Enterprise Emission Verification, Model Building Method, Device, Equipment and Storage Medium

Abstract

The present invention relates to the technical field of pollutant emission verification, and particularly to a method, device, equipment and storage medium for verifying enterprise emissions and constructing a model. The method of the present invention first obtains an enterprise emission data queue and a material consumption data set; then extracts frequency domain features from the enterprise emission data queue, and constructs a first verification vector with the obtained multiple frequency domain features and the material consumption data set; then inputs the first verification vector into multiple verification models respectively, and determines multiple first results according to the outputs of the multiple verification models; finally, selects a target result from the multiple first results as the first verified emission amount, and verifies the enterprise emission monitoring amount according to the first verified emission amount. The present invention predicts the emission amount based on the material consumption data and the emission data sequence that affect the enterprise emission. Therefore, the verification result is more reliable and credible, and the verification efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the verification of pollutant emissions, and in particular to a method, device, equipment and storage medium for verifying the emissions of enterprises and constructing a model. Background Art

[0002] Pollutant monitoring refers to the use of modern scientific and technological methods such as physics, chemistry, and biology to intermittently or continuously monitor and measure environmental chemical pollutants, physical and biological pollution factors, etc. on-site, and make a correct environmental quality assessment.

[0003] Pollutant monitoring data is affected by various factors during the collection process, resulting in data deviation. Before reporting periodic pollutant monitoring data, verification is usually required, and on-site verification at the enterprise may also be necessary when necessary. The traditional method is manual verification, which has a large workload and is prone to mistakes. For the situation of enterprises secretly discharging pollutants, it is difficult to detect even through enterprise visits. Moreover, due to the periodicity of data aggregation, when data anomalies are detected, a relatively long time has often passed, increasing the difficulty of investigation.

[0004] Based on this, it is necessary to develop and design a method for verifying the emissions of enterprises. Summary of the Invention

[0005] Embodiments of the present invention provide a method, device, equipment and storage medium for verifying the emissions of enterprises and constructing a model, which are used to solve the problem that the results of manual verification of enterprise emissions in the prior art are inaccurate.

[0006] In a first aspect, an embodiment of the present invention provides a method for verifying the emissions of enterprises, including:

[0007] Obtain an enterprise emission data queue and a material consumption data set, where the material consumption data set includes: energy consumption data and / or material consumption data;

[0008] Extract frequency domain features from the enterprise emission data queue, and construct a first verification vector by combining the obtained multiple frequency domain features with the material consumption data set;

[0009] Input the first verification vector into multiple verification models respectively, and determine multiple first results according to the outputs of the multiple verification models, where each first result corresponds to a verification model;

[0010] Select a target result from the multiple first results as the first verified emissions, and verify the enterprise emission monitoring quantity according to the first verified emissions, where the target result is the first result located in the emissions verification interval, and each verification model corresponds to an emissions verification interval.

[0011] In a possible implementation manner, extracting frequency domain features from the enterprise emission data queue and constructing a first verification vector with the obtained multiple frequency domain features and the material consumption data set includes:

[0012] Extracting multiple frequency domain features of the enterprise emission data queue according to a first formula, where the first formula is:

[0013]

[0014] In the formula, is the th frequency domain feature, is the th data of the enterprise emission data queue, is the total number of data in the enterprise emission data queue, is the natural constant, is the pi, is the fundamental frequency, is the imaginary unit, is the total number of enterprise emission data obtained within the fundamental wave period duration;

[0015] Arranging the multiple frequency domain features and the data in the material consumption data set according to a predetermined frequency feature order and a predetermined material order to obtain a first verification vector.

[0016] In a possible implementation manner, determining multiple first results according to the outputs of the multiple verification models includes:

[0017] Inputting the outputs of the multiple verification models into a second formula respectively to obtain multiple first results, where the second formula is:

[0018]

[0019] In the formula, is the first result, is the output of the verification model, is the upper limit of the emission verification interval of the verification model, is the lower limit of the emission verification interval of the verification model, is the scaling coefficient, is the first bias coefficient.

[0020] In a second aspect, an embodiment of the present invention provides a method for constructing an enterprise emission verification model, which is used to construct the enterprise emission verification method as described in the first aspect or any possible implementation manner of the first aspect. The method for constructing the enterprise emission verification model includes:

[0021] Obtain a verification basic model, a plurality of second verification vectors, and a plurality of second verification emissions, wherein each second verification vector corresponds to a second verification emission, and the second verification vector is constructed based on a historical material consumption data set and a plurality of frequency domain features extracted based on a historical enterprise emission data queue, and the historical material consumption data set includes historical energy consumption data and / or historical material consumption data;

[0022] dividing the plurality of second verification vectors into a plurality of first vector groups according to the intervals where the verified emission amounts are located, clustering the plurality of second verification vectors in each first vector group, reducing the plurality of second verification vectors in the first vector group according to the clustering result, and using the remaining second verification vectors as third verification vectors;

[0023] According to the corresponding second verified emissions, multiple third verification vectors are divided into multiple second vector groups, and multiple coefficients of the verification basic model are determined according to each second vector group and the multiple second verified emissions corresponding to the second vector group to obtain the enterprise emission verification model.

[0024] In a possible implementation, the multiple second verification vectors are divided into multiple first vector groups according to the interval where the verified emission amount is located, the multiple second verification vectors in each first vector group are clustered, the multiple second verification vectors in the first vector group are reduced according to the clustering result, and the remaining second verification vectors are used as third verification vectors, including:

[0025] Arranging the plurality of second verified emissions according to the magnitude of the values;

[0026] According to the intervals corresponding to the arrangement, the plurality of second verified emissions are divided into equal interval spans to obtain a plurality of emission groups, and according to the correspondence between the second verified emissions and the second verification vectors and the plurality of emission groups, the plurality of second verification vectors are divided into a plurality of first vector groups, wherein each first vector group corresponds to an emission group;

[0027] For each first vector group, perform the following steps respectively:

[0028] Get the number of clusters;

[0029] According to the number of clusters, randomly selecting second verification vectors of the number of clusters from the first vector group as multiple temporary cluster centers;

[0030] For each second verification vector in the first vector group, add the cluster where the temporary cluster center with the closest Euclidean distance is located;

[0031] Obtaining class centers of multiple classes, wherein the class center is a second verification vector that is closest to a class mean vector, and the class mean vector is a mean vector of multiple vectors in the class;

[0032] If there is a situation where some of the multiple temporary clustering centers are not class centers, then use the class centers of the multiple classes as the multiple temporary clustering centers, and jump to the step of adding each second verification vector in the first vector group to the class where the temporary clustering center with the closest Euclidean distance is located;

[0033] Otherwise, reduce the second verification vectors of the small - proportion classes, and use the second verification vectors in the remaining classes as the third verification vectors, where the small - proportion classes are the classes in which the ratio of the number of second verification vectors to the number of second verification vectors in the first vector group is less than the ratio threshold.

[0034] In a possible implementation manner, the verification basic model is:

[0035]

[0036] In the formula, is the output of the verification model, is the natural constant, is the joint variable, is the th first coefficient, is the th element of the verification vector, is the total number of verification vector elements, is the second bias coefficient, is the total number of times.

[0037] In a possible implementation manner, the method of dividing the multiple third verification vectors into multiple second vector groups according to the corresponding second verification emissions, and determining the multiple coefficients of the verification basic model according to each second vector group and the multiple second verification emissions corresponding to the second vector group to obtain the enterprise emissions verification model includes:

[0038] Arrange the multiple third verification vectors according to the values of the corresponding second verification emissions;

[0039] Successively take out a preset number of verification vectors from the arrangement of the third verification vectors to construct a second vector group;

[0040] For each second vector group, perform the following steps respectively:

[0041] Configure the number of coefficients in the verification basic model according to the number of verification vectors in the second vector group;

[0042] Traversingly take out the third verification vectors from the second vector group as the vectors to be processed;

[0043] According to the third formula, determine the verification value, where the third formula is:

[0044]

[0045] In the formula, is the verification value, is the second verification emission amount corresponding to the vector to be processed, is the scaling coefficient, is the upper limit of the verification emission amount interval corresponding to the second vector group, is the lower limit of the verification emission amount interval corresponding to the second vector group, is the first bias coefficient;

[0046] Input the vector to be processed into the basic verification model, determine the model deviation according to the output of the basic verification model and the verification value, and add the model deviation to the deviation queue;

[0047] If the values of the last several digits in the deviation queue are all less than the deviation threshold, then use the basic verification model as the enterprise emission verification model corresponding to the second vector group;

[0048] Otherwise, adjust multiple coefficients of the basic verification model according to the model deviation.

[0049] In a third aspect, an embodiment of the present invention provides an enterprise emission verification device for implementing the enterprise emission verification method described in the first aspect or any possible implementation manner of the first aspect above. The enterprise emission verification device includes:

[0050] A data acquisition module for acquiring an enterprise emission data queue and a material consumption data set, where the material consumption data set includes: energy consumption data and / or material consumption data;

[0051] A frequency domain analysis module for extracting frequency domain features from the enterprise emission data queue and constructing the obtained multiple frequency domain features and the material consumption data set into a first verification vector;

[0052] A verification result acquisition module for inputting the first verification vector into multiple verification models respectively, and determining multiple first results according to the outputs of the multiple verification models, where each first result corresponds to a verification model;

[0053] And,

[0054] A verification module for selecting a target result from the multiple first results as the first verification emission amount and verifying the enterprise emission monitoring amount according to the first verification emission amount, where the target result is the first result located in the emission verification interval, and each verification model corresponds to an emission verification interval.

[0055] Fourthly, an embodiment of the present invention provides an electronic device, including a memory and a processor. A computer program that can run on the processor is stored in the memory. When the processor executes the computer program, the steps of the method described in the first aspect or any possible implementation manner of the first aspect above are implemented.

[0056] Fifthly, an embodiment of the present invention provides a computer-readable storage medium. The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the steps of the method described in the first aspect or any possible implementation manner of the first aspect above are implemented.

[0057] The beneficial effects of the embodiments of the present invention compared with the prior art are as follows:

[0058] An embodiment of the present invention discloses a method for verifying enterprise emissions. First, an enterprise emission data queue and a material consumption data set are obtained. The material consumption data set includes energy consumption data and / or material consumption data. Then, frequency domain features are extracted from the enterprise emission data queue, and a plurality of obtained frequency domain features and the material consumption data set are constructed into a first verification vector. Next, the first verification vector is respectively input into a plurality of verification models, and a plurality of first results are determined according to the outputs of the plurality of verification models. Each first result corresponds to a verification model. Finally, a target result is selected from the plurality of first results as the first verified emissions, and the enterprise emission monitoring quantity is verified according to the first verified emissions. The target result is the first result located in the emissions verification interval, and each verification model corresponds to an emissions verification interval. The embodiment of the present invention predicts emissions based on the material consumption data and emission data sequence that affect enterprise emissions, and verifies enterprise emissions according to the prediction results. Since both factor and sequence data are integrated, the verification result is more reliable and credible, and the verification efficiency is improved. Description of the Drawings

[0059] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following will briefly introduce the drawings required for the embodiments or the description of the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0060] Figure 1 It is a flowchart of the enterprise emissions verification method provided by the embodiment of the present invention;

[0061] Figure 2 It is a flowchart of the enterprise emissions verification model construction method provided by the embodiment of the present invention;

[0062] Figure 3 It is the schematic diagram of the reduction process of the verification vector provided by the embodiment of the present invention;

[0063] Figure 4 It is the schematic diagram of the process of constructing the enterprise emission verification model through the second vector group provided by the embodiment of the present invention;

[0064] Figure 5 It is the functional block diagram of the enterprise emission verification device provided by the embodiment of the present invention;

[0065] Figure 6 It is the functional block diagram of the electronic device provided by the embodiment of the present invention. Specific Embodiments

[0066] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present invention. However, those skilled in the art should clearly understand that the present invention can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, and methods are omitted to avoid unnecessary details from interfering with the description of the present invention.

[0067] To make the purpose, technical solutions, and advantages of the present invention clearer, the following will be described through specific embodiments with reference to the accompanying drawings.

[0068] The following details the embodiments of the present invention. This example is implemented on the premise of the technical solution of the present invention, and detailed implementation manners and specific operation processes are given. However, the protection scope of the present invention is not limited to the following embodiments.

[0069] Figure 1 It is the flowchart of the enterprise emission verification method provided by the embodiment of the present invention.

[0070] As Figure 1 shown, it shows the implementation flowchart of the enterprise emission verification method provided by the first aspect of the embodiment of the present invention, which is described in detail as follows:

[0071] In step 101, an enterprise emission data queue and a material consumption data set are obtained, where the material consumption data set includes: energy consumption data and / or material consumption data.

[0072] In step 102, frequency domain feature extraction is performed on the enterprise emission data queue, and a plurality of obtained frequency domain features and the material consumption data set are constructed into a first verification vector.

[0073] In some embodiments, extracting frequency domain features from the enterprise emission data queue and constructing a first verification vector by using the obtained multiple frequency domain features and the material consumption data set includes:

[0074] Extracting multiple frequency domain features of the enterprise emission data queue according to a first formula, where the first formula is:

[0075]

[0076] In the formula, is the th frequency domain feature, is the th data of the enterprise emission data queue, is the total number of data in the enterprise emission data queue, is the natural constant, is the pi, is the fundamental wave frequency, is the imaginary unit, is the total number of enterprise emission data obtained within the fundamental wave period duration;

[0077] Arranging the multiple frequency domain features and the data in the material consumption data set according to a predetermined frequency feature order and a predetermined material order to obtain a first verification vector.

[0078] Exemplarily, embodiments of the present invention determine whether the enterprise emission monitoring data is credible by obtaining the previous enterprise emission data and the contemporaneous material consumption data and analyzing the emission data sequence and the contemporaneous material consumption. For example, for certain pollutant indicators, when obtaining the monitoring value, the pollutant emission data at multiple time nodes in a previous period is also obtained and arranged in the order of time nodes to obtain an enterprise emission data queue. Then, based on the time node of the pollutant monitoring value, the material consumption amount relative to the previous time node is obtained, such as the electricity consumption amount, the water consumption amount, the raw material consumption amount, etc.

[0079] Embodiments of the present invention aim to comprehensively verify whether the emission data is credible from the time series dimension and the dimension of factors affecting the emission data through the emission data sequence. In terms of the analysis of the time series dimension, embodiments of the present invention transform the time domain data into frequency domain data for easy analysis and reduce the dimension of the data. In one application scenario, the first formula is applied to extract the frequency domain features of the enterprise emission data queue:

[0080]

[0081] In the formula, is the th frequency domain feature, is the th data in the enterprise emission data queue, is the total number of data in the enterprise emission data queue, is the natural constant, is the pi, is the fundamental frequency, is the imaginary unit, is the total number of enterprise emission data obtained within the fundamental wave period duration.

[0082] Multiple frequency domain features obtained through the above formula, and then arranged with the material consumption data in a preset order to obtain a verification vector.

[0083] In step 103, the first verification vector is respectively input into multiple verification models, and multiple first results are determined according to the outputs of the multiple verification models, where each first result corresponds to a verification model.

[0084] In some embodiments, the determining multiple first results according to the outputs of the multiple verification models includes:

[0085] The outputs of the multiple verification models are respectively input into a second formula to obtain multiple first results, where the second formula is:

[0086]

[0087] In the formula, is the first result, is the output of the verification model, is the upper limit of the emission verification interval of the verification model, is the lower limit of the emission verification interval of the verification model, is the scaling coefficient, is the first bias coefficient.

[0088] In step 104, a target result is selected from the multiple first results as the first verified emission, and the enterprise emission monitoring quantity is verified according to the first verified emission, where the target result is the first result located in the emission verification interval, and each verification model corresponds to an emission verification interval.

[0089] Exemplarily, in the embodiment of the present invention, the verification vectors obtained in the foregoing steps are respectively input into multiple verification models, and each model outputs a result. The verification models in the embodiment of the present invention are modeled according to the segmented emissions. In other words, different pollutant emissions correspond to different verification models. The advantage of doing this is to fit the non-linear relationship with multiple relatively linear models. Here, the non-linearity refers to the non-linear relationship between emissions, time domain features, and material consumption data.

[0090] Due to the existence of multiple models, when we obtain the verification vector, we may not be clear which model to input to determine the verified emissions. In fact, for all verification models, the pollutant emission data is mapped to the interval [0 - 1], and scaling and offset are performed. For example, after scaling, it is [0.1 - 0.9]. If the verification vector is not applicable to a certain verification model, after the verification vector is input into the verification model, the data output by the verification model will exceed this interval. For example, when the verification model is input into an inapplicable model, the output value of the model is greater than 0.9 or less than 0.1.

[0091] And since the model is mapped for the interval, after the model outputs, we need to perform a secondary conversion. The embodiment of the present invention applies the second formula for conversion to restore the emission estimation value. The second formula is:

[0092]

[0093] In the formula, is the first result, is the output of the verification model, is the upper limit of the emission verification interval of the verification model, is the lower limit of the emission verification interval of the verification model, is the scaling coefficient, is the first bias coefficient.

[0094] When the emission estimation value is obtained according to the model, it can be compared with the measured value to determine whether the enterprise emission monitoring quantity is correct.

[0095] Regarding the construction of the enterprise emission verification model, the second aspect of the embodiment of the present invention is discussed in detail.

[0096] Figure 2 is the flowchart of the method for constructing the enterprise emission verification model provided by the embodiment of the present invention.

[0097] As Figure 2 shown, it shows the implementation flowchart of the method for constructing the enterprise emission verification model provided by the second aspect of the embodiment of the present invention, which is described in detail as follows:

[0098] In step 201, obtain the verification basic model, multiple second verification vectors, and multiple second verified emissions. Among them, each second verification vector corresponds to a second verified emission. The second verification vector is constructed based on the historical material consumption data set and multiple frequency domain features extracted from the historical enterprise emission data queue. The historical material consumption data set includes historical energy consumption data and / or historical material consumption data;

[0099] In some embodiments, the verification basic model is:

[0100]

[0101] Wherein, is the output of the verification model, is the natural constant, is the combined variable, is the th first coefficient, is the th element of the verification vector, is the total number of elements of the verification vector, is the second bias coefficient, is the total degree.

[0102] Exemplarily, the embodiment of the present invention completes the model construction by determining multiple parameters of the verification basic model based on multiple verification vectors and multiple emission amounts corresponding to these verification vectors.

[0103] The verification basic model is:

[0104]

[0105] Wherein, is the output of the verification model, is the natural constant, is the combined variable, is the th first coefficient, is the th element of the verification vector, is the total number of elements of the verification vector, is the second bias coefficient, is the total degree.

[0106] In terms of the verification vector (to distinguish from the verification model in the first aspect, the verification vector in the second aspect is the second verification vector), each second verification vector corresponds to a second verification emission amount, which has the same relationship and acquisition method as the verification vector and the verification emission amount in the first aspect. The second verification vector is obtained by extracting the frequency domain features from the historical enterprise emission data queue constructed based on the multiple enterprise emission data in the previous time period of the second verification emission amount, and combining the obtained multiple frequency domain features with the vector constructed from the historical material consumption data, where the historical material consumption data is in the same period as the second verification emission amount.

[0107] In step 202, the multiple second verification vectors are divided into multiple first vector groups according to the interval where the verification emission amount is located, the multiple second verification vectors in each first vector group are clustered, the multiple second verification vectors in the first vector group are reduced according to the clustering result, and the remaining second verification vectors are used as the third verification vectors.

[0108] In some embodiments, dividing the multiple second verification vectors into multiple first vector groups according to the interval where the verified emissions are located, clustering the multiple second verification vectors in each first vector group, reducing the multiple second verification vectors in the first vector group according to the clustering result, and using the remaining second verification vectors as third verification vectors includes:

[0109] Arranging the multiple second verified emissions according to the magnitude of the values;

[0110] Dividing the multiple second verified emissions at equal interval spans according to the intervals corresponding to the arrangement to obtain multiple emission groups, and dividing the multiple second verification vectors into multiple first vector groups according to the correspondence between the second verified emissions and the second verification vectors and the multiple emission groups, where each first vector group corresponds to one emission group;

[0111] For each first vector group, perform the following steps respectively:

[0112] Obtain the number of clusters;

[0113] According to the number of clusters, randomly select the second verification vectors of the number of clusters from the first vector group as multiple temporary cluster centers;

[0114] For each second verification vector in the first vector group, add it to the class where the temporary cluster center with the closest Euclidean distance is located;

[0115] Obtain the class centers of the multiple classes, where the class center is the second verification vector with the closest distance to the class mean vector, and the class mean vector is the mean vector of the multiple vectors in the class;

[0116] If there is a situation where some of the multiple temporary cluster centers are not class centers, then use the class centers of the multiple classes as the multiple temporary cluster centers, and jump to the step of for each second verification vector in the first vector group, add it to the class where the temporary cluster center with the closest Euclidean distance is located;

[0117] Otherwise, reduce the second verification vectors of the small - proportion classes, and use the second verification vectors in the remaining classes as the third verification vectors, where the small - proportion class is a class in which the ratio of the number of second verification vectors in the class to the number of second verification vectors in the first vector group is less than the proportion threshold.

[0118] Exemplarily, as Figure 3 shown, in the embodiment of the present invention, after obtaining multiple second verification vectors 301, some of the vectors are reduced, aiming to retain more reliable and trustworthy vectors and avoid the problem of model overfitting caused by data noise.

[0119] The embodiments of the present invention perform grouped subtraction on multiple second verification vectors 301. The grouping method is arranged according to the second verification emissions 302, for example, arranged from small to large. After the arrangement, according to the equal intervals of the second verification emissions 302, it is divided into multiple first vector groups 303. For example, after the second verification emissions 302 are arranged, its value is N100 - N8100, and the equal interval division interval is N800, then it can be divided into 10 intervals, where the first interval is N100 - N900. Note that the number of vectors in the first vector group 303 obtained by this division method is different. Usually, the vector group closer to the median of the emissions has more vectors, while the vector group closer to the extreme value has fewer vectors.

[0120] Then, for each vector group, clustering is performed. Usually, the number of clusters is specified and clustering is performed according to the number of clusters. After clustering is completed to obtain multiple classes 304, the clustering results are analyzed. For those classes 304 with a small number of vectors in the vector group, subtraction will be performed. Generally, a ratio threshold is set. For example, the ratio threshold is 5%. That is to say, if the number of vectors in a certain class 304 is less than 5% of the number of vectors in the vector group, all the vectors in this class 304 will be deleted. In this way, vectors with strong noise are deleted, preparing for the accuracy of model construction.

[0121] In step 203, according to the corresponding second verification emissions, multiple third verification vectors are divided into multiple second vector groups. According to each second vector group and the multiple second verification emissions corresponding to the second vector group, multiple coefficients of the verification basic model are determined, and an enterprise emissions verification model is obtained.

[0122] In some embodiments, the step of dividing multiple third verification vectors into multiple second vector groups according to the corresponding second verification emissions, and determining multiple coefficients of the verification basic model according to each second vector group and the multiple second verification emissions corresponding to the second vector group, and obtaining an enterprise emissions verification model includes:

[0123] Arrange multiple third verification vectors according to the value of the corresponding second verification emissions;

[0124] Successively take out a preset number of verification vectors from the arrangement of the third verification vectors to construct a second vector group;

[0125] For each second vector group, perform the following steps respectively:

[0126] Configure the number of coefficients in the verification basic model according to the number of verification vectors in the second vector group;

[0127] Traversingly take out the third verification vectors from the second vector group as the vectors to be processed;

[0128] Determine a verification value according to a third formula, where the third formula is:

[0129]

[0130] In the formula, is the verification value, is the second verification emission amount corresponding to the vector to be processed, is the scaling factor, is the upper limit of the verification emission amount interval corresponding to the second vector group, is the lower limit of the verification emission amount interval corresponding to the second vector group, is the first offset coefficient;

[0131] Input the vector to be processed into the basic verification model, determine a model deviation according to the output of the basic verification model and the verification value, and add the model deviation to a deviation queue;

[0132] If the values of the last several digits in the deviation queue are all less than a deviation threshold, use the basic verification model as the enterprise emission verification model corresponding to the second vector group;

[0133] Otherwise, adjust multiple coefficients of the basic verification model according to the model deviation.

[0134] Exemplarily, as Figure 4 shown, the embodiment of the present invention performs grouped modeling based on the reduced verification vectors. Specifically, for the sake of distinction, the verification vectors obtained through the previous step of reduction are used as the third verification vectors 401, and they are sorted again according to the values of the corresponding verification emission amounts 302, and then grouped equally according to the number of vectors, obtaining the second vector group 402. Then, according to the number of vectors in the second vector group 402, configure the number of coefficients of the basic verification model 403. For example, by adjusting the total number of times, make the number of coefficients of the basic verification model 403 not more than the number of vectors in the second vector group 402. In this way, there will be no problems of multiple solutions or no solutions for the model coefficients, preventing overfitting and underfitting problems caused by unreasonable coefficient settings.

[0135] Then, the vectors in the second vector group 402 are traversed and extracted and input into the basic verification model 403 (the coefficients of the basic verification model 403 are generally randomly initialized, and some non-zero values are randomly set). The result 404 output by the model is compared with the verification value determined according to the verification emission amount corresponding to the vector, obtaining a model deviation 405. The verification value is determined according to the third formula:

[0136]

[0137] In the formula, is the verification value, is the second verified emission amount corresponding to the vector to be processed, is the scaling factor, is the upper limit of the verified emission amount interval corresponding to the second vector group, is the lower limit of the verified emission amount interval corresponding to the second vector group, is the first offset coefficient.

[0138] Generally, the scaling factor takes a value less than 1, for example, 0.8, and the first offset coefficient is a value less than the difference between 1 and the scaling factor. For example, when the scaling factor is 0.8, the first offset coefficient is 0.1.

[0139] The model deviation 405 is added to the deviation queue 406. Whether the condition for terminating the iteration is satisfied is determined according to the last few digits of the deviation queue 406 (for example, five consecutive digits). For example, if the absolute values of the model deviations 405 of the last few digits are all less than the threshold (for example, 5%), the iteration is terminated, and the verification basic model 403 can be used as the enterprise emission verification model at this time.

[0140] If the condition for terminating the iteration is not satisfied, an optimization algorithm is adopted. For example, the gradient descent method, the particle swarm method, or the genetic algorithm is used to optimize the model coefficients.

[0141] For example, the gradient descent method (Gradient Descent) is an iterative algorithm for optimizing the objective function. Its main purpose is to find the minimum value of the objective function (in some cases, it can also be used to find the maximum value. In the implementation mode of the present invention, it is to find the minimum value of the absolute value of the model deviation).

[0142] The gradient descent method uses the gradient of the objective function to determine the update direction of the parameters. The objective function of the present invention is clearly the difference between the model output and the verification value. The gradient is a vector, and each component of it is the partial derivative of the objective function with respect to the corresponding parameter. The direction of the gradient represents the direction in which the function rises fastest at this point, so its opposite direction is the direction of the fastest descent.

[0143] In each iteration, the gradient of the objective function is calculated according to the current parameter values.

[0144] Then, the parameters are updated according to the following formula:

[0145]

[0146] Where, is the value of the current parameter, is the updated parameter value, is the learning rate, is the gradient of the objective function at at.

[0147] The learning rate controls the step size of each parameter update. If the learning rate is too large, it may cause the algorithm to fail to converge or oscillate near the minimum value; if the learning rate is too small, the convergence speed of the algorithm will be very slow.

[0148] Through iterative optimization by the optimization algorithm, the coefficients will eventually converge to obtain a relatively accurate model.

[0149] The implementation manner of the enterprise emission verification method of the present invention first obtains an enterprise emission data queue and a material consumption data set, where the material consumption data set includes: energy consumption data and / or material consumption data; then extracts frequency domain features from the enterprise emission data queue, and constructs the obtained multiple frequency domain features and the material consumption data set into a first verification vector; then inputs the first verification vector into multiple verification models respectively, and determines multiple first results according to the outputs of the multiple verification models, where each first result corresponds to a verification model; finally, selects a target result from the multiple first results as the first verified emission amount, and verifies the enterprise emission monitoring amount according to the first verified emission amount, where the target result is the first result located in the emission verification interval, and each verification model corresponds to an emission verification interval. The implementation manner of the present invention predicts the emission amount based on the material consumption data and the emission data sequence that affect the enterprise emission, and verifies the enterprise emission amount according to the prediction result. Since it combines two types of data, namely factors and sequences, the verification result is more reliable and credible, and the verification efficiency is improved.

[0150] It should be understood that the magnitudes of the sequence numbers of the steps in the above implementation manner do not mean the order of execution. The execution order of each process should be determined according to its function and internal logic, and should not constitute any limitation to the implementation process of the implementation manner of the present invention.

[0151] The following is the device implementation manner of the present invention. For the details not described in detail therein, reference can be made to the corresponding method implementation manner above.

[0152] Figure 5 is the functional block diagram of the enterprise emission verification device provided by the implementation manner of the present invention. Refer to Figure 5 The enterprise emission verification device includes: a data acquisition module 501, a frequency domain analysis module 502, a verification result acquisition module 503, and a verification module 504, where:

[0153] The data acquisition module 501 is used to acquire an enterprise emission data queue and a material consumption data set, where the material consumption data set includes: energy consumption data and / or material consumption data;

[0154] A frequency domain analysis module 502 is configured to extract frequency domain features from the enterprise emission data queue, and construct a first verification vector by using the obtained multiple frequency domain features and the material consumption data set.

[0155] A verification result acquisition module 503 is configured to input the first verification vector into multiple verification models respectively, and determine multiple first results according to the outputs of the multiple verification models, where each first result corresponds to a verification model.

[0156] A verification module 504 is configured to select a target result from the multiple first results as the first verified emission amount, and verify the enterprise emission monitoring amount according to the first verified emission amount, where the target result is the first result located in the emission verification interval, and each verification model corresponds to an emission verification interval.

[0157] Figure 6 It is a functional block diagram of an electronic device provided by an embodiment of the present invention. As Figure 6 shown, the electronic device 6 of this embodiment includes: a processor 600 and a memory 601, and a computer program 602 that can run on the processor 600 is stored in the memory 601. When the processor 600 executes the computer program 602, the steps in the above-mentioned various enterprise emission verification methods and embodiments are implemented, such as Figure 1 the steps 101 to 104 shown.

[0158] Exemplarily, the computer program 602 can be divided into one or more modules / units, and the one or more modules / units are stored in the memory 601 and executed by the processor 600 to complete the present invention.

[0159] The electronic device 6 can be a computing device such as a desktop computer, a notebook, a palm computer, and a cloud server. The electronic device 6 may include, but is not limited to, a processor 600 and a memory 601. Those skilled in the art can understand that Figure 6 it is only an example of the electronic device 6, and does not constitute a limitation on the electronic device 6. It may include more or fewer components than shown in the figure, or combine some components, or different components. For example, the electronic device 6 may further include an input / output device, a network access device, a bus, etc.

[0160] The so-called processor 600 may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0161] The memory 601 may be an internal storage unit of the electronic device 6, such as the hard disk or memory of the electronic device 6. The memory 601 may also be an external storage device of the electronic device 6, such as a plug-in hard disk equipped on the electronic device 6, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. Further, the memory 601 may also include both the internal storage unit and the external storage device of the electronic device 6. The memory 601 is used to store the computer program 602 and other programs and data required by the electronic device 6. The memory 601 may also be used to temporarily store data that has been output or is to be output.

[0162] Those skilled in the art can clearly understand that, for the convenience and brevity of description, only the above-mentioned division of each functional unit and module is used as an example. In actual applications, the above-mentioned functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the device is divided into different functional units or modules to complete all or part of the functions described above. Each functional unit and module in the embodiments can be integrated into one processing unit, or each unit can exist physically alone, 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 a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of this application. The specific working processes of the units and modules in the above system can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated here.

[0163] In the above embodiments, the descriptions of the various embodiments have their own emphases. For the parts not detailed or recorded in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0164] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present invention.

[0165] In the embodiments provided by the present invention, it should be understood that the disclosed device / electronic device and method can be implemented in other ways. For example, the device / electronic device embodiments described above are merely illustrative. For example, the division of the modules or units is only a logical function division. In actual implementation, there may be other division methods. For example, 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 displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces. The indirect couplings or communication connections of the devices or units can be in electrical, mechanical or other forms.

[0166] The units described as separate components may or may not be physically separated. The components displayed as units may or may not be physical units, that is, they can be located in one place, or can be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] In addition, the functional units in each embodiment of the present invention can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit. The above integrated units can be implemented in the form of hardware or in the form of software functional units.

[0168] When 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 such understanding, to implement all or part of the processes in the above-described implementation manners of the present invention, it can also be completed by instructing 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 a processor, the steps of the above-described various method and apparatus implementation manners can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disc, computer memory, read-only memory (ROM), random access memory (RAM), electrical carrier signal, telecommunication signal, and software distribution medium, etc.

[0169] The above-described implementation manners are only used to illustrate the technical solutions of the present invention, rather than to limit it; although the present invention has been described in detail with reference to the foregoing implementation manners, those of ordinary skill in the art should understand that: they can still modify the technical solutions recorded in the foregoing implementation manners, or perform equivalent replacements on some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the various implementation manners of the present invention, and should all be included in the protection scope of the present invention.

Claims

1. A method for verifying enterprise emissions, characterized in that: include: Obtaining an enterprise emission data queue and a material consumption data set, wherein the enterprise emission data queue is a data queue obtained by arranging the pollutant emission data of multiple time nodes in a period before the time node of the enterprise emission monitoring amount in the order of time nodes, and the material consumption data set is a data set of the same period as the enterprise emission data queue, and the material consumption data set includes: energy consumption data and / or material consumption data; Extract frequency domain features from the enterprise emission data queue, and construct a first verification vector using the obtained multiple frequency domain features and the material consumption data set, wherein the first verification vector is a vector obtained by arranging the multiple frequency domain features and the data in the material consumption data set in a preset order; The first verification vectors are respectively input into a plurality of verification models, and the outputs of the plurality of verification models are respectively input into a second formula to obtain a plurality of first results, wherein each first result corresponds to a verification model, and the second formula is: In the formula, For the first result, To verify the output of the model, To verify the upper limit of the emission verification range of the model, To verify the lower limit of the emission verification range of the model, is the scaling factor, is the first bias coefficient; Selecting a target result from the multiple first results as a first verified emission amount, and verifying the enterprise's emission monitoring amount according to the first verified emission amount, wherein the target result is a first result located in an emission amount verification interval, and each verification model corresponds to an emission amount verification interval; Among them, the enterprise emission verification model is constructed based on multiple verification vectors and the emissions corresponding to the verification vectors to determine multiple parameters of the verification basic model. The construction process of the enterprise emission verification model includes: A verification basic model, a plurality of second verification vectors, and a plurality of second verification emissions are obtained, wherein each second verification vector corresponds to a second verification emission, and the second verification vector is constructed based on a historical material consumption data set and a plurality of frequency domain features extracted based on a historical enterprise emission data queue, and the historical material consumption data set includes historical energy consumption data and / or historical material consumption data, and the verification basic model is: In the formula, To verify the output of the model, is a natural constant, is a joint variable, For the The first coefficient, is the first elements, To verify the total number of vector elements, is the second bias coefficient, is the total number of times; dividing the plurality of second verification vectors into a plurality of first vector groups according to the intervals where the verified emission amounts are located, clustering the plurality of second verification vectors in each first vector group, reducing the plurality of second verification vectors in the first vector group according to the clustering result, and using the remaining second verification vectors as third verification vectors; According to the corresponding second verified emissions, multiple third verification vectors are divided into multiple second vector groups, and multiple coefficients of the verification basic model are determined according to each second vector group and the multiple second verified emissions corresponding to the second vector group to obtain the enterprise emission verification model.

2. The method for verifying enterprise emissions according to claim 1, characterized in that: The extracting frequency domain features from the enterprise emission data queue and constructing a first verification vector using the obtained multiple frequency domain features and the material consumption data set includes: A plurality of frequency domain features of the enterprise emission data queue are extracted according to a first formula, wherein the first formula is: In the formula, For the Sub-frequency domain features, For the first data, The total amount of data in the enterprise emission data queue, is a natural constant, is the circumference of a circle, is the fundamental frequency, is an imaginary unit, The total number of enterprise emission data obtained during the base wave period; The plurality of frequency domain features and the data in the material consumption data set are arranged according to a predetermined frequency feature order and a predetermined material order to obtain a first verification vector.

3. The enterprise emission verification method according to claim 1, characterized in that: The method of dividing the plurality of second verification vectors into a plurality of first vector groups according to the interval where the verified emission amount is located, clustering the plurality of second verification vectors in each first vector group, reducing the plurality of second verification vectors in the first vector group according to the clustering result, and using the remaining second verification vectors as third verification vectors includes: Arranging the plurality of second verified emissions according to the magnitude of the values; According to the intervals corresponding to the arrangement, the plurality of second verified emissions are divided into equal interval spans to obtain a plurality of emission groups, and according to the correspondence between the second verified emissions and the second verification vectors and the plurality of emission groups, the plurality of second verification vectors are divided into a plurality of first vector groups, wherein each first vector group corresponds to an emission group; For each first vector group, perform the following steps respectively: Get the number of clusters; According to the number of clusters, randomly selecting second verification vectors of the number of clusters from the first vector group as multiple temporary cluster centers; For each second verification vector in the first vector group, add the cluster where the temporary cluster center with the closest Euclidean distance is located; Obtaining class centers of multiple classes, wherein the class center is a second verification vector that is closest to a class mean vector, and the class mean vector is a mean vector of multiple vectors in the class; If there is a situation where the multiple temporary clustering centers are not class centers, the class centers of the multiple classes are used as multiple temporary clustering centers, and the process jumps to the step of adding the class where the temporary clustering center with the closest Euclidean distance is located for each second verification vector in the first vector group; Otherwise, the second verification vectors of the small-proportion class are eliminated, and the second verification vectors in the remaining classes are used as the third verification vectors, wherein the small-proportion class is a class in which the ratio of the number of second verification vectors in the class to the number of second verification vectors in the first vector group is less than the ratio threshold.

4. The enterprise emission verification method according to claim 1, characterized in that: The method of dividing the plurality of third verification vectors into a plurality of second vector groups according to the corresponding second verification emission amounts, determining a plurality of coefficients of the verification basic model according to each second vector group and the plurality of second verification emission amounts corresponding to the second vector group, and obtaining the enterprise emission verification model includes: Arranging a plurality of third verification vectors according to values ​​corresponding to the second verification emission amounts; Sequentially taking out a preset number of verification vectors from the third verification vector arrangement to construct a second vector group; For each second vector group, perform the following steps respectively: Configure the number of coefficients in the verification basic model according to the number of verification vectors in the second vector group; traversally taking out a third verification vector from the second vector group as a vector to be processed; According to the third formula, the verification value is determined, wherein the third formula is: In the formula, To verify the value, is the second verified emission corresponding to the vector to be processed, is the scaling factor, is the upper limit of the verified emission range corresponding to the second vector group, is the lower limit of the verified emission range corresponding to the second vector group, is the first bias coefficient; Input the vector to be processed into the verification basic model, determine the model deviation according to the output of the verification basic model and the verification value, and add the model deviation to the deviation queue; If the values ​​of the last digits in the deviation queue are all smaller than the deviation threshold, the verification basic model is used as the enterprise emission verification model corresponding to the second vector group; Otherwise, multiple coefficients of the basic model are adjusted and verified according to the model deviation.

5. An enterprise emission verification device, characterized in that: Used to implement the enterprise emission verification method according to any one of claims 1-2, the enterprise emission verification device comprises: A data acquisition module, used to acquire an enterprise emission data queue and a material consumption data set, wherein the material consumption data set includes: energy consumption data and / or material consumption data; A frequency domain analysis module, used for extracting frequency domain features from the enterprise emission data queue, and constructing a first verification vector with the obtained multiple frequency domain features and the material consumption data set; a verification result acquisition module, configured to input the first verification vector into a plurality of verification models respectively, and determine a plurality of first results according to outputs of the plurality of verification models, wherein each first result corresponds to a verification model; as well as, A verification module is used to select a target result from the multiple first results as the first verified emission amount, and verify the enterprise emission monitoring amount based on the first verified emission amount, wherein the target result is the first result located in the emission verification interval, and each verification model corresponds to an emission verification interval.

6. 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 4 are implemented.

7. 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 according to any one of claims 1 to 4 are implemented.

Images