array corresponding to the historical feature matrix to be input;

[0047] Adding the multiple model deviations to corresponding deviation queues respectively;

[0048] If there is a queue whose last few digits are smaller than the deviation threshold among the multiple deviation queues, the intermediate model corresponding to the queue whose last few digits are smaller than the deviation threshold is used as the gas concentration prediction model;

[0049] Otherwise, the coefficients of the multiple intermediate models are adjusted respectively according to the multiple deviation queues, and the process jumps to the step of taking out matrices from the multiple historical feature matrices according to the order of time nodes as the historical feature matrix to be input.

[0050] In a possible implementation, the prediction basic model is:

[0051]

[0052] In the formula, QC cn is the cnth output of the predicted basic model, mMax is the total number of rows of intermediate nodes, nMax is the total number of columns of intermediate nodes, and wo cn,a is the ath coefficient of the cnth output node, wh m,n,a is the ath coefficient of the node in the mth row and nth column, h a,n-1 is the output of the intermediate node in row a and column n-1, n k is the column number of the input node, who m,n,a is the ath feeding coefficient of the mth row feeding into the column node, x b is the bth input variable, wi m,1,b is the b-th coefficient of the middle node in the first column of the m-th row, bmax is the total number of input variables, e is a natural constant, and C is the bias coefficient.

[0053] In a possible implementation manner, adjusting the coefficients of the multiple intermediate models respectively according to the multiple deviation queues includes:

[0054] For each intermediate model, perform the following steps:

[0055] According to the minimum value in the deviation queue, determine the historical optimal coefficient array;

[0056] Determine a global optimal coefficient array according to the minimum value of the last bits of the plurality of deviation queues;

[0057] The coefficients of the intermediate model are adjusted according to the third formula, wherein the third formula is:

[0058]

[0059] In the formula, is the ath coefficient of the intermediate model before adjustment, w a,opta is the ath coefficient of the global optimal coefficient array, w a,opth is the ath coefficient of the historical optimal coefficient array, is the global optimal distance, is the historical optimal distance, k1 is the first coefficient, k2 is the second coefficient, k a Substitute the coefficients for the first time, k h Substituting the coefficients for the second time, is the ath coefficient of the adjusted intermediate model, and distb is the basic disturbance.

[0060] In a second aspect, an embodiment of the present invention provides an insulating gas decomposition product concentration device, which is used to implement the insulating gas decomposition product concentration determination method described in the first aspect or any possible implementation of the first aspect, wherein the insulating gas decomposition product concentration device includes:

[0061] A gas concentration data acquisition module, used to acquire multiple gas concentration data queues, wherein the gas concentration data queues represent the characteristics of gas concentration changing over time;

[0062] A frequency domain analysis module, used to divide the queue period into a plurality of segmented periods, and construct a plurality of frequency domain feature vectors extracted from each segmented data set into a frequency domain matrix, wherein each segmented data set includes a plurality of data segments extracted from the gas concentration data queue, the plurality of data segments in the segmented data set correspond to the same segmented period, and the frequency domain feature vectors are extracted according to the data segments;

[0063] A feature extraction module is used to perform convolution and pooling operations on each frequency domain matrix to obtain a feature matrix;

[0064] as well as,

[0065] The gas concentration determination module is used to extract matrices from multiple feature matrices in sequence and input them into the gas concentration prediction model. When the fluctuation of the output of the model is less than a threshold, the concentrations of multiple gases are determined according to the output of the model.

[0066] In a third aspect, an embodiment of the present invention provides an electronic device, comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, and when the processor executes the computer program, it implements the steps of the method described in the first aspect or any possible implementation method of the first aspect.

[0067] In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method described in the first aspect or any possible implementation of the first aspect are implemented.

[0068] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0069] The embodiment of the present invention discloses a method for determining the concentration of insulating gas decomposition products. The present invention first obtains multiple gas concentration data queues, wherein the gas concentration data queues characterize the characteristics of gas concentration changing over time; then the multiple gas concentration data queues are segmented and grouped into multiple segmented data sets, and multiple frequency domain feature vectors extracted from each segmented data set are constructed into a frequency domain matrix, wherein each segmented data set includes multiple data segments extracted from the gas concentration data queue, multiple data segments in the segmented data set correspond to the same segmented time period, and the frequency domain feature vectors are extracted according to the data segments; then convolution and pooling operations are performed on each frequency domain matrix to obtain a feature matrix; finally, matrices are sequentially extracted from multiple feature matrices and input into a gas concentration prediction model, and when the fluctuation of the output of the model is less than a threshold, the concentrations of multiple gases are determined according to the output of the model. The present invention extracts frequency domain features and constructs a frequency domain matrix from the data obtained by segmenting the gas concentration data queue, so that the time series information is frequency-domained, so that the data dimension is reduced, and the frequency domain matrix includes more factor information, which provides data guarantee for determining the concentration of decomposition products. The present invention reduces the data dimension through convolution and pooling, reduces the amount of calculation and prevents overfitting, so that the data is more accurate. The embodiment of the present invention determines the concentration of the decomposition products of the insulating gas by inputting the characteristic matrix multiple times when the output value is stable. Therefore, the stable value of the sensor can be obtained in time, and the influence of different gases on the sensor is eliminated, and the concentration result is more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0071] Figure 1 is a flow chart of a method for determining the concentration of decomposition products of insulating gas provided in an embodiment of the present invention;

[0072] Figure 2 It is a schematic diagram of the construction process of the gas concentration prediction model provided by an embodiment of the present invention;

[0073] Figure 3 is a functional block diagram of an insulating gas decomposition product concentration device provided in an embodiment of the present invention;

[0074] Figure 4 It is a functional block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0075] In the following description, specific details such as specific system structures and technologies are provided for the purpose of illustration rather than limitation so as to provide a thorough understanding of the embodiments of the present invention. However, it should be clear to those skilled in the art that the present invention may be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, and methods are omitted so as not to obscure the description of the present invention with unnecessary details.

[0076] In order to make the purpose, technical solutions and advantages of the present invention more clear, a specific implementation method will be described below in conjunction with the accompanying drawings.

[0077] The following is a detailed description of an embodiment of the present invention. This example is implemented based on the technical solution of the present invention, and provides a detailed implementation method and a specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0078] Figure 1 A flow chart of a method for determining the concentration of decomposition products of insulating gas provided in an embodiment of the present invention.

[0079] like Figure 1 As shown, it shows a flowchart of the method for determining the concentration of decomposition products of insulating gas provided by an embodiment of the present invention, which is described in detail as follows:

[0080] In step 101, a plurality of gas concentration data queues are acquired, wherein the gas concentration data queues represent the characteristics of gas concentration changing over time.

[0081] In step 102, the multiple gas concentration data queues are segmented and grouped into multiple segmented data sets, and multiple frequency domain feature vectors extracted from each segmented data set are constructed into a frequency domain matrix, wherein each segmented data set includes data segments extracted from multiple gas concentration data queues, multiple data segments in the segmented data set correspond to the same segmented time period, and the frequency domain feature vectors are extracted based on the data segments.

[0082] In some embodiments, segmenting and grouping the plurality of gas concentration data queues into a plurality of segmented data sets, and constructing a plurality of frequency domain feature vectors extracted from each segmented data set into a frequency domain matrix, comprises:

[0083] Get the basic cycle;

[0084] Determine the length of the segmented period according to the basic cycle duration, wherein the length of the segmented period is an integer multiple of the basic cycle duration;

[0085] For each split period, perform the following steps:

[0086] extracting data segments from the plurality of gas concentration data queues according to segmentation time periods to construct segmented data sets;

[0087] According to a preset gas concentration sequence, extracting data segments from the segmented data set as data segments to be processed;

[0088] A plurality of frequency domain features of the data segment to be processed are extracted according to a first formula, and the extracted plurality of frequency domain features are constructed into a frequency domain feature vector, wherein the first formula is:

[0089]

[0090] Wherein, GF(k) is the kth frequency domain feature, qnMax is the total number of data in the data segment to be processed, QCD(qn) is the qnth data in the data segment to be processed, e is a natural constant, j is an imaginary unit, π is the pi, ω0 is the frequency corresponding to the basic cycle, and tN is the number of gas concentration data obtained within the duration of the basic cycle;

[0091] Adding the frequency domain eigenvector to the frequency domain matrix;

[0092] If the traversal of the segmented data set is not completed, the process jumps to the step of extracting data segments from the segmented data set as data segments to be processed according to the preset gas sequence.

[0093] For example, as mentioned above, in current technical means, when the gas concentration changes, it takes a certain amount of time for the sensor's detection value to stabilize, and due to the presence of multiple gases, the mutual interference of the gases will affect the detection value returned by the sensor.

[0094] The embodiments of the present invention are intended to determine the stable value of the sensor detection value when it is stable from the early data stream returned by the sensor, and to eliminate the interference of other gases on the data returned by the sensor as much as possible.

[0095] The gas concentration data queue of the present invention is obtained based on the return value of the sensor. Generally speaking, one gas concentration data queue corresponds to one sensor, that is, corresponds to the concentration value of one gas.

[0096] In order to facilitate data flow analysis, the embodiment of the present invention divides each data flow (gas concentration data queue) multiple times, extracts frequency domain features from the data corresponding to the same time period, and constructs a frequency domain matrix, so that each time period corresponds to a frequency domain matrix. Since the matrix is ​​a combination of the transformation of time domain features and multiple gas concentration data features, the concentration determination method of the present invention is a gas concentration determination method based on time series and factor analysis.

[0097] In terms of frequency domain matrix construction, the present invention first obtains the basic cycle, which is generally determined by the duration of the insulating gas when an abnormality occurs. For example, an integer multiple of the duration of a typical discharge of a GIS device is used as the basic cycle. The length of the segmentation period is determined based on the duration of the basic cycle. Note that the length of the segmentation period is an integer multiple of the length of the basic cycle. Then, data segments are extracted from each gas concentration data queue according to the segmentation period, and the obtained multiple data segments are constructed as a segmentation data set.

[0098] Then, according to the preset gas concentration order, data segments are extracted from the segmented data set for frequency domain feature extraction. The first formula is used in one application scenario. The first formula is:

[0099]

[0100] Wherein, GF(k) is the kth frequency domain feature, qnMax is the total number of data in the data segment to be processed, QCD(qn) is the qnth data in the data segment to be processed, e is a natural constant, j is an imaginary unit, π is the pi, ω0 is the frequency corresponding to the basic cycle, and tN is the number of gas concentration data obtained within the duration of the basic cycle.

[0101] The frequency domain features obtained by the above formula are constructed as frequency domain feature vectors and added to the frequency domain matrix. Data segments are extracted from the segmented data set, and frequency domain vectors are extracted and added to the frequency domain matrix repeatedly to complete the construction of the frequency domain matrix.

[0102] In step 103, convolution and pooling operations are performed on each frequency domain matrix to obtain a feature matrix.

[0103] In some implementations, performing convolution and pooling operations on each frequency domain matrix to obtain a feature matrix includes:

[0104] For each frequency domain matrix, perform the following steps:

[0105] Get the convolution operator;

[0106] Extracting a data block of the same type as the convolution operator from the frequency domain matrix according to the position index as a data block to be processed;

[0107] Performing a dot product operation on the convolution operator and the data block to be processed to obtain a dot product result;

[0108] Adding the dot product result to the convolution matrix according to the position index;

[0109] If the traversal of the frequency domain matrix is ​​not completed, the position index is offset according to the preset step length, and the process jumps to the step of extracting a data block of the same type as the convolution operator from the frequency domain matrix according to the position index as the data block to be processed;

[0110] Otherwise, perform maximum pooling on the convolution matrix to obtain a feature matrix.

[0111] Exemplarily, if the concentration of gas decomposition products is determined according to the frequency domain matrix obtained in the above steps, the amount of calculation will be too large due to the large dimension of the frequency domain matrix. On the other hand, there are some low-value elements and continuous elements in the frequency domain vector in the frequency domain matrix, which have a relatively small impact on the result of determining the concentration of decomposition products. The embodiment of the present invention reduces the dimension of the frequency domain matrix obtained in the above steps through convolution and pooling operations.

[0112] First, we get the convolution operator. For example, a difference operator is:

[0113]

[0114] According to the obtained differential operator and position index, a data block of the same type as the convolution operator is extracted from the frequency domain matrix, and a dot product operation is performed on the data block to be processed to obtain a dot product result;

[0115] Then add the dot product result to the convolution matrix according to the position index. Then adjust the position of the position index and repeat the dot product result calculation steps to complete the convolution operation of the entire matrix and obtain the convolution matrix.

[0116] The convolution matrix is ​​subjected to maximum pooling to obtain the feature matrix.

[0117] Max Pooling is a downsampling operation. Its main purpose is to reduce the dimension of the input features, reduce the amount of data calculation and the number of parameters, and extract the most significant feature information.

[0118] Suppose we have a feature map of size (here and represent the number of rows and columns of the feature map respectively). We define a pooling window of size (usually or) and a stride. The pooling window starts from the upper left corner of the feature map and slides on the feature map according to the stride. For each sub-region covered by the pooling window, the maximum value in this sub-region is selected as the output value.

[0119] Max pooling can reduce the amount of calculation: especially when processing large-scale data, the amount of calculation is very large. Max pooling can reduce the data size, thereby reducing the amount of calculation of subsequent layers. Max pooling can prevent overfitting: overfitting refers to the situation where the model performs well on the training data but performs poorly on the test data. Max pooling can be regarded as an abstraction of the data. It only retains the most significant feature information and discards some local detail information, so that the model has better generalization ability and helps prevent overfitting. Max pooling can extract the main features: it can extract the main features in the data. Because max pooling selects the maximum value within the pooling window, these maximum values ​​usually represent the most significant features in the local area, which is beneficial to subsequent tasks such as classification.

[0120] In step 104, matrices are sequentially extracted from a plurality of feature matrices and input into a gas concentration prediction model. When the fluctuation of the output of the model is less than a threshold, the concentrations of the plurality of gases are determined according to the output of the model.

[0121] In some embodiments, extracting matrices from a plurality of feature matrices in sequence and inputting them into a gas concentration prediction model, and determining the concentrations of a plurality of gases according to the output of the model when the fluctuation of the output of the model is less than a threshold, comprises:

[0122] Extract matrices from multiple feature matrices in sequence as the matrix to be input;

[0123] Inputting the matrix to be input into the gas concentration prediction model to obtain a plurality of second outputs;

[0124] The output deviation rate is determined according to the second formula, the plurality of second outputs and the plurality of first outputs, wherein the plurality of first outputs are the plurality of outputs of the gas concentration prediction model in the previous time, and the second formula is:

[0125]

[0126] Where Bias is the output deviation rate, is the cnth second output, is the cnth first output, cnMax is the total number of output nodes of the gas concentration prediction model;

[0127] Adding the output deviation rate to a deviation rate queue;

[0128] If the values ​​of the last multiple bits of the deviation rate queue are all less than the deviation rate threshold, the multiple second outputs are used as the concentrations of the multiple gases;

[0129] Otherwise, jump to the step of sequentially extracting matrices from a plurality of feature matrices as matrices to be input.

[0130] In some embodiments, the gas concentration prediction model is constructed based on a plurality of historical feature matrices, and the construction process of the gas concentration prediction model includes:

[0131] Obtain a prediction basic model, multiple historical feature matrices, and multiple historical gas concentration arrays, wherein each historical feature matrix corresponds to a historical gas concentration array, and the time node corresponding to the historical gas concentration array is later than the time node corresponding to the historical feature matrix;

[0132] Randomly generate a plurality of coefficient arrays, and substitute the plurality of coefficient arrays into the prediction basic model to generate a plurality of intermediate models;

[0133] Taking out a matrix from the plurality of historical feature matrices according to the order of time nodes as the historical feature matrix to be input;

[0134] Inputting the historical feature matrices to be input into the multiple intermediate models respectively to obtain multiple output arrays, wherein each output array corresponds to an intermediate model;

[0135] Determine a plurality of model deviations according to the plurality of output arrays and a target historical gas concentration array, wherein the target historical gas concentration array is a historical gas concentration array corresponding to the historical feature matrix to be input;

[0136] Adding the multiple model deviations to corresponding deviation queues respectively;

[0137] If there is a queue whose last few digits are smaller than the deviation threshold among the multiple deviation queues, the intermediate model corresponding to the queue whose last few digits are smaller than the deviation threshold is used as the gas concentration prediction model;

[0138] Otherwise, the coefficients of the multiple intermediate models are adjusted respectively according to the multiple deviation queues, and the process jumps to the step of taking out matrices from the multiple historical feature matrices according to the order of time nodes as the historical feature matrix to be input.

[0139] In some embodiments, the prediction base model is:

[0140]

[0141] In the formula, QC cn is the cnth output of the predicted basic model, mMax is the total number of rows of intermediate nodes, nMax is the total number of columns of intermediate nodes, and wo cn,a is the ath coefficient of the cnth output node, wh m,n,a is the ath coefficient of the node in the mth row and nth column, h a,n-1 is the output of the intermediate node in row a and column n-1, n k is the column number of the input node, whom,n,a is the ath feeding coefficient of the mth row feeding into the column node, x b is the bth input variable, wi m,1,b is the b-th coefficient of the middle node in the first column of the m-th row, bmax is the total number of input variables, e is a natural constant, and C is the bias coefficient.

[0142] In some implementations, adjusting the coefficients of the plurality of intermediate models respectively according to the plurality of deviation queues comprises:

[0143] For each intermediate model, perform the following steps:

[0144] According to the minimum value in the deviation queue, determine the historical optimal coefficient array;

[0145] Determine a global optimal coefficient array according to the minimum value of the last bits of the plurality of deviation queues;

[0146] The coefficients of the intermediate model are adjusted according to the third formula, wherein the third formula is:

[0147]

[0148] In the formula, is the ath coefficient of the intermediate model before adjustment, w a,opta is the ath coefficient of the global optimal coefficient array, w a,opth is the ath coefficient of the historical optimal coefficient array, is the global optimal distance, is the historical optimal distance, k1 is the first coefficient, k2 is the second coefficient, k a Substitute the coefficients for the first time, k h Substituting the coefficients for the second time, is the ath coefficient of the adjusted intermediate model, and distb is the basic disturbance.

[0149] Exemplarily, the present invention sequentially inputs multiple characteristic matrices into a gas concentration prediction model, and when the fluctuation of the output of the model is less than a threshold, the concentrations of multiple gases are determined according to the output of the model. The model has inputs corresponding to multiple elements of the characteristic matrix and multiple outputs corresponding to gas concentration prediction values.

[0150] like Figure 2 As shown, this figure is a schematic diagram of the construction process of the gas concentration prediction model provided by an embodiment of the present invention.

[0151] The construction process of the gas concentration prediction model provided by the embodiment of the present invention first obtains a prediction basic model, multiple historical feature matrices 201, and multiple historical gas concentration arrays 202. In one application scenario, the prediction basic model is:

[0152]

[0153] In the formula, QC cn is the cnth output of the predicted basic model, mMax is the total number of rows of intermediate nodes, nMax is the total number of columns of intermediate nodes, and wo cn,a is the ath coefficient of the cnth output node, wh m,n,a is the ath coefficient of the node in the mth row and nth column, h a,n-1 is the output of the intermediate node in row a and column n-1, n k is the column number of the input node, who m,n,a is the ath feeding coefficient of the mth row feeding into the column node, x b is the bth input variable, wi m,1,b is the b-th coefficient of the middle node in the first column of the m-th row, bmax is the total number of input variables, e is a natural constant, and C is the bias coefficient.

[0154] We can see that this model has a lot of coefficients that need to be determined.

[0155] On the other hand, each historical feature matrix 201 has a one-to-one correspondence with each historical gas concentration array 202, and the time node corresponding to the historical gas concentration array 202 is later than the time node corresponding to the historical feature matrix 201;

[0156] Then, multiple coefficient arrays are randomly generated, and the multiple coefficient arrays are substituted into the prediction basic model to obtain multiple intermediate models 203. Then, matrices are taken out from multiple historical feature matrices 201 according to the order of time nodes and input into multiple intermediate models 203 respectively to obtain multiple output arrays 204. Since each historical feature matrix 201 has a one-to-one correspondence with each historical gas concentration array 202, we can determine the model deviation based on the output array 204 and the target historical gas concentration array. Note that the target historical gas concentration array is the historical gas concentration array corresponding to the input historical feature matrix. The obtained model deviation is added to the corresponding deviation queue 205. Note that each deviation queue 205 corresponds to an intermediate model 203.

[0157] If there is a queue whose last few bits are less than the deviation threshold, the intermediate model 203 corresponding to this queue is used as the gas concentration prediction model, and the iteration ends. If multiple deviation queues do not meet the above conditions, the coefficients of multiple intermediate models 203 are adjusted according to the deviation queue 205, and the steps of taking out matrices from multiple historical feature matrices 201 according to the order of time nodes and inputting them into the intermediate model 203 are repeated. Through the above iterative process, the model construction can be completed.

[0158] In terms of adjusting the coefficients of the intermediate model according to the multiple deviation queues, the embodiment of the present invention determines the historical optimal coefficient array according to the minimum value in each deviation queue; then determines the global optimal coefficient array according to the minimum value in the last position of the multiple deviation queues; and then adjusts the coefficients of the intermediate model according to the third formula:

[0159]

[0160] In the formula, is the ath coefficient of the intermediate model before adjustment, w a,opta is the ath coefficient of the global optimal coefficient array, w a,opth is the ath coefficient of the historical optimal coefficient array, is the global optimal distance, is the historical optimal distance, k1 is the first coefficient, k2 is the second coefficient, k a Substitute the coefficients for the first time, k h Substituting the coefficients for the second time, is the ath coefficient of the adjusted intermediate model, and distb is the basic disturbance.

[0161] The embodiment of the method for determining the concentration of decomposition products of insulating gas of the present invention first obtains multiple gas concentration data queues, wherein the gas concentration data queues characterize the characteristics of gas concentration changing over time; then the multiple gas concentration data queues are segmented and grouped into multiple segmented data sets, and multiple frequency domain feature vectors extracted from each segmented data set are constructed into a frequency domain matrix, wherein each segmented data set includes multiple data segments extracted from the gas concentration data queue, multiple data segments in the segmented data set correspond to the same segmented time period, and the frequency domain feature vectors are extracted according to the data segments; then convolution and pooling operations are performed on each frequency domain matrix to obtain a feature matrix; finally, matrices are sequentially extracted from multiple feature matrices and input into a gas concentration prediction model, and when the fluctuation of the output of the model is less than a threshold, the concentrations of multiple gases are determined according to the output of the model. The present invention extracts frequency domain features and constructs a frequency domain matrix from the data obtained by segmenting the gas concentration data queue, so that the time series information is frequency-domained, so that the data dimension is reduced, and the frequency domain matrix includes more factor information, which provides data guarantee for determining the concentration of decomposition products. The present invention reduces the data dimension through convolution and pooling, reduces the amount of calculation and prevents overfitting, and makes the data more accurate. The embodiment of the present invention determines the concentration of the decomposition products of the insulating gas by inputting the characteristic matrix multiple times when the output value is stable. Therefore, the stable value of the sensor can be obtained in time, and the influence of different gases on the sensor is eliminated, and the concentration result is more accurate.

[0162] It should be understood that the size of the serial numbers of the steps in the above implementation does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the implementation method of the present invention.

[0163] The following is an implementation of the device of the present invention. For details not described in detail, reference may be made to the corresponding method implementation described above.

[0164] Figure 3 is a functional block diagram of an insulating gas decomposition product concentration device provided in an embodiment of the present invention, referring to Figure 3 The insulating gas decomposition product concentration device includes: a gas concentration data acquisition module 301, a frequency domain analysis module 302, a feature extraction module 303 and a gas concentration determination module 304, wherein:

[0165] The gas concentration data acquisition module 301 is used to acquire a plurality of gas concentration data queues, wherein the gas concentration data queues represent the characteristics of gas concentration changing over time;

[0166] A frequency domain analysis module 302 is used to divide the queue period into a plurality of segmented periods, and construct a plurality of frequency domain feature vectors extracted from each segmented data set into a frequency domain matrix, wherein each segmented data set includes a plurality of data segments extracted from the gas concentration data queue, the plurality of data segments in the segmented data set correspond to the same segmented period, and the frequency domain feature vectors are extracted according to the data segments;

[0167] A feature extraction module 303 is used to perform convolution and pooling operations on each frequency domain matrix to obtain a feature matrix;

[0168] The gas concentration determination module 304 is used to sequentially extract matrices from multiple feature matrices and input them into the gas concentration prediction model. When the fluctuation of the output of the model is less than a threshold, the concentrations of multiple gases are determined according to the output of the model.

[0169] Figure 4 is a functional block diagram of an electronic device provided by an embodiment of the present invention. Figure 4 As shown, the electronic device 4 of this embodiment includes: a processor 400 and a memory 401, wherein the memory 401 stores a computer program 402 that can be run on the processor 400. When the processor 400 executes the computer program 402, the steps in the above-mentioned insulating gas decomposition product concentration determination method and the embodiment are implemented, for example Figure 1 Steps 101 to 104 are shown.

[0170] Exemplarily, the computer program 402 may be divided into one or more modules / units, and the one or more modules / units are stored in the memory 401 and executed by the processor 400 to implement the present invention.

[0171] The electronic device 4 may be a computing device such as a desktop computer, a notebook, a PDA, or a cloud server. The electronic device 4 may include, but is not limited to, a processor 400 and a memory 401. Those skilled in the art will appreciate that Figure 4 It is only an example of the electronic device 4 and does not constitute a limitation of the electronic device 4. It may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, the electronic device 4 may also include input and output devices, network access devices, buses, etc.

[0172] The processor 400 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0173] The memory 401 may be an internal storage unit of the electronic device 4, such as a hard disk or memory of the electronic device 4. The memory 401 may also be an external storage device of the electronic device 4, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device 4. Further, the memory 401 may also include both an internal storage unit and an external storage device of the electronic device 4. The memory 401 is used to store the computer program 402 and other programs and data required by the electronic device 4. The memory 401 may also be used to temporarily store data that has been output or is to be output.

[0174] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In actual applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the implementation method can be integrated into a processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method implementation method, which will not be repeated here.

[0175] In the above-mentioned embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0176] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0177] In the embodiments provided by the present invention, it should be understood that the disclosed devices / electronic devices and methods can be implemented in other ways. For example, the device / electronic device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0178] 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, that is, they 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 present embodiment.

[0179] In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0180] If the integrated module / unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present invention implements all or part of the processes in the above-mentioned implementation method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, it can implement the steps of the above-mentioned various methods and device implementation methods. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device that can carry the computer program code, recording medium, U disk, mobile hard disk, disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal and software distribution medium.

[0181] The above-described embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for determining the concentration of decomposition products of insulating gas, characterized in that: include: Acquire multiple gas concentration data queues, wherein the gas concentration data queues represent characteristics of gas concentration changing over time; The plurality of gas concentration data queues are segmented and grouped into a plurality of segmented data sets, and a plurality of frequency domain feature vectors extracted from each segmented data set are constructed into a frequency domain matrix, wherein each segmented data set includes a data segment extracted from the plurality of gas concentration data queues, the plurality of data segments in the segmented data set correspond to the same segmented time period, and the frequency domain feature vector is extracted according to the data segment; Perform convolution and pooling operations on each frequency domain matrix to obtain a feature matrix; Matrix inputs are sequentially extracted from a plurality of feature matrices and fed into a gas concentration prediction model. When the fluctuation of the output of the model is less than a threshold, the concentrations of the plurality of gases are determined according to the output of the model.

2. The method for determining the concentration of decomposition products of insulating gas according to claim 1, characterized in that: The method of segmenting and grouping the plurality of gas concentration data queues into a plurality of segmented data sets, and constructing a plurality of frequency domain feature vectors extracted from each segmented data set into a frequency domain matrix, comprises: Get the basic cycle; Determine the length of the segmented period according to the basic cycle duration, wherein the length of the segmented period is an integer multiple of the basic cycle duration; For each split period, perform the following steps: extracting data segments from the plurality of gas concentration data queues according to segmentation time periods to construct segmented data sets; According to a preset gas concentration sequence, extracting data segments from the segmented data set as data segments to be processed; A plurality of frequency domain features of the data segment to be processed are extracted according to a first formula, and the extracted plurality of frequency domain features are constructed into a frequency domain feature vector, wherein the first formula is: Wherein, GF(k) is the kth frequency domain feature, qnMax is the total number of data in the data segment to be processed, QCD(qn) is the qnth data in the data segment to be processed, e is a natural constant, j is an imaginary unit, π is the pi, ω0 is the frequency corresponding to the basic cycle, and tN is the number of gas concentration data obtained within the duration of the basic cycle; Adding the frequency domain eigenvector to the frequency domain matrix; If the traversal of the segmented data set is not completed, the process jumps to the step of extracting data segments from the segmented data set as data segments to be processed according to the preset gas sequence.

3. The method for determining the concentration of decomposition products of insulating gas according to claim 1, characterized in that: The convolution and pooling operations are performed on each frequency domain matrix to obtain a feature matrix, including: For each frequency domain matrix, perform the following steps: Get the convolution operator; Extracting a data block of the same type as the convolution operator from the frequency domain matrix according to the position index as a data block to be processed; Performing a dot product operation on the convolution operator and the data block to be processed to obtain a dot product result; Adding the dot product result to the convolution matrix according to the position index; If the traversal of the frequency domain matrix is ​​not completed, the position index is offset according to the preset step length, and the process jumps to the step of extracting a data block of the same type as the convolution operator from the frequency domain matrix according to the position index as the data block to be processed; Otherwise, perform maximum pooling on the convolution matrix to obtain a feature matrix.

4. The method for determining the concentration of decomposition products of insulating gas according to any one of claims 1 to 3, characterized in that: The step of sequentially extracting matrices from a plurality of feature matrices and inputting them into a gas concentration prediction model, and determining the concentrations of a plurality of gases according to the output of the model when the fluctuation of the output of the model is less than a threshold, includes: Extract matrices from multiple feature matrices in sequence as the matrix to be input; Inputting the matrix to be input into the gas concentration prediction model to obtain a plurality of second outputs; The output deviation rate is determined according to the second formula, the plurality of second outputs and the plurality of first outputs, wherein the plurality of first outputs are the plurality of outputs of the gas concentration prediction model in the previous time, and the second formula is: Where Bias is the output deviation rate, is the cnth second output, is the cnth first output, cnMax is the total number of output nodes of the gas concentration prediction model; Adding the output deviation rate to a deviation rate queue; If the values ​​of the last multiple bits of the deviation rate queue are all less than the deviation rate threshold, the multiple second outputs are used as the concentrations of the multiple gases; Otherwise, jump to the step of sequentially extracting matrices from a plurality of feature matrices as matrices to be input.

5. The method for determining the concentration of decomposition products of insulating gas according to any one of claims 1 to 3, characterized in that: The gas concentration prediction model is constructed according to a plurality of historical characteristic matrices, and the construction process of the gas concentration prediction model includes: Obtain a prediction basic model, multiple historical feature matrices, and multiple historical gas concentration arrays, wherein each historical feature matrix corresponds to a historical gas concentration array, and the time node corresponding to the historical gas concentration array is later than the time node corresponding to the historical feature matrix; Randomly generate a plurality of coefficient arrays, and substitute the plurality of coefficient arrays into the prediction basic model to generate a plurality of intermediate models; Taking out a matrix from the plurality of historical feature matrices according to the order of time nodes as the historical feature matrix to be input; Inputting the historical feature matrices to be input into the multiple intermediate models respectively to obtain multiple output arrays, wherein each output array corresponds to an intermediate model; Determine a plurality of model deviations according to the plurality of output arrays and a target historical gas concentration array, wherein the target historical gas concentration array is a historical gas concentration array corresponding to the historical feature matrix to be input; Adding the multiple model deviations to corresponding deviation queues respectively; If there is a queue whose last few digits are smaller than the deviation threshold among the multiple deviation queues, the intermediate model corresponding to the queue whose last few digits are smaller than the deviation threshold is used as the gas concentration prediction model; Otherwise, the coefficients of the multiple intermediate models are adjusted respectively according to the multiple deviation queues, and the process jumps to the step of taking out matrices from the multiple historical feature matrices according to the order of time nodes as the historical feature matrix to be input.

6. The method for determining the concentration of decomposition products of insulating gas according to claim 5, characterized in that: The prediction basic model is: In the formula, QC cn is the cnth output of the predicted basic model, mMax is the total number of rows of intermediate nodes, nMax is the total number of columns of intermediate nodes, and wo cn,a is the ath coefficient of the cnth output node, wh m,n,a is the ath coefficient of the node in the mth row and nth column, h a,n-1 is the output of the intermediate node in row a and column n-1, n k is the column number of the input node, who m,n,a is the ath feeding coefficient of the mth row feeding into the column node, x b is the bth input variable, wi m,1,b is the b-th coefficient of the middle node in the first column of the m-th row, bmax is the total number of input variables, e is a natural constant, and C is the bias coefficient.

7. The method for determining the concentration of decomposition products of insulating gas according to claim 5, characterized in that: The adjusting the coefficients of the plurality of intermediate models respectively according to the plurality of deviation queues comprises: For each intermediate model, perform the following steps: According to the minimum value in the deviation queue, determine the historical optimal coefficient array; Determine a global optimal coefficient array according to the minimum value of the last bits of the plurality of deviation queues; The coefficients of the intermediate model are adjusted according to the third formula, wherein the third formula is: In the formula, is the ath coefficient of the intermediate model before adjustment, w a,opta is the ath coefficient of the global optimal coefficient array, w a,ipth is the ath coefficient of the historical optimal coefficient array, is the global optimal distance, is the historical optimal distance, k1 is the first coefficient, k2 is the second coefficient, k a Substitute the coefficients for the first time, k h Substituting the coefficients for the second time, is the ath coefficient of the adjusted intermediate model, and distb is the basic disturbance.

8. An insulating gas decomposition product concentration device, characterized in that: Used to implement the insulating gas decomposition product concentration determination method according to any one of claims 1 to 7, the insulating gas decomposition product concentration device comprises: A gas concentration data acquisition module, used to acquire multiple gas concentration data queues, wherein the gas concentration data queues represent the characteristics of gas concentration changing over time; A frequency domain analysis module, used to divide the queue period into a plurality of segmented periods, and construct a plurality of frequency domain feature vectors extracted from each segmented data set into a frequency domain matrix, wherein each segmented data set includes a plurality of data segments extracted from the gas concentration data queue, the plurality of data segments in the segmented data set correspond to the same segmented period, and the frequency domain feature vectors are extracted according to the data segments; A feature extraction module is used to perform convolution and pooling operations on each frequency domain matrix to obtain a feature matrix; as well as, The gas concentration determination module is used to extract matrices from multiple feature matrices in sequence and input them into the gas concentration prediction model. When the fluctuation of the output of the model is less than a threshold, the concentrations of multiple gases are determined according to the output of the model.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program that can be run on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method as claimed in any one of claims 1 to 7 are implemented.

Images