Production suspension and restriction identification method, device, equipment and storage medium

Abstract

The present application provides a method, device, equipment and storage medium for production suspension and restriction identification. It relates to the field of industrial data processing technology. The method includes: judging the first production suspension and restriction state and the second production suspension and restriction state based on the production load coefficient and the online monitoring production coefficient, and obtaining the production status label data; using the production load coefficient, the actual power consumption, the multi-dimensional flue gas parameter coefficient and the multi-dimensional actual flue gas parameter value as the model input, and the production status label data as the output, to establish a production suspension and restriction identification model, and train the production suspension and restriction identification model, and use the trained production suspension and restriction identification model to realize production suspension and restriction identification. The present application uses the enterprise's electricity consumption data and online monitoring data, couples the enterprise's production load coefficient and the multi-dimensional flue gas parameter coefficient distribution, uses a non-parametric, dynamic threshold setting method to judge the production status, and constructs a production suspension and restriction identification model based on the attention mechanism to improve the efficiency and effect of production status identification.

Description

Technical Field

[0001] The present application relates to the technical field of industrial data processing, and in particular to a method, device, equipment and storage medium for identifying production suspension or restriction. Background Art

[0002] At present, after the heavy pollution weather emergency is launched, the pollution situation is severe, the time is tight, the duration is short, and the public attention is high, and the requirements for industrial enterprises to start suspension and restriction of production in a timely manner are high. At present, after the heavy pollution weather emergency is launched, supervision is still mainly based on "human defense", and supervision is carried out by on-site verification of enterprises.

[0003] Common existing methods include methods based on manual experience judgment, methods based on statistics, and methods based on machine learning.

[0004] Methods based on manual experience judgment are often reliable in judging the production status of enterprises, and experts can adjust the judgment criteria according to actual conditions. However, manual judgment is time-consuming and inefficient, and is not suitable for processing large-scale data sets.

[0005] Statistical methods use statistical indicators such as mean and standard deviation to perform statistical analysis based on historical data of a certain period of time. They can be repeatedly applied to similar data sets, but the generalization ability of methods that rely solely on statistics needs to be improved.

[0006] Currently, machine learning-based methods often use traditional machine learning algorithms such as logistic regression and random forest. Traditional machine learning algorithms are easy to implement and highly stable, but their expressive power is relatively limited and they are easily affected by noise and outliers. Summary of the invention

[0007] The present application provides a method, device, equipment and storage medium for identifying production suspension and restriction. By coupling the production load coefficient of the enterprise and the multi-dimensional flue gas parameter coefficient distribution and constructing a deep learning production suspension and restriction identification model, it is possible to quickly and accurately identify enterprises that have not implemented the production suspension and restriction control requirements during heavy pollution periods, realize remote supervision of enterprises included in the heavy pollution emergency supervision list, and improve the effectiveness of law enforcement and supervision.

[0008] In a first aspect, the present application provides a method for identifying production suspension or restriction, comprising:

[0009] Determine the first production suspension or restriction state based on the production load factor;

[0010] Determine the second production suspension or restriction state based on the online monitoring production coefficient;

[0011] According to the first production suspension or restriction status and the second production suspension or restriction status, obtaining production status label data;

[0012] Taking the production load coefficient, actual power consumption, multi-dimensional flue gas parameter coefficients and multi-dimensional actual flue gas parameter values ​​as model input, and taking the production status label data as output, a production suspension and restriction identification model is established, and the production suspension and restriction identification model is trained. The trained production suspension and restriction identification model is used to realize production suspension and restriction identification.

[0013] In a possible design, determining the first production suspension or restriction state based on the production load factor includes:

[0014] Get the hourly electricity consumption of the enterprise;

[0015] The production load factor is calculated by the following formula:

[0016] p j = e j / B

[0017] Among them, p j is the production load factor at time j, e j is the electricity consumption at timestamp j, and B is the production benchmark value, which is used to characterize the electricity consumption of the enterprise under normal production conditions;

[0018] The production load coefficient is fitted with a skewed distribution, and the skewed distribution is transformed into a normal distribution to obtain the transformed production load coefficient PT, PT=(pt1,pt2,....,pt t ), where pt j is the transformed production load coefficient at timestamp j, where j=1,2,...,t;

[0019] First dynamic threshold and the second dynamic threshold From and Select from the set and calculate using the following formula and :

[0020] ;

[0021] in, The first dynamic threshold Preselected set, where , The second dynamic threshold Preselected set, where , is the mean value of the transformed production load factor PT, is the standard deviation of the transformed production load factor PT, is a pre-selected set of negative standard scores under normal distribution, Preselect a set of positive standard scores for the normal distribution;

[0022] according to and Determine the first dynamic threshold and the second dynamic threshold ;

[0023] Select the maximum value of G corresponding to and As and The best value in the set, where ; The calculation formula of G is:

[0024] ;

[0025] Among them, G is the evaluation index of production status division effect, which is less than Production load factor ; greater than Production load factor ; Greater than or equal to and less than or equal to Production load factor ;

[0026] Based on the first dynamic threshold and the second dynamic threshold , determine the first suspension and restriction status by the following method:

[0027] pt j > When the first suspension or restriction of production status is normal production of the enterprise;

[0028] pt j < When the first suspension or restriction status is the suspension of production of the enterprise;

[0029] ≤pt j ≤ At this time, the first production suspension and restriction status is enterprise production restriction.

[0030] In a possible design, judging the second production suspension or restriction state based on the online monitoring production coefficient includes:

[0031] Obtain the hourly data of the enterprise's flue gas parameters for n periods of time and calculate the flue gas parameter coefficients at different timestamps ;

[0032] According to the difference in the contribution of each flue gas parameter to the production status of the enterprise, the weight of each flue gas parameter is dynamically adjusted. For n flue gas parameters, the entropy weight method is used to determine the weight W=(W1,W2,....,W i,....,W n ), where W i is the weight of the i-th smoke parameter;

[0033] The smoke parameter data of the jth timestamp for the i-th smoke parameter Normalize it and get , the formula is as follows:

[0034] ;

[0035] in, is the normalized data based on the jth timestamp of the ith smoke parameter, min is the minimum function, max is the maximum function, X i is the hourly data of the i-th flue gas parameter;

[0036] Normalized data based on the jth timestamp of the i-th smoke parameter , the ratio value is obtained by the following formula:

[0037] ;

[0038] in, is the proportional value;

[0039] The entropy E of the i-th flue gas parameter is calculated by the following formula: i :

[0040] E i = ;

[0041] The W of the i-th flue gas parameter is calculated by the following formula i :

[0042] ;

[0043] The flue gas parameters are weighted according to the obtained weight values, and the formula is as follows:

[0044] ;

[0045] Among them, v j is the online monitoring production coefficient at time stamp j, , and Indicates the first, second and nth smoke parameter coefficients at timestamp j;

[0046] Calculate the third dynamic threshold and the fourth dynamic threshold ;

[0047] Based on the third dynamic threshold and the fourth dynamic threshold , determine the second production suspension status by the following method:

[0048] v j > When the second suspension or restriction of production status is normal production of the enterprise;

[0049] v j < When the second suspension or restriction status is that the enterprise stops production;

[0050] ≤v j ≤ The second production suspension and restriction status is enterprise production restriction.

[0051] In a possible design, the production status label data is obtained according to the first production suspension or restriction status and the second production suspension or restriction status, including:

[0052] For the jth time stamp, if the production load factor pt j Production status and online monitoring of production coefficient v j If the production status is the same, the j-timestamp production status label data s is retained. j , then there are d timestamp production status label data represented by S=(s1,s2,,...., sd), otherwise the j timestamp data is not retained.

[0053] In a possible design, the production suspension and restriction identification model includes:

[0054] MLP layer, including 4 linear layers and 1 dropout layer; wherein the 4 linear layers use tanh, tanh, tanh and relu activation functions respectively;

[0055] N Multi-Head Attention layers, which use the output of the MLP layer as input to describe the importance of the input and use the important inputs to form a vector representation of the production state;

[0056] The Linear-Softmax layer is used to normalize the vector representation of the production status through the Linear layer and softmax to obtain the production status probability.

[0057] In a possible design, the output of the production suspension and restriction identification model is the highest probability value and its corresponding production status.

[0058] In a second aspect, the present application provides a device for identifying production suspension or restriction, the device comprising:

[0059] A first state determination module is configured to determine a first production suspension or restriction state based on a production load factor;

[0060] A second state judgment module is configured to judge a second production suspension or restriction state based on the online monitoring production coefficient;

[0061] A status label generating module is configured to obtain production status label data according to the first production suspension or restriction status and the second production suspension or restriction status;

[0062] The model building training module is configured to use the production load coefficient, actual power consumption, multi-dimensional flue gas parameter coefficients and multi-dimensional actual flue gas parameter values ​​as model inputs, and the production status label data as outputs, to establish a production suspension and restriction identification model, and to train the production suspension and restriction identification model, and to realize production suspension and restriction identification using the trained production suspension and restriction identification model.

[0063] In a third aspect, an embodiment of the present application provides an electronic device, comprising: at least one processor and a memory; the memory stores computer execution instructions; the at least one processor executes the computer execution instructions stored in the memory, so that the at least one processor executes the production suspension and restriction identification method described in the first aspect and various possible designs of the first aspect.

[0064] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer execution instructions are stored. When a processor executes the computer execution instructions, the production suspension and restriction identification method described in the first aspect and various possible designs of the first aspect is implemented.

[0065] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program. When the computer program is executed by a processor, it implements the production suspension and restriction identification method described in the first aspect and various possible designs of the first aspect.

[0066] The main advantages of the production suspension and restriction identification method, device, equipment and storage medium provided by this application are summarized as follows:

[0067] (1) Using the enterprise electricity consumption data and online monitoring data, the production load factor and multi-dimensional flue gas parameter coefficient are defined based on expert experience.

[0068] (2) The production load coefficient and the multi-dimensional flue gas parameter coefficient distribution are coupled, and a non-parametric and dynamic threshold setting method is used to generate production status label data. The production status label data is verified by an expert mechanism to ensure authenticity and accuracy.

[0069] (3) Using the enterprise's production load factor, actual electricity consumption, multi-dimensional flue gas parameter coefficients, and multi-dimensional actual flue gas parameter data, an Attention-based deep learning model for production suspension and restriction identification is constructed to improve the efficiency and effectiveness of production status identification, improve the accuracy of on-site inspections by law enforcement personnel, and enhance the effectiveness of law enforcement and supervision. BRIEF DESCRIPTION OF THE DRAWINGS

[0070] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0071] Figure 1 A flowchart of a method for identifying a suspension or restriction of production provided in an embodiment of the present application;

[0072] Figure 2 An architectural diagram of a production suspension or restriction identification model provided in an embodiment of the present application;

[0073] Figure 3 A schematic diagram of the structure of a production suspension or restriction identification device provided in an embodiment of the present application.

[0074] The above drawings have shown clear embodiments of the present application, which will be described in more detail later. These drawings and text descriptions are not intended to limit the scope of the present application in any way, but to illustrate the concept of the present application to those skilled in the art by referring to specific embodiments. DETAILED DESCRIPTION

[0075] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The implementations described in the following exemplary embodiments do not represent all implementations consistent with the present application. Instead, they are merely examples of devices and methods consistent with some aspects of the present application as detailed in the appended claims.

[0076] In the technical solution of this application, the collection, storage, use, processing, transmission, provision and disclosure of information such as financial data or user data involved shall comply with the provisions of relevant laws and regulations and shall not violate public order and good morals.

[0077] It should be noted that in the embodiments of the present application, certain software, components, models and other existing solutions in the industry may be mentioned, and they should be regarded as exemplary. Their purpose is only to illustrate the feasibility of implementing the technical solution of the present application, but it does not mean that the applicant has or will necessarily use the solution.

[0078] The technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems are described in detail below with specific embodiments. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0079] The embodiment of the present application provides a method for identifying production suspension and restriction, which uses the enterprise's electricity consumption data and online monitoring data, couples the enterprise's production load coefficient and the multi-dimensional flue gas parameter coefficient distribution, uses a non-parametric, dynamic threshold setting method to judge the production status, and constructs an attention-based production suspension and restriction identification model to improve the efficiency and effect of production status identification. This method can be used to remotely monitor the implementation of production suspension and restriction control requirements by enterprises during heavy pollution weather.

[0080] Figure 1 A flowchart of a method for identifying a production suspension or restriction provided in an embodiment of the present application. Figure 1 As shown, the production suspension and restriction identification method includes the following steps 1 to 4.

[0081] Step 1: Determine the production suspension or restriction of the enterprise based on the electricity consumption data. That is, determine the first production suspension or restriction status based on the production load factor.

[0082] In some embodiments, step 1 is specifically implemented by the following steps 1.1 to 1.3.

[0083] Step 1.1. Calculate the production load factor.

[0084] Get the hourly electricity consumption E, E of the enterprise j =(e1,e2,...., e j ,....,e t ), t is the time window size, where e j is the power consumption at timestamp j.

[0085] The production load factor represents the relationship between the hourly electricity consumption of an enterprise and the production benchmark value. The production load factor P with a time stamp of t is P=(p1,p2,....,p t ), where p j is the production load factor at timestamp j. The calculation formula of the production load factor is as follows.

[0086] p j = e j / B

[0087] For the calculation of the production benchmark value B, the hourly electricity consumption data of the enterprise is used for dbscan clustering, and adaptive clustering is performed according to the data density. Then the mean of each category is calculated, and the highest mean is selected as the production benchmark value B of the enterprise, which can characterize the electricity consumption of the enterprise under normal production conditions.

[0088] Step 1.2: Calculate the first dynamic threshold and the second dynamic threshold .

[0089] Determine the first dynamic threshold and the second dynamic threshold The following steps can be used to divide the enterprise into production, reduced production and normal production states:

[0090] Step 1.21, data distribution fitting. Considering that the data distribution of the production load coefficient P shows a skewed trend, the skewed distribution fitting is performed on the production load coefficient P, and the skewed distribution is transformed into a normal distribution to obtain the transformed production load coefficient PT, PT=(pt1,pt2,....,pt t ), where pt j is the production load factor at time stamp j. For right-skewed distribution, use logarithmic transformation to transform it into normal distribution, and for left-skewed distribution, use inverse transformation to transform it into right-skewed distribution, and then transform it into normal distribution.

[0091] Step 1.22: First dynamic threshold and the second dynamic threshold Calculation.

[0092] First dynamic threshold and the second dynamic threshold From and Select and The calculation formula is as follows.

[0093] ;

[0094] First dynamic threshold and the second dynamic threshold in accordance with and OK. The value range is -2 to -10, The value range of is 2 to 10. Select the maximum value of G corresponding to and ,in The calculation formula of G is as follows.

[0095] ;

[0096] Among them, G is the evaluation index of production status division effect, which is less than Production load factor ; greater than Production load factor ; Greater than or equal to and less than or equal to Production load factor .

[0097] Step 1.3: Determine production status.

[0098] pt j > When the first suspension or restriction of production status is normal production of the enterprise;

[0099] pt j < When the first suspension or restriction status is the suspension of production of the enterprise;

[0100] ≤pt j ≤ At this time, the first production suspension and restriction status is enterprise production restriction.

[0101] Step 2: Determine the production suspension or restriction of the enterprise based on the online monitoring data. That is, determine the second production suspension or restriction status based on the online monitoring production coefficient.

[0102] In some embodiments, step 2 can be implemented by the following steps 2.1 to 2.5.

[0103] Step 2.1, calculate the flue gas parameter coefficient.

[0104] Obtain the hourly data of n flue gas parameters of the enterprise, such as five flue gas parameters including oxygen content, flue gas temperature, flue gas flow, flow velocity, and humidity. i , , t is the time window size, and the smoke parameter coefficients are calculated in the same way as in step 1.1 above , , (i.e. calculated based on the production benchmark value, where the production benchmark value is calculated based on the flue gas parameter hourly data) then the flue gas parameter coefficient , the smoke parameter data of the jth timestamp of the i-th smoke parameter .

[0105] Step 2.2: Dynamic weight matching of flue gas parameter coefficients.

[0106] According to the difference in the contribution of each flue gas parameter to the production status of the enterprise, the weight of each flue gas parameter is dynamically adjusted. For the five flue gas parameters, the entropy weight method is used to determine the weight W=(W1,W2,...., W5) according to the variability of the flue gas parameters, where W i is the weight of the i-th smoke parameter.

[0107] Specifically, step 2.2 can be implemented by the following steps:

[0108] Step 2.21: Data preprocessing.

[0109] The smoke parameter data of the jth timestamp for the i-th smoke parameter Normalize it and get The formula is shown below.

[0110] ;

[0111] For the jth timestamp of the i-th smoke parameter , and the ratio is , the formula is as follows.

[0112] ;

[0113] Step 2.22: Calculate the entropy E of the i-th flue gas parameter i , the formula is as follows.

[0114] E i = ；

[0115] Step 2.23: Calculate W of the i-th flue gas parameter i , the formula is as follows.

[0116] ；

[0117] Step 2.3: Calculate the online monitoring production coefficient.

[0118] The flue gas parameter PS is weighted according to the obtained weight value W, and the formula is as follows, so as to obtain the online monitoring production coefficient V j , V j =(v1,v2,....,v t ), where v j is the production coefficient at timestamp j.

[0119] ;

[0120] Step 2.4: Calculate the third dynamic threshold and the fourth dynamic threshold and .

[0121] The third dynamic threshold and the fourth dynamic threshold and The calculation principle is the same as that of the first dynamic threshold and the second dynamic threshold. As described in step 1.2 above, the production load coefficient P is replaced by the online monitoring production coefficient.

[0122] Step 2.5: Determine production status.

[0123] v j > When the second suspension or restriction of production status is normal production of the enterprise;

[0124] v j< When the second suspension or restriction status is that the enterprise stops production;

[0125] ≤v j ≤ The second production suspension and restriction status is enterprise production restriction.

[0126] Step 3: Obtain production status label data.

[0127] For the jth time stamp, if the production load factor pt j Production status and online monitoring of production coefficient v j If the production status is the same, the j-timestamp production status label data s is retained. j , then there are d timestamp production status tag data S=(s1,s2,,...., s d ), otherwise the timestamp data will not be retained. The production status label data is verified by experts to be consistent with the actual production status.

[0128] Step 4: Establish a production suspension and restriction identification model.

[0129] In some embodiments, step 4 is specifically implemented by the following steps 4.1 to 4.3.

[0130] Step 4.1, take the production load factor, actual power consumption, oxygen content, smoke temperature, smoke flow, flow velocity, humidity parameter coefficient, oxygen content, smoke temperature, smoke flow, flow velocity, humidity parameter value as model input, take the production status label data S as output, and establish the production suspension and restriction identification model. Figure 2 As shown in FIG. 1 , the production suspension and restriction identification model specifically includes:

[0131] MLP layer: MLP consists of 4 linear layers and 1 dropout. The linear layers use tanh, tanh, tanh and relu activation functions respectively.

[0132] Multi-Head Attention layer: Combined with the process of generating production status labels, considering the different importance of inputs to the production status, N Multi-Head Attention layers are used to describe the importance of the inputs, and important inputs are used to form a vector representation of the production status.

[0133] Linear & Softmax layer: The output is normalized through the Linear layer and softmax to obtain the production status probability.

[0134] Step 4.2: Divide the data set into training set, validation set, and test set. The accuracy rate is used as the classification effect evaluation indicator for model effect evaluation. Different parameter initialization methods are used to retain the model with the best validation set effect as the production suspension and restriction recognition model.

[0135] Step 4.3: After using the model for prediction, output the highest probability value and its corresponding production status.

[0136] The present application also provides a device for identifying production suspension or restriction. Figure 3 As shown, the production suspension and restriction identification device includes:

[0137] The first state determination module 301 is configured to determine a first production suspension or restriction state based on a production load factor;

[0138] The second state judgment module 302 is configured to judge the second production suspension or restriction state based on the online monitoring production coefficient;

[0139] The status label generating module 303 is configured to obtain production status label data according to the first production suspension or restriction status and the second production suspension or restriction status;

[0140] The model building training module 304 is configured to use the production load coefficient, actual power consumption, multi-dimensional flue gas parameter coefficients and multi-dimensional actual flue gas parameter values ​​as model inputs, and the production status label data as outputs, to establish a production suspension and restriction identification model, and to train the production suspension and restriction identification model, and to implement production suspension and restriction identification using the trained production suspension and restriction identification model.

[0141] In some embodiments, the first state determination module is further configured to:

[0142] Get the hourly electricity consumption of the enterprise;

[0143] The production load factor is calculated by the following formula:

[0144] p j = e j / B

[0145] Among them, p j is the production load factor at time j, e j is the electricity consumption at timestamp j, and B is the production benchmark value, which is used to characterize the electricity consumption of the enterprise under normal production conditions;

[0146] The production load coefficient is fitted with a skewed distribution, and the skewed distribution is transformed into a normal distribution to obtain the transformed production load coefficient PT, PT=(pt1,pt2,....,pt t ), where pt jis the transformed production load coefficient at timestamp j, where j=1,2,...,t;

[0147] Calculated by the following formula and :

[0148] ;

[0149] in, The first dynamic threshold Preselected set, where , The second dynamic threshold Preselected set, where , is the mean value of the transformed production load factor PT, is the standard deviation of the transformed production load factor PT, is a pre-selected set of negative standard scores under normal distribution, Preselect a set of positive standard scores for the normal distribution;

[0150] according to and Determine the first dynamic threshold and the second dynamic threshold ;

[0151] Select the maximum value of G corresponding to and As and The best value in the set, where ; The calculation formula of G is:

[0152] ;

[0153] Among them, G is the evaluation index of production status division effect, which is less than Production load factor ; greater than Production load factor ; Greater than or equal to and less than or equal to Production load factor .

[0154] Based on the first dynamic threshold and the second dynamic threshold , determine the first suspension and restriction status by the following method:

[0155] pt j > When the first suspension or restriction of production status is normal production of the enterprise;

[0156] pt j < When the first suspension or restriction status is the suspension of production of the enterprise;

[0157] ≤pt j ≤ At this time, the first production suspension and restriction status is enterprise production restriction.

[0158] In some embodiments, the second state determination module is further configured to:

[0159] Obtain the hourly data of the enterprise's flue gas parameters for n periods of time and calculate the flue gas parameter coefficients at different timestamps ;

[0160] According to the difference in the contribution of each flue gas parameter to the production status of the enterprise, the weight of each flue gas parameter is dynamically adjusted. For n flue gas parameters, the entropy weight method is used to determine the weight W=(W1,W2,....,W i ,....,W n ), where W i is the weight of the i-th smoke parameter;

[0161] The smoke parameter data of the jth timestamp for the i-th smoke parameter Normalize it and get , the formula is as follows:

[0162] ;

[0163] in, is the normalized data based on the jth timestamp of the ith smoke parameter, min is the minimum function, max is the maximum function, X i is the hourly data of the i-th flue gas parameter;

[0164] Normalized data based on the jth timestamp of the i-th smoke parameter , the ratio value is obtained by the following formula:

[0165] ;

[0166] in, is the proportional value;

[0167] The entropy E of the i-th flue gas parameter is calculated by the following formula: i :

[0168] E i = ;

[0169] The W of the i-th flue gas parameter is calculated by the following formulai :

[0170] ;

[0171] The flue gas parameters are weighted according to the obtained weight values, and the formula is as follows:

[0172] ;

[0173] Among them, v j is the online monitoring production coefficient at time stamp j, , and Indicates the first, second and nth smoke parameter coefficients at timestamp j;

[0174] Calculate the third dynamic threshold and the fourth dynamic threshold ;

[0175] Based on the third dynamic threshold and the fourth dynamic threshold , determine the second production suspension status by the following method:

[0176] v j > When the second suspension or restriction of production status is normal production of the enterprise;

[0177] v j < When the second suspension or restriction status is that the enterprise stops production;

[0178] ≤v j ≤ The second production suspension and restriction status is enterprise production restriction.

[0179] In some embodiments, the status label generation module is further configured to:

[0180] For the jth time stamp, if the production load factor pt j Production status and online monitoring of production coefficient v j If the production status is the same, the j-timestamp production status label data s is retained. j , then there are d timestamp production status label data represented by S=(s1,s2,,...., sd), otherwise the j timestamp data is not retained.

[0181] In some embodiments, the production suspension and restriction identification model includes:

[0182] MLP layer, including 4 linear layers and 1 dropout layer; wherein the 4 linear layers use tanh, tanh, tanh and relu activation functions respectively;

[0183] N Multi-Head Attention layers, which use the output of the MLP layer as input to describe the importance of the input and use the important inputs to form a vector representation of the production state;

[0184] The Linear-Softmax layer is used to normalize the vector representation of the production status through the Linear layer and softmax to obtain the production status probability.

[0185] In some embodiments, the output of the production suspension and restriction identification model is the highest probability value and its corresponding production status.

[0186] An embodiment of the present application provides an electronic device, which may include: a processor and a memory, wherein the processor and the memory may communicate with each other; illustratively, the processor and the memory communicate with each other via a communication bus.

[0187] The processor executes the computer execution instructions stored in the memory, so that the processor executes the scheme in the above embodiment. The processor can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application-specific integrated circuit ASIC, a field programmable gate array FPGA or other programmable logic devices, discrete gates or transistor logic devices, and discrete hardware components.

[0188] The communication bus can be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The system bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, only one thick line is used in the figure, but it does not mean that there is only one bus or one type of bus. The transceiver is used to realize the communication between the database access device and other computers (such as clients, read-write libraries, and read-only libraries). The memory may include random access memory (RAM) and may also include non-volatile memory.

[0189] The electronic device provided in the embodiment of the present application may be the terminal device of the above embodiment.

[0190] An embodiment of the present application also provides a computer-readable storage medium, in which computer instructions are stored. When the computer instructions are executed on a computer, the computer executes the technical solution of the production suspension and restriction identification method of the above embodiment.

[0191] An embodiment of the present application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium. When at least one processor executes the computer program, it can implement the technical solution of the production suspension and restriction identification method in the above embodiment.

[0192] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules is only a logical function division, and there may be other division methods in actual implementation, such as multiple modules 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 an indirect coupling or communication connection through some interfaces, devices or modules, which can be electrical, mechanical or other forms.

[0193] The modules described as separate components may or may not be physically separated, and the components shown as modules 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 modules may be selected according to actual needs to implement the solution of this embodiment.

[0194] In addition, each functional module in each embodiment of the present application can be integrated into one processing unit, or each module can exist physically separately, or two or more modules can be integrated into one unit. The above-mentioned module-composed unit can be implemented in the form of hardware or in the form of hardware plus software functional units.

[0195] The above-mentioned integrated module implemented in the form of a software function module can be stored in a computer-readable storage medium. The above-mentioned software function module is stored in a storage medium, including a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) or a processor to perform some steps of the methods of various embodiments of the present application.

[0196] It should be understood that the above processor can be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in the invention can be directly embodied as being executed by a hardware processor, or can be executed by a combination of hardware and software modules in the processor.

[0197] The memory may include a high-speed RAM memory, and may also include a non-volatile storage NVM, such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk or an optical disk, etc.

[0198] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For ease of representation, the bus in the drawings of this application is not limited to only one bus or one type of bus.

[0199] The above storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic memory, flash memory, magnetic disk or optical disk. The storage medium can be any available medium that can be accessed by a general or special purpose computer.

[0200] An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be a component of the processor. The processor and the storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the storage medium can also exist as discrete components in an electronic control unit or a main control device.

[0201] Those skilled in the art can understand that all or part of the steps of implementing the above-mentioned method embodiments can be completed by hardware related to program instructions. The aforementioned program can be stored in a computer-readable storage medium. When the program is executed, the steps of the above-mentioned method embodiments are executed; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk and other media that can store program codes.

[0202] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for identifying production suspension or restriction, characterized in that: The method comprises: Determine the first production suspension or restriction state based on the production load factor; Determine the second production suspension or restriction state based on the online monitoring production coefficient; According to the first production suspension or restriction status and the second production suspension or restriction status, obtaining production status label data; Taking the production load factor, actual power consumption, multi-dimensional flue gas parameter coefficients and multi-dimensional actual flue gas parameter values ​​as model inputs and the production status label data as outputs, an Attention-based production suspension and restriction recognition model is established, and the production suspension and restriction recognition model is trained, and the production suspension and restriction recognition model after training is used to realize production suspension and restriction recognition; The first production suspension or restriction state is determined based on the production load factor, including: Get the hourly electricity consumption of the enterprise; Calculate production load factor; The production load coefficient is fitted with a skewed distribution, and the skewed distribution is transformed into a normal distribution to obtain the transformed production load coefficient PT, PT=(pt1,pt2,....,pt t ), where pt j is the transformed production load coefficient at timestamp j, where j=1,2,...,t; First dynamic threshold and the second dynamic threshold From and Select from the set and calculate using the following formula and : ； in, The first dynamic threshold Preselected set, where , The second dynamic threshold Preselected set, where , is the mean value of the transformed production load factor PT, is the standard deviation of the transformed production load factor PT, is a pre-selected set of negative standard scores under normal distribution, Preselect a set of positive standard scores for the normal distribution; according to and Determine the first dynamic threshold and the second dynamic threshold ; Select the maximum value of G corresponding to and As and The best value in the set, where ; The calculation formula of G is: ； Among them, G is the evaluation index of production status division effect, which is less than Production load factor ; greater than Production load factor ; Greater than or equal to and less than or equal to Production load factor ; Based on the first dynamic threshold and the second dynamic threshold , determine the first production suspension or restriction status.

2. The method for identifying production suspension or restriction according to claim 1, characterized in that: The production load factor is calculated by the following formula: p j = e j / B Among them, p j is the production load factor at time j, e j is the electricity consumption at timestamp j, and B is the production benchmark value, which is used to characterize the electricity consumption of the enterprise under normal production conditions; Based on the first dynamic threshold and the second dynamic threshold , determine the first suspension and restriction status by the following method: pt j > When the first suspension or restriction of production status is normal production of the enterprise; pt j < When the first suspension or restriction status is the suspension of production of the enterprise; ≤pt j ≤ At this time, the first production suspension and restriction status is enterprise production restriction.

3. The method for identifying production suspension or restriction according to claim 1, characterized in that: The second production suspension or restriction state is determined based on the online monitoring production coefficient, including: Obtain the hourly data of the enterprise's flue gas parameters for n periods of time and calculate the flue gas parameter coefficients at different timestamps ; According to the difference in the contribution of each flue gas parameter to the production status of the enterprise, the weight of each flue gas parameter is dynamically adjusted. For n flue gas parameters, the entropy weight method is used to determine the weight W=(W1, W2,....,W i ,....,W n ), where W i is the weight of the i-th smoke parameter; The smoke parameter data of the jth timestamp for the i-th smoke parameter Normalize it and get , the formula is as follows: ； in, is the normalized data based on the jth timestamp of the ith smoke parameter, min is the minimum function, max is the maximum function, X i is the hourly data of the i-th flue gas parameter; Normalized data based on the jth timestamp of the i-th smoke parameter , the ratio value is obtained by the following formula: ； in, is the proportional value; The entropy E of the i-th flue gas parameter is calculated by the following formula: i : E i = ； The W of the i-th flue gas parameter is calculated by the following formula i : ； The flue gas parameters are weighted according to the obtained weight values, and the formula is as follows: ； Among them, v j is the online monitoring production coefficient at time stamp j, , and Indicates the 1st, 2nd and nth smoke parameter coefficients at timestamp j; Calculate the third dynamic threshold and the fourth dynamic threshold ; Based on the third dynamic threshold and the fourth dynamic threshold , determine the second production suspension status by the following method: v j > When the second suspension or restriction of production status is normal production of the enterprise; v j < When the second suspension or restriction status is that the enterprise stops production; ≤v j ≤ The second production suspension and restriction status is enterprise production restriction.

4. The method for identifying production suspension or restriction according to claim 1, characterized in that: According to the first production suspension or restriction status and the second production suspension or restriction status, production status label data is obtained, including: For the jth time stamp, if the production load factor pt j Production status and online monitoring of production coefficient v j If the production status is the same, the j-timestamp production status label data s is retained. j , then there are d timestamp production status label data represented by S=(s1,s2,,...., sd), otherwise the j timestamp data is not retained.

5. The method for identifying production suspension or restriction according to claim 1, characterized in that: The production suspension and restriction identification model includes: MLP layer, including 4 linear layers and 1 dropout layer; wherein the 4 linear layers respectively use tanh, tanh, tanh and relu activation functions; N Multi-Head Attention layers, which use the output of the MLP layer as input to describe the importance of the input and use the important inputs to form a vector representation of the production state; The Linear-Softmax layer is used to normalize the vector representation of the production status through the Linear layer and softmax to obtain the production status probability.

6. The method for identifying production suspension or restriction according to claim 5, characterized in that: The output of the production suspension and restriction identification model is the highest probability value and its corresponding production status.

7. A device for identifying production suspension or restriction, characterized in that: The device comprises: A first state determination module is configured to determine a first production suspension or restriction state based on a production load factor; A second state judgment module is configured to judge a second production suspension or restriction state based on the online monitoring production coefficient; A status label generating module is configured to obtain production status label data according to the first production suspension or restriction status and the second production suspension or restriction status; The model building training module is configured to use the production load factor, actual power consumption, multi-dimensional flue gas parameter coefficients and multi-dimensional actual flue gas parameter values ​​as model inputs, and the production status label data as outputs, to establish an attention-based production suspension and restriction recognition model, and to train the production suspension and restriction recognition model, and to realize production suspension and restriction recognition using the trained production suspension and restriction recognition model; The first state determination module is further configured to: Get the hourly electricity consumption of the enterprise; Calculate production load factor; The production load coefficient is fitted with a skewed distribution, and the skewed distribution is transformed into a normal distribution to obtain the transformed production load coefficient PT, PT=(pt1,pt2,....,pt t ), where pt j is the transformed production load coefficient at timestamp j, where j=1,2,...,t; First dynamic threshold and the second dynamic threshold Respectively from and Select from the set and calculate using the following formula and : ； in, The first dynamic threshold Preselected set, where , The second dynamic threshold Preselected set, where , is the mean value of the transformed production load factor PT, is the standard deviation of the transformed production load factor PT, is a pre-selected set of negative standard scores under normal distribution, Preselect a set of positive standard scores for the normal distribution; according to and Determine the first dynamic threshold and the second dynamic threshold ; Select the maximum value of G corresponding to and As and The best value in the set, where ; The calculation formula of G is: ； Among them, G is the evaluation index of production status division effect, which is less than Production load factor ; greater than Production load factor ; Greater than or equal to and less than or equal to Production load factor ; Based on the first dynamic threshold and the second dynamic threshold , determine the first production suspension or restriction status.

8. An electronic device, characterized in that: include: A processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the production suspension and restriction identification method according to any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method for identifying production suspension or restriction as described in any one of claims 1 to 6.

10. A computer program product, comprising a computer program, wherein when the computer program is executed by a processor, the method for identifying production suspension or restriction as described in any one of claims 1 to 6 is implemented.

Images