CN115683255A

CN115683255A - Oil consumption metering system and method based on data collaborative analysis

Info

Publication number: CN115683255A
Application number: CN202211252777.8A
Authority: CN
Inventors: 张凯元
Original assignee: Shaanxi Junkai Electronic Technology Co ltd
Current assignee: Shaanxi Junkai Electronic Technology Co ltd
Priority date: 2022-10-13
Filing date: 2022-10-13
Publication date: 2023-02-03

Abstract

The application relates to the field of data analysis, and particularly discloses a fuel consumption metering system and a fuel consumption metering method based on data collaborative analysis, wherein high-dimensional implicit nonlinear correlation among all related data items influencing fuel consumption is mined by using a feature extractor based on a deep neural network model, meanwhile, high-dimensional implicit correlation features of a fuel consumption sequence are extracted by using the feature extractor based on the deep neural network model, then, collaborative analysis is carried out on Gao Weite of the fuel consumption sequence and measurement parameter features to mine the deep implicit correlation between measurement parameters and the fuel consumption, and a fuel consumption metering scheme based on the data collaborative analysis with higher measurement accuracy is constructed on the basis.

Description

Oil consumption metering system and method based on data collaborative analysis

Technical Field

The present application relates to the field of data analysis, and more particularly, to a fuel consumption metering system and method based on data collaborative analysis.

Background

When a user purchases a vehicle or uses the vehicle, the user may pay attention to the fuel consumption of the vehicle, because the fuel consumption is closely related to the vehicle cost of the user. For the oil consumption value of the automobile, the industry and the credit department can give an oil consumption reference value, but in the actual industry and life, the reference value and the actual oil consumption value of the automobile have larger deviation, and the accurate evaluation of the user on the automobile performance and the automobile use cost is influenced.

Patent application 201922008287.3 discloses an agricultural machinery oil consumption metering system, which cannot achieve accurate evaluation effect.

Therefore, a more optimized fuel consumption metering scheme based on data synergy analysis is expected.

Disclosure of Invention

The present application is proposed to solve the above-mentioned technical problems. The embodiment of the application provides a fuel consumption metering system and method based on data collaborative analysis, high-dimensional implicit nonlinear correlation between all relevant data items influencing fuel consumption is mined by using a feature extractor based on a deep neural network model, meanwhile, high-dimensional implicit correlation features of a fuel consumption sequence are extracted by using the feature extractor based on the deep neural network model, then, the Gao Weite of the fuel consumption sequence and measurement parameter features are subjected to collaborative analysis to mine the deep implicit correlation between measurement parameters and the fuel consumption, and a fuel consumption metering scheme based on the data collaborative analysis and with higher measurement accuracy is constructed based on the method.

According to an aspect of the present application, there is provided a fuel consumption metering system based on data integration analysis, including:

the system comprises a test data acquisition module, a data processing module and a data processing module, wherein the test data acquisition module is used for acquiring a plurality of groups of vehicle detection data, and each group of vehicle detection data comprises vehicle speed, running time, wind resistance, total work, power, heat efficiency, heat value and oil consumption value;

the relevant data coding module is used for arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in the plurality of groups of vehicle detection data into a measurement input matrix according to the data item dimension and the group dimension and then obtaining a measurement characteristic vector through a convolutional neural network model serving as a filter;

the oil consumption data coding module is used for arranging oil consumption values in the plurality of groups of vehicle detection data into oil consumption input vectors and then obtaining multi-scale neighborhood oil consumption characteristic vectors through the multi-scale neighborhood characteristic extraction module;

the co-integration analysis module is used for performing co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain a co-integration eigenvector matrix;

the detection data acquisition module is used for acquiring the vehicle speed values of a vehicle to be detected at a plurality of preset time points in a preset time period;

the detection data coding module is used for enabling the vehicle speed values of the plurality of preset time points to pass through a time sequence coder containing a one-dimensional convolution layer so as to obtain a vehicle speed characteristic vector;

the vector query module is used for multiplying the vehicle speed characteristic vector serving as a query characteristic vector by the coordinated characteristic matrix to obtain a decoding characteristic vector;

the decoding optimization module is used for correcting the characteristic values of all positions in the decoding characteristic vector based on the reciprocal of the mean value of the characteristic values of all the positions of the decoding characteristic vector to obtain a corrected decoding characteristic vector; and

and the oil consumption metering module is used for enabling the corrected decoding eigenvector to pass through a decoder to obtain a decoding value, and the decoding value is an oil consumption value.

In the above oil consumption metering system based on data collaborative analysis, the related data encoding module includes: the row vector construction unit is used for arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in each group of vehicle detection data into row vectors so as to obtain a plurality of row vectors; the two-dimensional arrangement unit is used for performing two-dimensional arrangement on the plurality of row vectors to obtain the measurement input matrix; and a deep convolutional encoding unit for performing convolutional processing, feature matrix-based pooling processing, and nonlinear activation processing on input data in forward pass of layers using the layers of the convolutional neural network model as a filter, respectively, to output the measurement feature vector from a last layer of the convolutional neural network model as a filter, wherein an input of a first layer of the convolutional neural network model as a filter is the measurement input matrix.

In the above oil consumption metering system based on data collaborative analysis, the oil consumption data encoding module includes: the first scale neighborhood convolutional coding unit is used for inputting the oil consumption input vector into a first convolutional layer of the multi-scale neighborhood feature extraction module to obtain a first scale oil consumption feature vector, wherein the first convolutional layer has a first one-dimensional convolutional kernel with a first length; a second scale neighborhood convolutional coding unit, configured to input the fuel consumption input vector into a second convolutional layer of the multi-scale neighborhood feature extraction module to obtain a second scale fuel consumption feature vector, where the second convolutional layer has a second one-dimensional convolutional kernel with a second length, and the first length is different from the second length; and the cascade unit is used for cascading the first scale oil consumption characteristic vector and the second scale oil consumption characteristic vector to obtain the oil consumption characteristic vector.

In the above oil consumption metering system based on data covariance analysis, the first scale neighborhood convolutional coding unit is further configured to: performing one-dimensional convolution coding on the oil consumption input vector by using a first convolution layer of the multi-scale neighborhood characteristic extraction module according to the following formula to obtain a first scale oil consumption characteristic vector;

wherein the formula is:

wherein a is the width of the first convolution kernel in the X direction, F (a) is a first convolution kernel parameter vector, G (X-a) is a local vector matrix operated with a convolution kernel function, w is the size of the first convolution kernel, and X represents the fuel consumption input vector.

In the oil consumption metering system based on data covariance analysis, the second scale neighborhood convolutional coding unit is further configured to: performing one-dimensional convolution coding on the oil consumption input vector by using a second convolution layer of the multi-scale neighborhood characteristic extraction module according to the following formula to obtain a second scale oil consumption characteristic vector;

wherein the formula is:

wherein b is the width of the second convolution kernel in the X direction, F (b) is a second convolution kernel parameter vector, G (X-b) is a local vector matrix operated with the convolution kernel function, m is the size of the second convolution kernel, and X represents the fuel consumption input vector.

In the oil consumption metering system based on data co-integration analysis, the co-integration analysis module is further configured to perform co-integration analysis on the measurement eigenvector and the multi-scale neighborhood oil consumption eigenvector by using the following formula to obtain the co-integration eigenvector;

wherein the formula is:

wherein v is ₁ Is the multi-scale neighborhood fuel consumption feature vector v ₂ For the measured feature vector, M _c And integrating the feature matrix for the coordination.

In the oil consumption metering system based on data collaborative analysis, the decoding optimization module is further configured to: based on the reciprocal of the mean value of the feature values of all the positions of the decoding feature vector, correcting the feature values of all the positions in the decoding feature vector by the following formula to obtain the corrected decoding feature vector; wherein the formula is:

where V represents the decoded feature vector,

the inverse of the mean value of the eigenvalues of all the positions in the decoded eigenvector, indicates dot-by-dot multiplication.

In the oil consumption metering system based on the data co-integration analysis, the oil consumption metering module is further configured to perform decoding regression on the corrected decoding eigenvector by using the decoder according to the following formula to obtain the decoded value; wherein the formula is:

where V' is the corrected decoded feature vector, Y is the decoded value, W is the weight matrix, B is the offset vector,

representing the matrix multiplication, h (-) is the activation function.

According to another aspect of the present application, a fuel consumption metering method based on data integration analysis is provided, which includes:

acquiring a plurality of groups of vehicle detection data, wherein each group of vehicle detection data comprises vehicle speed, running time, wind resistance, total work, power, heat efficiency, heat value and oil consumption value;

arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in the plurality of groups of vehicle detection data into a measurement input matrix according to the data item dimension and the group dimension, and then obtaining a measurement characteristic vector through a convolutional neural network model serving as a filter;

arranging the oil consumption values in the multiple groups of vehicle detection data into oil consumption input vectors, and then obtaining multi-scale neighborhood oil consumption characteristic vectors through a multi-scale neighborhood characteristic extraction module;

performing a co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain a co-integration eigenvector matrix;

acquiring vehicle speed values of a vehicle to be detected at a plurality of preset time points within a preset time period;

passing the vehicle speed values of the plurality of preset time points through a time sequence encoder comprising a one-dimensional convolution layer to obtain a vehicle speed characteristic vector;

multiplying the vehicle speed eigenvector serving as a query eigenvector by the co-integration eigenvector matrix to obtain a decoding eigenvector;

correcting the feature values of all positions in the decoded feature vector based on the reciprocal of the mean value of the feature values of all the positions of the decoded feature vector to obtain a corrected decoded feature vector; and

and enabling the corrected decoding eigenvector to pass through a decoder to obtain a decoding value, wherein the decoding value is the oil consumption value.

According to still another aspect of the present application, there is provided an electronic apparatus including: a processor; and a memory having stored therein computer program instructions that, when executed by the processor, cause the processor to perform the fuel consumption metering method as described above.

According to yet another aspect of the present application, a computer-readable medium is provided, on which computer program instructions are stored, which, when executed by a processor, cause the processor to perform the fuel consumption metering method as described above.

Compared with the prior art, the oil consumption metering system and the method based on the data collaborative analysis have the advantages that the characteristic extractor based on the deep neural network model is used for mining high-dimensional implicit nonlinear correlation among all related data items influencing oil consumption, meanwhile, the characteristic extractor based on the deep neural network model is used for extracting high-dimensional implicit correlation characteristics of an oil consumption sequence, then, the Gao Weite of the oil consumption sequence and the measured parameter characteristics are subjected to collaborative analysis to mine the deep implicit correlation between the measured parameters and the oil consumption, and the oil consumption metering scheme based on the data collaborative analysis and with higher measuring accuracy is constructed based on the oil consumption metering system and the method.

Drawings

The above and other objects, features and advantages of the present application will become more apparent by describing in more detail embodiments of the present application with reference to the attached drawings. The accompanying drawings are included to provide a further understanding of the embodiments of the application and are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description serve to explain the principles of the application. In the drawings, like reference numbers generally represent like parts or steps.

Fig. 1 illustrates an application scenario of a fuel consumption metering system based on data integration analysis according to an embodiment of the present application;

FIG. 2 illustrates a block diagram of a fuel consumption metering system based on data integration analysis according to an embodiment of the application;

FIG. 3 is a diagram illustrating a system architecture of a fuel consumption metering system based on data integration analysis according to an embodiment of the present application;

FIG. 4 is a block diagram illustrating a relevant data encoding module in a fuel consumption metering system based on data integration analysis according to an embodiment of the present application;

FIG. 5 is a block diagram illustrating an oil consumption data encoding module in an oil consumption metering system based on data integration analysis according to an embodiment of the present application;

FIG. 6 is a flow chart illustrating a method for fuel consumption measurement based on data synergy analysis according to an embodiment of the present application;

FIG. 7 illustrates a block diagram of an electronic device in accordance with an embodiment of the application.

Detailed Description

Hereinafter, example embodiments according to the present application will be described in detail with reference to the accompanying drawings. It should be understood that the described embodiments are only some embodiments of the present application and not all embodiments of the present application, and that the present application is not limited by the example embodiments described herein.

Overview of a scene

As described above, the industry and the credit department give the fuel consumption reference value, but in the actual industry and life, the reference value and the actual fuel consumption value of the automobile have large deviation, and the accurate evaluation of the automobile performance and the automobile use cost by the user is influenced. Therefore, a more optimized fuel consumption metering scheme is desired.

Research and test show that the actual fuel consumption of the vehicle is related to a plurality of factors including but not limited to vehicle speed, running time, wind resistance, total work, power, thermal efficiency, heat value and the like, and the correlation between the fuel consumption value and the influencing factors is a complex nonlinear correlation. Some manufacturers try to construct a more accurate statistical model of fuel consumption through the statistical model, but on one hand, the statistical model omits some parameter items for convenience of calculation, such as wind resistance, and on the other hand, the statistical model simply understands a nonlinear association as a combination of a plurality of linear sections in the modeling process, for example, setting different influence coefficients by taking speed as a boundary. Although the statistical model can relatively accurately calculate the fuel consumption, the statistical model is still lack of precision, and some parameter items are ignored, so that when the parameter items become important influence factors, the calculation precision of the fuel consumption is greatly deviated, for example, in windy weather, the wind resistance becomes an important influence factor.

In order to solve the problems, in the technical scheme of the application, a feature extractor based on a deep neural network model is used for mining high-dimensional implicit nonlinear correlation among all related data items influencing oil consumption, meanwhile, a feature extractor based on the deep neural network model is used for extracting high-dimensional implicit correlation features of an oil consumption sequence, and then the Gao Weite of the oil consumption sequence and measurement parameter features are subjected to collaborative analysis to mine the deep implicit correlation between measurement parameters and the oil consumption, so that an oil consumption metering scheme with higher measurement accuracy is constructed.

Specifically, a plurality of groups of vehicle detection data are obtained, wherein each group of vehicle detection data comprises vehicle speed, running time, wind resistance, total work, power, thermal efficiency, heat value and oil consumption value. And then, arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in the plurality of groups of vehicle detection data into a measurement input matrix according to the data item dimension and the group dimension, and then obtaining a measurement characteristic vector through a convolutional neural network model serving as a filter. That is, the oil consumption influencing parameter items in the plurality of sets of vehicle detection data are two-dimensionally structured, for example, the vehicle speed, the running time, the wind resistance, the total work, the power, the thermal efficiency and the heat value in each set of vehicle detection data are firstly arranged into row vectors to obtain a plurality of row vectors, and then the plurality of row vectors are two-dimensionally arranged to obtain the measurement input matrix; then, a convolutional neural network model having excellent performance in the local feature extraction field is used as a feature filter to extract high-dimensional local implicit features in the measurement input matrix, that is, high-dimensional implicit associated features of association between different oil consumption amount influence parameter items.

Meanwhile, the oil consumption values in the multiple groups of vehicle detection data are arranged into oil consumption input vectors, and then the multi-scale neighborhood oil consumption characteristic vectors are obtained through a multi-scale neighborhood characteristic extraction module. That is, after the oil consumption values in the multiple sets of vehicle detection data are subjected to one-dimensional vectorization, the multi-scale neighborhood feature extraction module is used for performing multi-scale one-dimensional convolution coding on the oil consumption input vector so as to extract the association mode features between the oil consumption values of different spans in the oil consumption input vector. It should be appreciated that different scales of one-dimensional convolutional encoding of the fuel consumption input vector by different scales of one-dimensional convolutional cores may utilize different receptive fields to capture richer and more hierarchical representations of features in the fuel consumption input vector.

And then, carrying out the co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain a co-integration eigenvector matrix. That is, the deep implicit multidimensional associations between fuel consumption characteristics and measurement characteristics are mined based on a collaborative analysis. In a specific example of the present application, the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector are subjected to a co-integration analysis by using the following formula to obtain the co-integration eigenvector matrix; wherein the formula is:

wherein v is ₁ Is the multi-scale neighborhood fuel consumption feature vector v ₂ For the measured feature vector, M _c And integrating the feature matrix for the coordination. Namely, at the data level, the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector are represented by a transfer matrix.

After the co-integration feature matrix is obtained, the co-integration feature matrix can be regarded as a feature library, and the feature vector obtained through actual measurement is used as a query vector, so that a decoding feature vector can be obtained based on feature query retrieval, and then a decoding result for representing the oil consumption value can be obtained by using a decoder.

In the technical solution of the present application, since the factor most convenient to measure actually is the vehicle speed value, in the technical solution of the present application, the vehicle speed value is used as an input vector, and the obtained vehicle speed values at a plurality of predetermined time points are mapped to a high-dimensional feature space by using a time-series encoder including a one-dimensional convolution layer to obtain a vehicle speed feature vector. And then, multiplying the vehicle speed feature vector serving as a query feature vector by the co-integration feature matrix to obtain a decoding feature vector. It should be understood that although the vehicle feature vector is used as the query vector, the matched decoded feature vector contains other parameter features and information that affect the oil consumption, and therefore, decoding the decoded feature vector by using the decoder can obtain a more accurate decoded value representing the oil consumption.

Particularly, in the technical solution of the present application, since each eigenvalue of the co-integration eigenvector obtained by performing co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector has a position attribute, for example, each position in the transfer matrix represents relationship information between corresponding positions of vectors, after matrix-multiplying the vehicle speed eigenvector as the query eigenvector with the co-integration eigenvector, each position of the obtained decoded eigenvector also has a position attribute.

However, when the decoded feature vector is subjected to decoding regression by a decoder, the decoding regression belongs to a substantial regression task without a position attribute, and thus a decoded value deviation may be caused by not considering position attribute information of the decoded feature vector.

Based on this, in the technical solution of the present application, the vectors for performing phase sensing on the decoded feature vector are preferably aggregated by location, and are represented as:

the phase perception characterization of the vector introduces the amplitude-phase quasi-real value-imaginary value characterization, and the vector is subjected to position splicing expansion of real value vectors based on the principle of Euler's formula, so that regression numerical value deviation possibly caused when the vector is subjected to a real value regression task without position attributes is compensated in a multi-layer perception mode. Thus, the accuracy of decoding regression is improved, i.e., a more accurate calculated value of oil consumption is obtained.

Based on this, this application has proposed a fuel consumption measurement system based on data integration analysis, and it includes: the system comprises a test data acquisition module, a data processing module and a data processing module, wherein the test data acquisition module is used for acquiring a plurality of groups of vehicle detection data, and each group of vehicle detection data comprises vehicle speed, running time, wind resistance, total work, power, heat efficiency, heat value and oil consumption value; the relevant data coding module is used for arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in the plurality of groups of vehicle detection data into a measurement input matrix according to the data item dimension and the group dimension and then obtaining a measurement characteristic vector through a convolutional neural network model serving as a filter; the oil consumption data coding module is used for arranging oil consumption values in the plurality of groups of vehicle detection data into oil consumption input vectors and then obtaining multi-scale neighborhood oil consumption characteristic vectors through the multi-scale neighborhood characteristic extraction module; the co-integration analysis module is used for performing co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain a co-integration eigenvector matrix; the detection data acquisition module is used for acquiring the vehicle speed values of a vehicle to be detected at a plurality of preset time points in a preset time period; the detection data coding module is used for enabling the vehicle speed values of the plurality of preset time points to pass through a time sequence coder containing a one-dimensional convolution layer so as to obtain a vehicle speed characteristic vector; the vector query module is used for multiplying the vehicle speed characteristic vector serving as a query characteristic vector by the co-integration characteristic matrix to obtain a decoding characteristic vector; the decoding optimization module is used for correcting the characteristic values of all positions in the decoding characteristic vector based on the reciprocal of the mean value of the characteristic values of all the positions of the decoding characteristic vector to obtain a corrected decoding characteristic vector; and the oil consumption metering module is used for enabling the corrected decoding eigenvector to pass through a decoder to obtain a decoding value, and the decoding value is the oil consumption value.

Fig. 1 illustrates an application scenario of a fuel consumption metering system based on data integration analysis according to an embodiment of the present application. As shown in fig. 1, in the application scenario, a plurality of sets of vehicle detection data are obtained through vehicle detection sensors (e.g., C1-Cn as illustrated in fig. 1), wherein each set of vehicle detection data includes a vehicle speed, a driving time, a wind resistance, a total work, a power, a thermal efficiency, a heat value, and a fuel consumption value; meanwhile, vehicle speed values of a vehicle to be detected at a plurality of predetermined time points within a predetermined time period are acquired through a vehicle speed sensor (for example, sp as illustrated in fig. 1); then, the plurality of sets of vehicle detection data and the vehicle speed values at the plurality of predetermined time points in the predetermined time period are input into a server (for example, S in fig. 1) deployed with a fuel consumption metering algorithm based on data coordination analysis, wherein the server can process the data with the fuel consumption metering algorithm based on data coordination analysis to generate the fuel consumption value.

Having described the general principles of the present application, various non-limiting embodiments of the present application will now be described with reference to the accompanying drawings.

Exemplary System

Fig. 2 illustrates a block diagram of a fuel consumption metering system based on data covariance analysis according to an embodiment of the application. As shown in fig. 2, a fuel consumption metering system 300 based on data integration analysis according to an embodiment of the present application includes: a test data acquisition module 310; a related data encoding module 320; a fuel consumption data encoding module 330; a co-integration analysis module 340; a detection data acquisition module 350; a detection data encoding module 360; a vector query module 370; a decode optimization module 380; and a fuel consumption metering module 390.

The test data acquisition module 310 is configured to acquire a plurality of sets of vehicle detection data, where each set of vehicle detection data includes a vehicle speed, a driving time, a wind resistance, a total work, a power, a thermal efficiency, a heat value, and a fuel consumption value; the related data encoding module 320 is configured to arrange vehicle speed, travel time, wind resistance, total work, power, thermal efficiency, and thermal value in the multiple sets of vehicle detection data into a measurement input matrix according to data item dimensions and group dimensions, and then obtain a measurement feature vector through a convolutional neural network model serving as a filter; the oil consumption data encoding module 330 is configured to arrange oil consumption values in the multiple groups of vehicle detection data into oil consumption input vectors, and then obtain multi-scale neighborhood oil consumption feature vectors through the multi-scale neighborhood feature extraction module; the co-integration analysis module 340 is configured to perform co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain a co-integration eigenvector matrix; the detection data acquisition module 350 is configured to acquire vehicle speed values of a vehicle to be detected at a plurality of predetermined time points within a predetermined time period; the detection data encoding module 360 is configured to pass the vehicle speed values at the plurality of predetermined time points through a time sequence encoder including a one-dimensional convolution layer to obtain a vehicle speed feature vector; the vector query module 370 is configured to multiply the vehicle speed feature vector, which is used as a query feature vector, by the co-integration feature matrix to obtain a decoded feature vector; the decoding optimization module 380 is configured to correct the feature values of each position in the decoded feature vector based on a reciprocal of a mean value of the feature values of all positions of the decoded feature vector to obtain a corrected decoded feature vector; and the oil consumption metering module 390 is configured to pass the corrected decoded eigenvector through a decoder to obtain a decoded value, where the decoded value is an oil consumption value.

Fig. 3 is a system architecture diagram illustrating a fuel consumption metering system 300 based on data synergy analysis according to an embodiment of the present application. As shown in fig. 3, first, a plurality of sets of vehicle detection data are obtained through the test data acquisition module 310, where each set of vehicle detection data includes a vehicle speed, a driving time, a wind resistance, a total work, a power, a thermal efficiency, a heat value, and a fuel consumption value; then, the related data encoding module 320 arranges the vehicle speed, the driving time, the wind resistance, the total work, the power, the thermal efficiency and the thermal value in the plurality of groups of vehicle detection data into a measurement input matrix according to the dimension of the data item and the dimension of the group, and then obtains a measurement feature vector through a convolutional neural network model serving as a filter; the oil consumption data encoding module 330 arranges the oil consumption values in the multiple groups of vehicle detection data into oil consumption input vectors, and then obtains multi-scale neighborhood oil consumption characteristic vectors through a multi-scale neighborhood characteristic extraction module; then, the co-integration analysis module 340 performs co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain a co-integration eigenvector matrix; the detection data acquisition module 350 acquires vehicle speed values of a vehicle to be detected at a plurality of preset time points within a preset time period; the detection data encoding module 360 passes the vehicle speed values of the plurality of predetermined time points through a time sequence encoder comprising a one-dimensional convolution layer to obtain a vehicle speed feature vector; the vector query module 370 multiplies the vehicle speed eigenvector as a query eigenvector by the co-integration eigenvector matrix to obtain a decoding eigenvector; the decoding optimization module 380 corrects the eigenvalues of each position in the decoded eigenvector based on the reciprocal of the mean value of the eigenvalues of all positions of the decoded eigenvector to obtain a corrected decoded eigenvector; further, the fuel consumption metering module 390 passes the corrected decoded eigenvector through a decoder to obtain a decoded value, where the decoded value is a fuel consumption value.

Specifically, in the operation process of the oil consumption metering system 300 based on data collaborative analysis, the test data acquisition module 310 is configured to acquire a plurality of sets of vehicle detection data, where each set of vehicle detection data includes a vehicle speed, a driving time, a wind resistance, a total work, a power, a thermal efficiency, a heat value, and an oil consumption value. The actual oil consumption of the vehicle is related to a plurality of factors, wherein the factors comprise vehicle speed, running time, wind resistance, total work, power, thermal efficiency and heat value, as much related data as possible is adopted, the metering accuracy of the actual oil consumption is higher, and in the technical scheme of the application, the data can be detected through the vehicle detection sensor.

Specifically, in the operation process of the oil consumption metering system 300 based on data collaborative analysis, the related data encoding module 320 is configured to arrange the vehicle speed, the driving time, the wind resistance, the total work, the power, the thermal efficiency, and the heat value in the plurality of groups of vehicle detection data into a measurement input matrix according to the data item dimension and the group dimension, and then obtain a measurement feature vector through a convolutional neural network model serving as a filter. It should be understood that the correlation between the fuel consumption value and the factors such as the vehicle speed, the running time, the wind resistance, the total work, the power, the thermal efficiency, the heat value and the like is a complex nonlinear correlation, and the influencing factors are arranged into a measurement input matrix according to the data item dimension and the group dimension and then pass through a convolutional neural network model serving as a filter to obtain a measurement characteristic vector, that is, the oil consumption influencing parameter items in the multiple groups of vehicle detection data are two-dimensionally structured, for example, the vehicle speed, the running time, the wind resistance, the total work, the power, the thermal efficiency and the heat value in each group of vehicle detection data are firstly arranged into a plurality of row vectors, and then the plurality of row vectors are two-dimensionally arranged to obtain the measurement input matrix; then, a convolutional neural network model having excellent performance in the local feature extraction field is used as a feature filter to extract high-dimensional local implicit features in the measurement input matrix, that is, high-dimensional implicit associated features of association between different oil consumption amount influence parameter items.

Fig. 4 is a block diagram illustrating a relevant data encoding module in a fuel consumption metering system based on data integration analysis according to an embodiment of the present application. As shown in fig. 4, the related data encoding module 320 includes: a row vector construction unit 321 configured to arrange the vehicle speed, the driving time, the wind resistance, the total work, the power, the thermal efficiency, and the heat value in each set of the vehicle detection data into row vectors to obtain a plurality of row vectors; a two-dimensional arrangement unit 322, configured to perform two-dimensional arrangement on the plurality of row vectors to obtain the measurement input matrix; and a deep convolutional encoding unit 323 for performing convolutional processing, feature matrix-based pooling processing, and nonlinear activation processing on input data in forward pass of layers using the layers of the convolutional neural network model as a filter, respectively, to output the measurement feature vector from the last layer of the convolutional neural network model as a filter, wherein an input of the first layer of the convolutional neural network model as a filter is the measurement input matrix.

Specifically, in the operation process of the oil consumption metering system 300 based on data co-integration analysis, the oil consumption data encoding module 330 is configured to arrange oil consumption values in the multiple sets of vehicle detection data into an oil consumption input vector, and then obtain a multi-scale neighborhood oil consumption feature vector through the multi-scale neighborhood feature extraction module. It should be appreciated that different scales of one-dimensional convolutional encoding of the fuel consumption input vector by different scales of one-dimensional convolutional cores may utilize different receptive fields to capture richer and more hierarchical representations of features in the fuel consumption input vector. In the technical scheme of the application, after the oil consumption values in the plurality of groups of vehicle detection data are subjected to one-dimensional vectorization, the multi-scale neighborhood feature extraction module is used for carrying out multi-scale one-dimensional convolution coding on the oil consumption input vector so as to extract the association mode features among the oil consumption values of different spans in the oil consumption input vector.

Fig. 5 illustrates a block diagram of an oil consumption data encoding module in an oil consumption metering system based on data integration analysis according to an embodiment of the present application. As shown in fig. 5, the oil consumption data encoding module 330 includes: a first scale neighborhood convolution unit 331, configured to input the fuel consumption input vector into a first convolution layer of the multi-scale neighborhood feature extraction module to obtain a first scale fuel consumption feature vector, where the first convolution layer has a first one-dimensional convolution kernel with a first length; a second scale neighborhood convolutional coding unit 332, configured to input the fuel consumption input vector into a second convolutional layer of the multi-scale neighborhood feature extraction module to obtain a second scale fuel consumption feature vector, where the second convolutional layer has a second one-dimensional convolutional kernel with a second length, and the first length is different from the second length; and the cascading unit 333 is configured to cascade the first scale fuel consumption feature vector and the second scale fuel consumption feature vector to obtain the fuel consumption feature vector.

More specifically, the first scale neighborhood convolution unit is further configured to: performing one-dimensional convolution coding on the oil consumption input vector by using a first convolution layer of the multi-scale neighborhood characteristic extraction module according to the following formula to obtain a first scale oil consumption characteristic vector;

wherein the formula is:

wherein the formula is:

Specifically, in the operation process of the oil consumption metering system 300 based on data co-integration analysis, the co-integration analysis module 340 is configured to perform co-integration analysis on the measurement eigenvector and the multi-scale neighborhood oil consumption eigenvector to obtain a co-integration eigenvector. It should be understood that the feature extractor based on the deep neural network model extracts the high-dimensional implicit association features of the oil consumption sequence, and then performs the synergistic analysis on the Gao Weite and the measured parameter features of the oil consumption sequence to mine the deep implicit association between the measured parameters and the oil consumption, so as to construct the oil consumption metering scheme with higher measurement accuracy. In a specific example of the present application, the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector are subjected to a co-integration analysis by using the following formula to obtain the co-integration eigenvector matrix; wherein the formula is:

wherein v is ₁ Is the multi-scale neighborhood fuel consumption feature vector v ₂ For the measured feature vector，M _c And integrating the feature matrix for the coordination. Namely, at the data level, the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector are represented by a transfer matrix.

Specifically, in the operation process of the oil consumption metering system 300 based on data collaborative analysis, the detection data acquisition module 350 is configured to acquire vehicle speed values of a vehicle to be detected at a plurality of predetermined time points within a predetermined time period. After the co-integration feature matrix is obtained, the co-integration feature matrix can be regarded as a feature library, and the feature vector obtained through actual measurement is used as a query vector, so that a decoding feature vector can be obtained based on feature query retrieval, and then a decoding result for representing the oil consumption value can be obtained by using a decoder. The vehicle speed is most convenient to detect, and the vehicle speed is used as an input vector, and different influence coefficients are set. Therefore, in one technical solution of the present application, first, vehicle speed values of a vehicle to be detected at a plurality of predetermined time points within a predetermined time period may be acquired by a vehicle speed sensor. However, because the wind resistance factor associated with the vehicle speed is an important relevant factor, a vehicle model with a relatively close vehicle appearance is selected as the vehicle to be measured during measurement, and the measurement accuracy is improved.

Specifically, in the operation process of the oil consumption metering system 300 based on data collaborative analysis, the detection data encoding module 360 is configured to pass the vehicle speed values at the plurality of predetermined time points through a time sequence encoder including a one-dimensional convolution layer to obtain a vehicle speed feature vector. That is, the detected vehicle speed value is taken as an input vector, and the obtained vehicle speed values at a plurality of predetermined time points are mapped into a high-dimensional feature space using a time-series encoder including a one-dimensional convolutional layer to obtain a vehicle speed feature vector.

Specifically, in the operation process of the oil consumption metering system 300 based on data covariance analysis, the vector query module 370 is configured to multiply the vehicle speed feature vector serving as a query feature vector by the covariance feature matrix to obtain a decoding feature vector. It should be understood that although the vehicle feature vector is used as the query vector, the matched decoded feature vector contains other parameter features and information affecting the oil consumption, so that the decoding of the decoded feature vector by using the decoder can obtain a more accurate decoded value representing the oil consumption. Particularly, in the technical solution of the present application, since each eigenvalue of the co-integration eigenvector obtained by performing co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector has a position attribute, for example, each position in the transfer matrix represents relationship information between corresponding positions of vectors, after matrix-multiplying the vehicle speed eigenvector as the query eigenvector with the co-integration eigenvector, each position of the obtained decoded eigenvector also has a position attribute.

Specifically, in the operation process of the oil consumption metering system 300 based on the data covariance analysis, the decoding optimization module 380 is configured to correct the eigenvalue of each position in the decoded eigenvector based on the reciprocal of the mean value of the eigenvalue of all positions of the decoded eigenvector to obtain the corrected decoded eigenvector. Since the decoding regression belongs to a substantial regression task without a position attribute when the decoding feature vector is subjected to decoding regression through a decoder, a decoding numerical deviation may be caused since position attribute information of the decoding feature vector is not considered. Based on this, in the technical solution of the present application, it is preferable that the vectors for performing phase sensing on the decoded feature vectors are aggregated by location, and are expressed as:

where V represents the decoded feature vector,

the inverse of the mean value of the eigenvalues of all the positions in the decoded eigenvector, indicates dot-by-dot multiplication. Here, the phase-aware characterization of the vector introduces a magnitude-phase quasi-real-imaginary characterization to stitch the vector by location for real-valued vectors based on the principle of Euler's formulaAnd expanding so as to compensate regression numerical value deviation possibly caused when the real-value regression task without the position attribute is carried out on the vector in a multi-layer perception mode. Thus, the accuracy of decoding regression is improved, i.e., a more accurate calculated value of oil consumption is obtained.

Specifically, in the operation process of the oil consumption metering system 300 based on data co-integration analysis, the oil consumption metering module 390 is configured to pass the corrected decoded feature vector through a decoder to obtain a decoded value, where the decoded value is an oil consumption value. In a specific example of the present application, the fuel consumption metering module is further configured to perform decoding regression on the corrected decoding feature vector by using the decoder according to the following formula to obtain the decoding value; wherein the formula is:

representing the matrix multiplication, h (-) is the activation function.

In summary, the oil consumption metering system 300 based on data collaborative analysis according to the embodiment of the present application is illustrated, a high-dimensional implicit nonlinear correlation between all relevant data items affecting oil consumption is mined by using a feature extractor based on a deep neural network model, meanwhile, a high-dimensional implicit correlation feature of an oil consumption sequence is extracted by using the feature extractor based on the deep neural network model, and then, collaborative analysis is performed on Gao Weite of the oil consumption sequence and a measurement parameter feature to mine a deep implicit correlation between a measurement parameter and the oil consumption, so as to construct an oil consumption metering scheme based on data collaborative analysis with higher measurement accuracy.

As described above, the fuel consumption metering system based on data collaborative analysis according to the embodiment of the present application can be implemented in various terminal devices. In one example, the fuel consumption metering system 300 based on data integration analysis according to the embodiment of the present application may be integrated into a terminal device as a software module and/or a hardware module. For example, the data-analysis-based fuel consumption metering system 300 may be a software module in the operating system of the terminal device, or may be an application developed for the terminal device; of course, the fuel consumption metering system 300 based on data covariance analysis may also be one of the hardware modules of the terminal device.

Alternatively, in another example, the data analysis-based fuel consumption metering system 300 and the terminal device may be separate devices, and the data analysis-based fuel consumption metering system 300 may be connected to the terminal device through a wired and/or wireless network and transmit the interaction information according to the agreed data format.

Exemplary method

Fig. 6 illustrates a flowchart of a fuel consumption metering method based on data covariance analysis according to an embodiment of the present application. As shown in fig. 6, the method for measuring oil consumption based on data collaborative analysis according to the embodiment of the present application includes the steps of: s110, obtaining a plurality of groups of vehicle detection data, wherein each group of vehicle detection data comprises vehicle speed, running time, wind resistance, total work, power, heat efficiency, heat value and oil consumption value; s120, arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in the plurality of groups of vehicle detection data into a measurement input matrix according to the data item dimension and the group dimension, and then obtaining a measurement characteristic vector through a convolutional neural network model serving as a filter; s130, arranging the oil consumption values in the plurality of groups of vehicle detection data into oil consumption input vectors, and then obtaining multi-scale neighborhood oil consumption characteristic vectors through a multi-scale neighborhood characteristic extraction module; s140, carrying out co-integration analysis on the measurement eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain a co-integration eigenvector matrix; s150, acquiring vehicle speed values of a vehicle to be detected at a plurality of preset time points within a preset time period; s160, passing the vehicle speed values of the plurality of preset time points through a time sequence encoder comprising a one-dimensional convolution layer to obtain a vehicle speed characteristic vector; s170, multiplying the vehicle speed feature vector serving as a query feature vector by the co-integration feature matrix to obtain a decoding feature vector; s180, correcting the feature values of all positions in the decoding feature vector based on the reciprocal of the mean value of the feature values of all the positions of the decoding feature vector to obtain a corrected decoding feature vector; and S190, enabling the corrected decoding characteristic vector to pass through a decoder to obtain a decoding value, wherein the decoding value is the oil consumption value.

In an example, in the above oil consumption metering method based on data covariance analysis, the step S120 includes: the row vector construction unit is used for arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in each group of vehicle detection data into row vectors so as to obtain a plurality of row vectors; the two-dimensional arrangement unit is used for performing two-dimensional arrangement on the plurality of row vectors to obtain the measurement input matrix; and a deep convolutional encoding unit for performing convolutional processing, feature matrix-based pooling processing, and nonlinear activation processing on input data in forward pass of layers using the layers of the convolutional neural network model as a filter, respectively, to output the measurement feature vector from a last layer of the convolutional neural network model as a filter, wherein an input of a first layer of the convolutional neural network model as a filter is the measurement input matrix.

In an example, in the above oil consumption metering method based on data covariance analysis, the step S130 includes: the first scale neighborhood convolution unit is used for inputting the oil consumption input vector into a first convolution layer of the multi-scale neighborhood feature extraction module to obtain a first scale oil consumption feature vector, wherein the first convolution layer has a first one-dimensional convolution kernel with a first length; the second scale neighborhood convolutional coding unit is used for inputting the fuel consumption input vector into a second convolutional layer of the multi-scale neighborhood feature extraction module to obtain a second scale fuel consumption feature vector, wherein the second convolutional layer has a second one-dimensional convolutional kernel with a second length, and the first length is different from the second length; and the cascade unit is used for cascading the first scale oil consumption characteristic vector and the second scale oil consumption characteristic vector to obtain the oil consumption characteristic vector.

Specifically, the fuel consumption input vector is input into a second convolution layer of the multi-scale neighborhood feature extraction module to obtain a second-scale fuel consumption feature vector, where the second convolution layer has a second one-dimensional convolution kernel with a second length, and the first length is different from the second length, and is further configured to: performing one-dimensional convolution coding on the oil consumption input vector by using a first convolution layer of the multi-scale neighborhood characteristic extraction module according to the following formula to obtain a first scale oil consumption characteristic vector;

wherein the formula is:

More specifically, the fuel consumption input vector is input into a second convolution layer of the multi-scale neighborhood feature extraction module to obtain a second-scale fuel consumption feature vector, where the second convolution layer has a second one-dimensional convolution kernel with a second length, and the first length is different from the second length, and is further configured to: performing one-dimensional convolution coding on the oil consumption input vector by using a second convolution layer of the multi-scale neighborhood characteristic extraction module according to the following formula to obtain a second scale oil consumption characteristic vector;

wherein the formula is:

In an example, in the above oil consumption metering method based on data covariance analysis, the step S140 is further configured to perform covariance analysis on the measurement eigenvector and the multi-scale neighborhood oil consumption eigenvector according to the following formula to obtain the covariance eigenvector matrix;

wherein the formula is:

wherein v is ₁ Is the multi-scale neighborhood fuel consumption eigenvector, v ₂ For the measured feature vector, M _c And integrating the feature matrix for the coordination.

In an example, in the above oil consumption metering method based on data synergistic analysis, the step S180 is further configured to: based on the reciprocal of the mean value of the feature values of all the positions of the decoding feature vector, correcting the feature values of all the positions in the decoding feature vector by the following formula to obtain the corrected decoding feature vector;

wherein the formula is:

where V represents the decoded feature vector,

In an example, in the oil consumption metering method based on data covariance analysis, the step S190 is further configured to perform decoding regression on the corrected decoded feature vector by using the decoder according to the following formula to obtain the decoded value;

wherein the formula is:

representing the matrix multiplication, h (-) is the activation function.

In summary, the oil consumption metering method based on the data collaborative analysis according to the embodiment of the present application is clarified, the high-dimensional implicit nonlinear correlation between all related data items affecting the oil consumption is mined by using the feature extractor based on the deep neural network model, meanwhile, the high-dimensional implicit correlation feature of the oil consumption sequence is extracted by using the feature extractor based on the deep neural network model, then the Gao Weite of the oil consumption sequence and the measurement parameter feature are collaboratively analyzed to mine the deep implicit correlation between the measurement parameter and the oil consumption, and based on this, the oil consumption metering scheme based on the data collaborative analysis with higher measurement accuracy is constructed.

Exemplary electronic device

Next, an electronic apparatus according to an embodiment of the present application is described with reference to fig. 7.

FIG. 7 illustrates a block diagram of an electronic device in accordance with an embodiment of the present application.

As shown in fig. 7, the electronic device 10 includes one or more processors 11 and memory 12.

The processor 11 may be a Central Processing Unit (CPU) or other form of processing unit having data processing capabilities and/or instruction execution capabilities, and may control other components in the electronic device 10 to perform desired functions.

Memory 12 may include one or more computer program products that may include various forms of computer-readable storage media, such as volatile memory and/or non-volatile memory. The volatile memory may include, for example, random Access Memory (RAM), cache memory (cache), and/or the like. The non-volatile memory may include, for example, read Only Memory (ROM), hard disk, flash memory, etc. One or more computer program instructions may be stored on the computer-readable storage medium and executed by the processor 11 to implement the functions of the fuel consumption metering system based on data synergy analysis of the various embodiments of the present application described above and/or other desired functions. Various content such as a co-integrated feature matrix may also be stored in the computer-readable storage medium.

In one example, the electronic device 10 may further include: an input device 13 and an output device 14, which are interconnected by a bus system and/or other form of connection mechanism (not shown).

The input device 13 may include, for example, a keyboard, a mouse, and the like.

The output device 14 can output various information including a decoded value and the like to the outside. The output devices 14 may include, for example, a display, speakers, a printer, and a communication network and its connected remote output devices, among others.

Of course, for simplicity, only some of the components of the electronic device 10 relevant to the present application are shown in fig. 7, and components such as buses, input/output interfaces, and the like are omitted. In addition, the electronic device 10 may include any other suitable components depending on the particular application.

Exemplary computer program product and computer-readable storage Medium

In addition to the above-described methods and apparatus, embodiments of the present application may also be a computer program product comprising computer program instructions that, when executed by a processor, cause the processor to perform the steps in the functions of the method for fuel consumption measurement based on data-covariance analysis according to various embodiments of the present application described in the above-mentioned "exemplary systems" section of this specification.

The computer program product may be written with program code for performing the operations of embodiments of the present application in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server.

Furthermore, embodiments of the present application may also be a computer-readable storage medium having stored thereon computer program instructions, which, when executed by a processor, cause the processor to perform the steps in the functions of the fuel consumption metering method based on data covariance analysis according to various embodiments of the present application, described in the "exemplary system" section above in this specification.

The computer-readable storage medium may take any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may include, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or a combination of any of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

The foregoing describes the general principles of the present application in conjunction with specific embodiments, however, it is noted that the advantages, effects, etc. mentioned in the present application are merely examples and are not limiting, and they should not be considered essential to the various embodiments of the present application. Furthermore, the foregoing disclosure of specific details is for the purpose of illustration and description and is not intended to be limiting, since the foregoing disclosure is not intended to be exhaustive or to limit the disclosure to the precise details disclosed.

The block diagrams of devices, apparatuses, systems referred to in this application are only given as illustrative examples and are not intended to require or imply that the connections, arrangements, configurations, etc. must be made in the manner shown in the block diagrams. These devices, apparatuses, devices, systems may be connected, arranged, configured in any manner, as will be appreciated by one skilled in the art. Words such as "including," "comprising," "having," and the like are open-ended words that mean "including, but not limited to," and are used interchangeably therewith. The words "or" and "as used herein mean, and are used interchangeably with, the word" and/or, "unless the context clearly dictates otherwise. The word "such as" is used herein to mean, and is used interchangeably with, the phrase "such as but not limited to".

It should also be noted that in the devices, apparatuses, and methods of the present application, the components or steps may be decomposed and/or recombined. These decompositions and/or recombinations are to be considered as equivalents of the present application.

The previous description of the disclosed aspects is provided to enable any person skilled in the art to make or use the present application. Various modifications to these aspects will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other aspects without departing from the scope of the application. Thus, the present application is not intended to be limited to the aspects shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

The foregoing description has been presented for purposes of illustration and description. Furthermore, the description is not intended to limit embodiments of the application to the form disclosed herein. While a number of example aspects and embodiments have been discussed above, those of skill in the art will recognize certain variations, modifications, alterations, additions and sub-combinations thereof.

Claims

1. A fuel consumption metering system based on data collaborative analysis is characterized by comprising:

the detection data acquisition module is used for acquiring vehicle speed values of a vehicle to be detected at a plurality of preset time points within a preset time period;

the vector query module is used for multiplying the vehicle speed characteristic vector serving as a query characteristic vector by the co-integration characteristic matrix to obtain a decoding characteristic vector;

2. The fuel consumption metering system based on data covariance analysis of claim 1, wherein the related data encoding module comprises:

the row vector construction unit is used for arranging the vehicle speed, the running time, the wind resistance, the total work, the power, the heat efficiency and the heat value in each group of vehicle detection data into row vectors so as to obtain a plurality of row vectors; and

a two-dimensional arrangement unit configured to perform two-dimensional arrangement on the plurality of row vectors to obtain the measurement input matrix; and

a deep convolutional coding unit, configured to perform convolutional processing, feature matrix-based pooling processing, and nonlinear activation processing on input data in forward pass of layers using the layers of the convolutional neural network model as the filter, respectively, to output the measurement feature vector from a last layer of the convolutional neural network model as the filter, where an input of a first layer of the convolutional neural network model as the filter is the measurement input matrix.

3. The fuel consumption metering system based on data covariance analysis of claim 2, wherein the fuel consumption data encoding module comprises:

the first scale neighborhood convolutional coding unit is used for inputting the oil consumption input vector into a first convolutional layer of the multi-scale neighborhood feature extraction module to obtain a first scale oil consumption feature vector, wherein the first convolutional layer has a first one-dimensional convolutional kernel with a first length;

a second scale neighborhood convolutional coding unit, configured to input the fuel consumption input vector into a second convolutional layer of the multi-scale neighborhood feature extraction module to obtain a second scale fuel consumption feature vector, where the second convolutional layer has a second one-dimensional convolutional kernel with a second length, and the first length is different from the second length; and

and the cascade unit is used for cascading the first scale oil consumption characteristic vector and the second scale oil consumption characteristic vector to obtain the oil consumption characteristic vector.

4. The fuel consumption metering system based on data covariance analysis of claim 3, wherein the first scale neighborhood convolutional encoding unit is further configured to: performing one-dimensional convolution coding on the oil consumption input vector by using a first convolution layer of the multi-scale neighborhood characteristic extraction module according to the following formula to obtain a first scale oil consumption characteristic vector;

wherein the formula is:

5. The fuel consumption metering system based on data covariance analysis of claim 4, wherein the second scale neighborhood convolution encoding unit is further configured to: performing one-dimensional convolution coding on the oil consumption input vector by using a second convolution layer of the multi-scale neighborhood characteristic extraction module according to the following formula to obtain a second scale oil consumption characteristic vector;

wherein the formula is:

6. The fuel consumption metering system based on data covariance analysis of claim 5, wherein the covariance analysis module is further configured to perform covariance analysis on the measured eigenvector and the multi-scale neighborhood fuel consumption eigenvector to obtain the covariance eigen matrix according to the following formula;

wherein the formula is:

7. The fuel consumption metering system based on data covariance analysis of claim 6, wherein the decoding optimization module is further configured to: based on the reciprocal of the mean value of the feature values of all the positions of the decoding feature vector, correcting the feature values of all the positions in the decoding feature vector by the following formula to obtain the corrected decoding feature vector;

wherein the formula is:

where V represents the decoded feature vector,

8. The fuel consumption metering system based on data collaborative analysis according to claim 7, wherein the fuel consumption metering module is further configured to use the decoder to perform decoding regression on the corrected decoded eigenvector by using the following formula to obtain the decoded value;

wherein the formula is:

representing the matrix multiplication, h (-) is the activation function.

9. A fuel consumption metering method based on data collaborative analysis is characterized by comprising the following steps:

arranging the oil consumption values in the plurality of groups of vehicle detection data into oil consumption input vectors, and then obtaining multi-scale neighborhood oil consumption characteristic vectors through a multi-scale neighborhood characteristic extraction module;

10. The fuel consumption metering method based on data covariance analysis according to claim 9, wherein the corrected decoded eigenvector is passed through a decoder to obtain a decoded value, and the decoded value is a fuel consumption value, and is further used for performing decoding regression on the corrected decoded eigenvector by using the decoder according to the following formula to obtain the decoded value;

wherein the formula is:

representing the matrix multiplication, h (-) is the activation function.