Diesel vehicle high emission identification and analysis method based on SCR technology

Abstract

The invention provides an SCR technology diesel vehicle high emission identification analysis method, which comprises the following steps: vehicle emission real-time test data is obtained, and the vehicle emission real-time test data comprises a combination of CO2 real-time emission data and NOX real-time emission data, or a combination of NOX real-time emission data and fuel consumption rate; performing correlation analysis on the vehicle emission real-time test data to obtain correlation results of the vehicle emission real-time test data under different working states of the SCR system; performing partition marking on the data points of the SCR system according to the correlation result; vehicle operation parameters are obtained, and a vehicle high emission analysis result is obtained through analysis according to the partition marking result and the vehicle operation parameters. By acquiring the real-time test data of vehicle emission and performing correlation analysis on the real-time test data of vehicle emission, high emission of pollutants of the diesel vehicle can be quickly identified, reasons for emission exceeding are searched in combination with the vehicle operation parameters, and efficient supervision of the high-emission diesel vehicle is supported.

Description

Technical Field

[0001] This application relates to the field of vehicle supervision, and particularly to a method for identifying and analyzing high emissions of SCR technology diesel vehicles. Background Art

[0002] The problem of mobile source pollution represented by motor vehicles has become increasingly prominent. In particular, nitrogen oxides (NO X ) account for 59.65% of the national total emissions and have become one of the most important local pollution sources of urban nitrogen dioxide (NO 2 ), fine particulate matter (PM2.5) and ozone (O 3 ). Diesel vehicles with a holding ratio of less than 10% account for 88.4% of the total vehicle NO X emissions. Therefore, strengthening the NO X emission control of diesel vehicles is a key task for continuously improving the air quality in China, and its pollution control has been listed as a landmark important battle in the "Blue Sky Defense War".

[0003] In order to reduce pollutant emissions, countries around the world have gradually tightened emission standards in recent years, prompting diesel vehicles to adopt more advanced emission control technologies (such as selective catalytic reduction, SCR). For example, the NO X emission limit of the national VI standard for heavy-duty diesel vehicles in China has been reduced by more than 90% compared with the national III standard. However, emission test studies in various countries have shown that the actual road NO X emission level of diesel vehicles is significantly higher than the requirements of laboratory type inspection limits, and the NO X emission of diesel vehicles has become one of the most difficult pollutants to control in motor vehicles. The main reasons for the ineffective actual road emission control of diesel vehicles include: X

[0004] (1) Emission fraud or cheating by manufacturers occurs from time to time; vehicle owners often add inferior urea (or even directly add water) or shield SCR components in order to save costs. These various emission fraud vehicles result in actual road pollutant emissions far higher than the standard limits;

[0005] (2) Enterprises often only design and optimize products according to the requirements of regulatory tests, resulting in poor NO X emission control of vehicles under actual complex working conditions and environmental conditions. For example, the ideal catalytic temperature range of SCR is 250 - 500 °C, and factors such as cold start, low temperature, and low load will all cause a significant increase in NO X emissions.

[0006] Therefore, strengthening the actual road NO X emission supervision of diesel vehicles will ensure that the tightened emission standards truly play their effectiveness and promote the NO X ​The top priority for substantial emissions reduction. With the frequent occurrences of cheating and falsification in diesel vehicle SCR systems and excessive emissions, how to achieve efficient identification and diagnostic root cause tracing of high-emission vehicles has become a key technical issue that urgently needs to be solved in mobile source emissions supervision. Summary of the Invention

[0007] To solve one of the above technical problems, the present invention provides a method for identifying and analyzing high emissions of diesel vehicles using SCR technology.

[0008] An embodiment of the present invention provides a method for identifying and analyzing high emissions of diesel vehicles using SCR technology, the method comprising:

[0009] Obtain real-time vehicle emission test data, the real-time vehicle emission test data including a combination of CO 2 real-time emission data and NO X real-time emission data, or a combination of NO X real-time emission data and fuel consumption rate;

[0010] Perform correlation analysis on the real-time vehicle emission test data to obtain a correlation result of the real-time vehicle emission test data under different working states of the SCR system;

[0011] Partition and label data points of the SCR system according to the correlation result;

[0012] Obtain vehicle operation parameters, and analyze and obtain a high-emission vehicle analysis result based on the result of the partition label and the vehicle operation parameters.

[0013] Preferably, the process of performing correlation analysis on the real-time vehicle emission test data to obtain a correlation result of the real-time vehicle emission test data under different working states of the SCR system includes:

[0014] Calculate the correlation coefficient of the real-time vehicle emission test data;

[0015] When the correlation coefficient is greater than a preset value, identify the working state of the diesel vehicle SCR system as a failure state.

[0016] Preferably, the process of partitioning and labeling data points of the SCR system according to the correlation result includes:

[0017] Establish a coordinate system according to the correlation result of the real-time vehicle emission test data, and organize the data points in the coordinate system into a data point set;

[0018] Respectively divide the data point set into a high-emission area and an emission-compliant area, and further divide the high-emission area into an SCR system non-working area and an SCR system abnormal area.

[0019] Preferably, the process of dividing the data point set into high-emission regions includes:

[0020] Perform a first normalization on the data point set to obtain a first normalized data point set;

[0021] Perform distribution statistics on the first normalized data point set to obtain a distribution set;

[0022] Determine the axis of symmetry of the high-emission region according to the distribution set, and perform a second normalization on the first normalized data point set to obtain a second normalized data point set;

[0023] Define a coordinate transformation function, classify the second normalized data point set based on a model to obtain an outlier set, obtain a positive example set and a negative example set of the second normalized data point set according to the outlier set, and estimate the boundary of the high-emission region according to the coordinate transformation function, the positive example set and the negative example set.

[0024] Preferably, the process of performing a first normalization on the data point set to obtain a first normalized data point set includes:

[0025] Obtain the maximum and minimum values of the abscissa and ordinate of the real-time emission data of each type in the data point set;

[0026] Perform normalization on the data points in the data point set to obtain a first normalized data point set.

[0027] Preferably, the process of performing distribution statistics on the first normalized data point set to obtain a distribution set includes:

[0028] Set a narrowband half-width and an elevation angle set, where the elevation angle set contains elevation angles covering all data points in the first normalized data point set;

[0029] Establish two range functions in the coordinate system according to the narrowband half-width and the elevation angle set;

[0030] Count the data points in the first normalized data point set located between the two range functions and divide by the number of all data points in the first normalized data point set to obtain the relative density index for each elevation angle;

[0031] Perform set sorting on the relative density index to obtain a distribution set.

[0032] Preferably, the process of determining the axis of symmetry of the high-emission region according to the distribution set and performing a second normalization on the first normalized data point set to obtain a second normalized data point set includes:

[0033] Set the angle between the ideal axis of symmetry of the high-emission region and the horizontal axis of the coordinate system;

[0034] Differentiate the said distribution set to obtain a difference set, and inversely search for the first non-zero value in the first normalized data point set at the position of the elevation angle corresponding to the minimum value in the difference set, and obtain the elevation angle corresponding to the first non-zero value;

[0035] If the elevation angle corresponding to the first non-zero value is less than the preset limit value, there is no high-emission area;

[0036] If the elevation angle corresponding to the first non-zero value is greater than or equal to the preset limit value, and the relative density index corresponding to the elevation angle corresponding to the first non-zero value is the maximum value in the distribution set, assign the elevation angle corresponding to the first non-zero value to the angle between the ideal axis of symmetry and the horizontal axis of the coordinate system;

[0037] If the elevation angle corresponding to the first non-zero value is greater than or equal to the preset limit value, and the relative density index corresponding to the elevation angle corresponding to the first non-zero value is not the maximum value in the distribution set, perform a second differentiation on the difference set to obtain a second-order difference set, start from the elevation angle corresponding to the first non-zero value, perform a forward search on the second-order difference set until the first positive value is found, obtain the elevation angle corresponding to the first positive value, obtain the maximum value between the relative density index corresponding to the elevation angle corresponding to the first positive value and the relative density index corresponding to the elevation angle corresponding to the first non-zero value in the difference set, record the elevation angle corresponding to the maximum value, and assign the elevation angle corresponding to the maximum value to the angle between the ideal axis of symmetry and the horizontal axis of the coordinate system;

[0038] Multiply the abscissas of all data points in the first normalized data point set by a normalization coefficient to obtain a second normalized data point set.

[0039] Preferably, the process of classifying the second normalized data point set based on a model to obtain an outlier set, obtaining a positive example set and a negative example set of the second normalized data point set according to the outlier set, and estimating the boundary of the high-emission area according to the coordinate conversion function, the positive example set and the negative example set includes:

[0040] Classify the second normalized data point set based on a model to obtain an outlier set;

[0041] Select the outliers located above the left of the line y = x in the coordinate system to obtain a negative example set;

[0042] Differentiate the second normalized data point set relative to the negative example set to obtain a positive example set;

[0043] Train the negative example set and the positive example set to obtain a training model;

[0044] Convert the data points into coordinate values through the coordinate conversion function, and input the coordinate values into the training model to output negative or positive examples.

[0045] Preferably, the process of dividing the data point set into an emission compliance area includes: in the data point set, dividing the area formed by the data points with the ordinate lower than the preset emission limit into the emission compliance area.

[0046] Preferably, the process of further dividing the high emission area into an SCR system non-operating area and an SCR system abnormal area includes:

[0047] In the data point set, swap the abscissa and ordinate of each data point located in the high emission area and then input it into the training model to obtain an output value;

[0048] The area where the output value is equal to 0 is the SCR system abnormal area, and the area where the output value is equal to 1 is the SCR system non-operating area.

[0049] The beneficial effects of the present invention are as follows: By obtaining the real-time vehicle emission test data and performing a correlation analysis on the real-time vehicle emission test data, the present invention can quickly identify the high emissions of diesel vehicle pollutants, and combine the vehicle operation parameters to find the reasons for emission exceedance, supporting the efficient supervision of in-use high-emission diesel vehicles. Description of the Drawings

[0050] The drawings described herein are used to provide a further understanding of the present application, and constitute a part of the present application. The schematic embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation of the present application. In the drawings:

[0051] Figure 1 It is a flowchart of a method for identifying and analyzing high emissions of SCR technology diesel vehicles according to Embodiment 1 of the present invention;

[0052] Figure 2 It is a correlation diagram of the CO 2 real-time emission concentration and NO X real-time emission concentration under different working states of the SCR system of different diesel vehicles according to Embodiment 1 of the present invention;

[0053] Figure 3 It is a correlation diagram of the CO 2 emission rate and NO X emission rate under different working states of the SCR system of different diesel vehicles according to Embodiment 1 of the present invention;

[0054] Figure 4 It is a correlation diagram of the fuel consumption rate and NO under different working states of the SCR system of different diesel vehicles according to Embodiment 1 of the present inventionX Schematic diagram of the correlation of emission rates;

[0055] Figure 5 Schematic diagram for dividing the working states of the SCR system described in Embodiment 1 of the present invention;

[0056] Figure 6 Schematic diagram of the diagnostic cause-tracing process for high pollutant emissions caused by a diesel vehicle with an SCR system described in Embodiment 1 of the present invention;

[0057] Figure 7 Schematic diagram of the principle of a high-emission identification and analysis system for SCR technology diesel vehicles described in Embodiment 2 of the present invention. Detailed implementation manners

[0058] In order to make the technical solutions and advantages in the embodiments of the present application clearer, the following further describes the exemplary embodiments of the present application in detail with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than an exhaustive list of all embodiments. It should be noted that, without conflict, the embodiments in the present application and the features in the embodiments can be combined with each other.

[0059] Embodiment 1

[0060] As Figure 1 shown, this embodiment proposes a method for identifying and analyzing high emissions of SCR technology diesel vehicles, and the method includes:

[0061] S101. Obtain real-time vehicle emission test data.

[0062] Specifically, for the diesel vehicle to be analyzed and processed, real-time vehicle emission test data can be collected. The real-time vehicle emission test data includes a combination of real-time CO 2 emission data and NO X emission data, or a combination of NO X emission data and fuel consumption rate. The method of collecting and obtaining the real-time vehicle emission test data can adopt two methods: on-road vehicle testing or remote online monitoring. Among them, the on-road vehicle testing method is to install a portable emission test system on the vehicle and collect the test data when the vehicle is driving on the actual road through the portable emission test system. The remote online monitoring method is to obtain the real-time vehicle test data in real time through remote OBD online monitoring data. The above two data collection methods can be selected according to the actual situation.

[0063] S102. Perform correlation analysis on the real-time vehicle emission test data to obtain the correlation results of the real-time vehicle emission test data in different working states of the SCR system.

[0064] Specifically, in this embodiment, the working state of the SCR system of a diesel vehicle is identified by calculating the correlation coefficient of the real-time vehicle emission test data. In this embodiment, the working state of the SCR system of a diesel vehicle is mainly identified through the correlation analysis of two types of data in the real-time vehicle emission test data. Among them, these two types of data are mainly CO 2 real-time emission data and NO X real-time emission data. This embodiment lists three methods for identifying the working state of the SCR system of a diesel vehicle through correlation analysis, specifically:

[0065] First, the working state of the SCR system of a diesel vehicle is identified through the correlation analysis of the CO 2 real-time emission concentration and the NO X real-time emission concentration.

[0066] First, the CO 2 real-time emission concentration (%) and the NO X real-time emission concentration (ppm) are obtained through means such as on-vehicle testing on actual roads or remote online monitoring. The CO 2 real-time emission concentration and the NO X real-time emission concentration are second-by-second transient data, with strong real-time performance. Then, the correlation coefficient (Pearson's r) of the CO 2 real-time emission concentration and the NO X real-time emission concentration is calculated. The CO 2 real-time emission concentration and the NO X real-time emission concentration are highly correlated. When the correlation coefficient r > 0.85, it can be judged that the SCR system of the diesel vehicle has completely failed or is not working. If the SCR system of the diesel vehicle is working properly, then the correlation between the CO 2 real-time emission concentration and the NO X real-time emission concentration is extremely weak, r < 0.5, as Figure 2 shown. Therefore, the working state of the SCR system of a diesel vehicle can be intuitively identified through the correlation coefficient of the CO 2 real-time emission concentration and the NO X real-time emission concentration. And when the correlation coefficient of the CO 2 real-time emission concentration and the NO X real-time emission concentration is greater than the preset value, the working state of the SCR system of the diesel vehicle is a failure state, and it may have completely failed or may not have been put into operation.

[0067] Second, the working state of the SCR system of a diesel vehicle is identified through the correlation analysis of the CO 2 emission rate and the NO X emission rate.

[0068] Due to CO 2 emission rate and NO X emission rate cannot be directly collected by sensors / test equipment. Therefore, before analyzing the correlation between CO 2 emission rate and NO X emission rate, it is first necessary to calculate the CO 2 emission rate and NO X emission rate. In this embodiment, the real-time emission concentrations (%) of CO 2 and NO X in ppm are obtained through on-vehicle tests on actual roads, and based on the real-time emission concentrations of CO 2 and NO X respectively, the CO 2 emission rate (g / s) and NO X emission rate (g / s) are obtained. The calculation process is as follows:

[0069]

[0070] Where ER i is the emission rate of pollutant i, with the unit of mg / s; C i is the real-time emission concentration of pollutant i, with the unit of ppm; M i is the molar mass of pollutant i, with the unit of g / mol; V 0 is the molar volume of an ideal gas under standard conditions, with the unit of L / mol; Q Exh. is the real-time exhaust gas flow rate, with the unit of kg / s; ρ Exh. is the exhaust gas density, with the unit of kg / m 3 . According to GB 17691-2018, the exhaust gas density of diesel vehicles is taken as 1.2943 kg / m 3 .

[0071] In the above formula, M i is a fixed value for a certain gas, and V 0 , Q Exh. and ρ Exh. have the same values for different pollutants i. Therefore, the correlation of ER i directly depends on C i . Based on this, the CO 2 emission rate and NO X emission rate can be obtained respectively from the real-time emission concentrations of CO 2 and NO X . The CO 2 emission rate and NO XThe emission rate is highly correlated, with a correlation coefficient r > 0.85, and it can be determined that the SCR system of the diesel vehicle has completely failed or is not working. If the SCR system of the diesel vehicle is working properly, then the CO 2 emission rate and NO X emission rate have extremely weak correlation, r < 0.5, as Figure 3 shown. Therefore, the working state of the SCR system of the diesel vehicle can be visually identified through the correlation coefficient of the CO 2 emission rate and NO X emission rate, and when the correlation coefficient of the CO 2 emission rate and NO X emission rate is greater than the preset value, the working state of the SCR system of the diesel vehicle is a failure state, and it may have completely failed or may not have been put into operation.

[0072] Thirdly, identify the working state of the SCR system of the diesel vehicle through the correlation analysis of the fuel consumption rate and the NO X emission rate.

[0073] The fuel consumption rate is directly related to the CO 2 emission rate. Therefore, the working state of the SCR system of the diesel vehicle can also be identified based on the correlation analysis of the fuel consumption rate (g / s) and the NO X emission rate (g / s). Among them, the NO X emission rate can be obtained by calculating the real-time emission concentration of NO X . The calculation process can be seen in detail in the second method above and will not be elaborated here. As Figure 4 shown, the fuel consumption rate and the NO X emission rate are highly correlated, with a correlation coefficient r > 0.85, and it can be determined that the SCR system of the diesel vehicle has completely failed or is not working. If the SCR system of the diesel vehicle is working properly, then the correlation between the fuel consumption rate and the NO X emission rate is extremely weak, r < 0.5. Therefore, the working state of the SCR system of the diesel vehicle can be visually identified through the correlation coefficient of the fuel consumption rate and the NO X emission rate, and when the correlation coefficient of the fuel consumption rate and the NO X emission rate is greater than the preset value, the working state of the SCR system of the diesel vehicle is a failure state, and it may have completely failed or may not have been put into operation.

[0074] In this embodiment, by obtaining the real-time vehicle emission test data and performing correlation analysis on multiple data in the real-time vehicle emission test data, it is possible to quickly identify whether the SCR system of the diesel vehicle is in a failure state, supporting the efficient supervision of in-use high-emission diesel vehicles.

[0075] S103. Mark the data points of the SCR system according to the correlation results.

[0076] Specifically, in this embodiment, taking Figure 2 the correlation diagram of the CO 2 real-time emission concentration and NO X real-time emission concentration under different working states of the SCR system shown in 2 as an example, the CO X real-time emission concentration is used as the horizontal axis (x-axis) of the coordinate system, and the NO X real-time emission concentration is used as the vertical axis (y-axis) of the coordinate system. In this coordinate system, the data points above the NO 2 real-time emission concentration limit can all be collectively referred to as high-emission data points. When the SCR system fails completely, the CO X real-time emission concentration and the NO X real-time emission concentration are highly correlated, approximately proportional, and with the fitting line as the boundary, the residual sets of the upper-side sample points and the lower-side sample points basically follow the same distribution with the same parameters. Therefore, the fitting line and the surrounding data points can be recorded as the SCR system complete failure data points. When the SCR system works normally, the NO X real-time emission concentration corresponding to the same coordinate point will decrease to varying degrees. Therefore, the data points between the lower right of the fitting line and above the NO

[0077] real-time emission concentration limit can be recorded as partially failed data points. Based on this, in this embodiment, the data points in the coordinate system are sorted out to obtain a data point set, and then the data point set is divided into a high-emission area and an emission compliance area respectively, and then the high-emission area is further divided into an SCR system non-working area and an SCR system abnormal area.

[0078] For dividing the data point set into a high-emission area, this embodiment uses the following method for division:

[0079] Perform a normalization process on the data point set once to obtain a first normalized data point set;

[0080] Perform a distribution statistics on the first normalized data point set to obtain a distribution set;

[0081] Determine the axis of symmetry of the high-emission area according to the distribution set, and perform a secondary normalization process on the first normalized data point set to obtain a second normalized data point set;

[0082] Specifically, still taking Figure 2CO under the working conditions of different diesel vehicle SCR systems shown 2 Real-time emission concentration and NO X Taking the correlation diagram of real-time emission concentration as an example, for CO 2 Real-time emission concentration is used as the horizontal axis (x-axis) of the coordinate system, and NO X Real-time emission concentration is used as the data point set D of the vertical axis (y-axis) of the coordinate system, and record CO 2 Real-time emission concentration and NO X The maximum and minimum values of real-time emission concentration are MAX_CO2, MIN_CO2, MAX_NOX, and MIN_NOX respectively, and the data points in the above data point set D are normalized based on MaxMinScaler to make the distribution of data points in the first quadrant balanced, and the first normalized data point set D 0 .

[0083] Based on the distribution characteristics of data points in the high-emission area, this embodiment uses a narrowband area corresponding to a specific elevation angle for data point distribution statistics. Specifically, first set the narrowband half-width b and the elevation angle r, and use the narrowband half-width b and the elevation angle r to establish two range functions in the coordinate system, namely the straight line y = tan(r / 180·π)·x - b and the straight line y = tan(r / 180·π)·x + b. Then for the first normalized data point set D 0 Count the data points falling on the above two straight lines, and divide the counting result by the number of all data points in the first normalized data point set D 0 to obtain the relative density index E at the elevation angle r. In the actual calculation process, a set of ascending-order elevation angle sets S 0 that can effectively cover all data points in the first normalized data point set D r can be specified, for example, S r = {0°, 1°, 2°, …, 89°}. Then calculate the relative density index corresponding to each elevation angle according to the above calculation process, and organize the relative density index to obtain the distribution set S E = {E 0 , E 1 , E 2 , …, E 89}.

[0084] For the ideal axis of symmetry of the data point set of the emission compliance area to be determined, this embodiment can set the angle between the ideal axis of symmetry and the horizontal axis as R. Take the difference of the distribution set S E to obtain the difference set S D , and record the elevation angle r D corresponding to the minimum value in the difference set S min . Among them, the difference set S DThe minimum value represents the first normalized data point set D 0 The position where the data point distribution changes most sharply from dense to sparse is used as the boundary for searching, which can avoid the interference of a small number of data points with relatively sparse distribution in the upper left of the first quadrant. min Search the position in reverse until the first non-zero value is found, and record the elevation angle r corresponding to the first non-zero value zero , used to represent the distribution set S E The stationary point that exists here.

[0085] At this time, it is necessary to zero For discussion, this embodiment proposes the following three situations.

[0086] The first one, if r zero If the value is less than the preset limit, it can be determined that the currently obtained symmetry axis coincides with the emission compliance area, so there is no high emission area. In actual calculations, the preset limit can be defined as arctan((T_NOX-MIN_NOX) / (MAX_NOX-MIN_NOX)), where T_NOX is the preset emission limit, that is, the specified NO X Real-time emission concentration.

[0087] The second type, if r zero Greater than the preset limit and in the distribution set S E The corresponding zero The relative density index of the distribution set S E The maximum value of R = r zero .

[0088] The third type, if r zero If neither of the above two situations applies, it means that the distribution peak of the data points in the high emission area may be covered by other data points, and further search is required. Specifically, take the difference set S D The difference of , we get the second-order difference set S DD For the second-order difference set S DD , from r zero Starting from position , search forward to the first positive value, and record the elevation angle r corresponding to the first positive value lower Get r lower to r zero Difference Set S D The maximum value of the maximum value, and record the corresponding elevation angle r localmax The elevation angle r localmax Yes lower to r zero The first normalized data point set D in this interval 0 The position where the data point distribution changes from dense to sparse most steadily is the position most likely to coincide with the covered peak. In this case, let R = rlocalmax 。

[0089] Through the above three cases, the axis of symmetry of the high-emission area can be determined. Then, multiply the abscissa of all data points in the first normalized data point set D 0 by the normalization coefficient M = tan(R / 180·π) to obtain the refined second normalized data point set D 1 。

[0090] After obtaining the axis of symmetry of the high-emission area, define the coordinate transformation function:

[0091] CO2(x) = M·(x - MIN_CO2) / (MAX_CO2 - MIN_CO2),

[0092] NOX(y) = (y - MIN_NOX) / (MAX_NOX - MIN_NOX),

[0093] This coordinate transformation function can be used for the coordinate transformation of the input of the original data points.

[0094] After determining the axis of symmetry of the high-emission area, it is necessary to predict the boundary of the high-emission area to delimit the range of the high-emission area.

[0095] In this embodiment, first, the negative example labeling based on the outlier algorithm is used to distinguish the positive and negative example sets. Specifically, the outlier algorithm, such as one-class support vector machine, is used to classify and obtain the outlier point set of the second normalized data point set D 1 . Select the outlier points located above the left of the line y = x and label them as negative sample points to obtain the negative example set D n . The complement D n of this negative example set D p = D 1 - D n is the positive example set.

[0096] Then, the boundary prediction is performed according to this positive and negative example set. Specifically, select a classification algorithm, such as support vector machine classifier, and train according to this negative example set D n and positive example set D p to obtain the training model Model. This training model Model can receive the coordinate values of the data points calculated by the coordinate transformation function, and input the coordinate values into this training model Model. The negative example outputs 0, and the positive example outputs 1.

[0097] For dividing the data point set into an emission compliance area, in this embodiment, the area formed by the data points with the ordinate lower than the preset emission limit value T_NOX in the data point set is divided into the emission compliance area.

[0098] To further distinguish the emission non-compliance caused by the SCR system anomaly and the SCR system not working. In this embodiment, in the high emission area, the SCR system anomaly area and the SCR system not working area are further distinguished, as Figure 5 shown. Specifically, in the data point set, for each data point (x j , y j ), first swap its abscissa and ordinate, and then input it into the training model Model for training to obtain the output value predict = Model(NOX(y j ), CO2(x j )). Among them, swapping the abscissa and ordinate makes use of the same distribution characteristic of the residual and the symmetric characteristic after the fitting line coefficient is normalized. When the output value predict of the training model Model is 0, it means that this data point is caused by the SCR system anomaly or the poor working efficiency of the SCR system. If the output value predict = 1, it means that this data point is caused by the SCR system not working. Accordingly, the area where the data points with the output value predict = 0 are located is the SCR system anomaly area or the area with poor SCR system working efficiency, and the area where the data points with the output value predict = 1 are located is the area where the SCR system does not work.

[0099] S104. Obtain the vehicle operation parameters, and parse the vehicle high emission analysis result according to the result of the partition marking and the vehicle operation parameters.

[0100] Specifically, based on the data points in the high emission area, by coupling the corresponding vehicle operation parameters, including vehicle instantaneous operation parameters, after-treatment status or environmental condition information, etc., the main reasons for the high pollutant emissions of diesel vehicles with SCR systems can be parsed.

[0101] As Figure 6 shown are some possible reasons for the high vehicle pollutant emissions proposed in this embodiment, including but not limited to the high correlation between the CO 2 or instantaneous fuel consumption data and the NO X instantaneous data caused by the complete failure of the SCR system, the too high SCR system inlet temperature caused by the deterioration of the SCR system catalyst, the low engine coolant temperature caused by the vehicle being in the cold start mode, the vehicle being in the low speed and low load state caused by adverse working conditions, the abnormal vehicle operation caused by extreme environments, etc. These factors are all possible reasons for the high vehicle pollutant emissions. This embodiment can parse and check the above reasons in turn.

[0102] In some optional embodiments, after obtaining the real-time vehicle emission test data, the SCR technology diesel vehicle high-emission identification and analysis method further includes: a step of aligning the time axis of the real-time vehicle emission test data to uniformly align different time series data on the time axis for further data analysis and processing. The embodiment does not make special limitations on the way of time axis alignment.

[0103] Embodiment 2

[0104] Corresponding to Embodiment 1, as Figure 7 shown, this embodiment proposes an SCR technology diesel vehicle high-emission identification and analysis system, which includes:

[0105] A data acquisition module for obtaining real-time vehicle emission test data, where the real-time vehicle emission test data includes a combination of CO 2 real-time emission data and NO X real-time emission data, or a combination of NO X real-time emission data and fuel consumption rate;

[0106] A correlation analysis module for performing a correlation analysis on the real-time vehicle emission test data to obtain the correlation results of the real-time vehicle emission test data in different working states of the SCR system;

[0107] A partition marking module for partitioning and marking the data points of the SCR system according to the correlation results;

[0108] A cause analysis module for obtaining vehicle operation parameters and parsing to obtain the vehicle high-emission analysis results according to the results of the partition marking and the vehicle operation parameters.

[0109] The working principle and implementation process of the SCR technology diesel vehicle high-emission identification and analysis system proposed in this embodiment can refer to the content recorded in Embodiment 1, and will not be elaborated in this embodiment. By obtaining the real-time vehicle emission test data and performing a correlation analysis on the real-time vehicle emission test data, this embodiment can quickly identify the high emissions of diesel vehicle pollutants and combine the vehicle operation parameters to find the reasons for excessive emissions, supporting the efficient supervision of in-use high-emission diesel vehicles.

[0110] Embodiment 3

[0111] This embodiment proposes a computer storage medium, including computer instructions. When the computer instructions run on an electronic device, the electronic device executes the following method:

[0112] Obtain real-time vehicle emission test data, where the real-time vehicle emission test data includes CO 2 real-time emission data and NO XA combination of real-time emission data, or NO X A combination of real-time emission data and fuel consumption rate;

[0113] Performing a correlation analysis on the real-time vehicle emission test data to obtain the correlation results of the real-time vehicle emission test data under different operating states of the SCR system;

[0114] Partitioning and marking the data points of the SCR system according to the correlation results;

[0115] Obtaining vehicle operating parameters, and parsing according to the results of the partitioning and marking and the vehicle operating parameters to obtain the vehicle high-emission analysis results.

[0116] Obviously, those skilled in the art can make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalent technologies, this application is also intended to include these changes and variations.

Claims

1. A method for identifying and analyzing high emissions from SCR technology diesel vehicles, characterized in that: The method comprises: Acquire vehicle emission real-time test data, including CO2 real-time emission data and NO X A combination of real-time emissions data, or NO X A combination of real-time emissions data and fuel consumption rates; Performing correlation analysis on the vehicle emission real-time test data to obtain correlation results of the vehicle emission real-time test data under different working conditions of the SCR system; Partitioning and marking the data points of the SCR system according to the correlation results; The vehicle operating parameters are obtained, and the vehicle high emission analysis results are obtained according to the partition marking results and the vehicle operating parameters analysis.

2. The method according to claim 1, characterized in that The process of performing correlation analysis on the vehicle emission real-time test data to obtain correlation results of the vehicle emission real-time test data under different working states of the SCR system includes: Calculate the correlation coefficient of vehicle emission real-time test data; When the correlation coefficient is greater than a preset value, it is identified that the working state of the SCR system of the diesel vehicle is a failure state.

3. The method according to claim 1, characterized in that: The process of marking the data points of the SCR system according to the correlation results comprises: Establishing a coordinate system according to the correlation results of the vehicle emission real-time test data, and arranging the data points in the coordinate system into a data point set; The data point set is divided into a high emission area and an emission standard area, and the high emission area is further divided into an SCR system non-working area and an SCR system abnormal area.

4. The method according to claim 3, characterized in that The process of dividing the data point set into high emission areas includes: Perform a normalization process on the data point set to obtain a first normalized data point set; Performing distribution statistics on the first normalized data point set to obtain a distribution set; Determining a symmetry axis of a high emission area according to the distribution set, and performing secondary normalization processing on the first normalized data point set to obtain a second normalized data point set; A coordinate transformation function is defined, and the second normalized data point set is classified based on the model to obtain an outlier point set, and a positive example set and a negative example set of the second normalized data point set are obtained according to the outlier point set, and a boundary of the high emission area is estimated according to the coordinate transformation function, the positive example set and the negative example set.

5. The method according to claim 4, characterized in that The process of performing a normalization process on the data point set to obtain a first normalized data point set includes: Obtain the maximum and minimum values ​​of the horizontal and vertical coordinates of each type of real-time emission data in the data point set; The data points in the data point set are normalized to obtain a first normalized data point set.

6. The method according to claim 5, characterized in that The process of performing distribution statistics on the first normalized data point set to obtain a distribution set includes: Setting a narrowband half-width and an elevation angle set, wherein the elevation angle set includes elevation angles covering all data points in the first normalized data point set; Establishing two range functions in the coordinate system according to the narrowband half-width and elevation angle set; Counting the data points between the two range functions in the first normalized data point set and dividing the count by the number of all data points in the first normalized data point set to obtain a relative density index for each elevation angle; The relative density indexes are collected and sorted to obtain a distribution set.

7. The method according to claim 6, characterized in that The process of determining the symmetry axis of the high emission area according to the distribution set and performing secondary normalization processing on the first normalized data point set to obtain the second normalized data point set includes: Set the angle between the ideal symmetry axis of the high emission area and the horizontal axis of the coordinate system; Differentiating the distribution set to obtain a differential set, reversely searching for the first non-zero value in the first normalized data point set at the position of the elevation angle corresponding to the minimum value in the differential set, and obtaining the elevation angle corresponding to the first non-zero value; If the elevation angle corresponding to the first non-zero value is less than a preset limit, there is no high emission area; If the elevation angle corresponding to the first non-zero value is greater than or equal to a preset limit value, and the relative density index corresponding to the elevation angle corresponding to the first non-zero value is the maximum value in the distribution concentration, then the elevation angle corresponding to the first non-zero value is assigned to the angle between the ideal symmetry axis and the horizontal axis of the coordinate system; If the elevation angle corresponding to the first non-zero value is greater than or equal to the preset limit value, and the relative density index corresponding to the elevation angle corresponding to the first non-zero value is not the maximum value in the distribution set, then perform a secondary difference on the difference set to obtain a second-order difference set, start from the elevation angle corresponding to the first non-zero value, perform a forward search on the second-order difference set until the first positive value is found, obtain the elevation angle corresponding to the first positive value, obtain the maximum value between the relative density index corresponding to the elevation angle corresponding to the first positive value and the relative density index corresponding to the elevation angle corresponding to the first non-zero value in the difference set, record the elevation angle corresponding to the maximum value, and assign the elevation angle corresponding to the maximum value to the angle between the ideal symmetry axis and the horizontal axis of the coordinate system; The horizontal coordinates of all data points in the first normalized data point set are multiplied by a normalization coefficient to obtain a second normalized data point set.

8. The method according to claim 7, characterized in that The process of classifying the second normalized data point set based on the model to obtain an outlier point set, obtaining a positive example set and a negative example set of the second normalized data point set according to the outlier point set, and estimating the boundary of the high emission area according to the coordinate conversion function, the positive example set and the negative example set includes: Classifying the second normalized data point set based on the model to obtain an outlier point set; Select the outlier point located at the upper left of the y=x line in the coordinate system to obtain the negative example set; Differentiating the second normalized data point set from the negative example set to obtain a positive example set; Training the negative example set and the positive example set to obtain a training model; The data points are converted into coordinate values ​​by the coordinate conversion function, the coordinate values ​​are input into the training model and a negative example or a positive example is output.

9. The method according to claim 8, wherein the process of dividing the data point set into emission compliance areas comprises: In the data point set, the area formed by the data points whose ordinates are lower than the preset emission limit value is divided into the emission standard-reaching area.

10. The method according to claim 9, characterized in that The process of further dividing the high emission area into an SCR system non-operating area and an SCR system abnormal area includes: In the data point set, the horizontal and vertical coordinates of each data point located in the high emission area are swapped and then input into the training model to obtain an output value; The area where the data points with the output value equal to 0 are located is the abnormal area of ​​the SCR system, and the area where the data points with the output value equal to 1 are located is the non-working area of ​​the SCR system.

Images