CN116992247B - Abnormal data detection method of tail gas analyzer - Google Patents
Abnormal data detection method of tail gas analyzer Download PDFInfo
- Publication number
- CN116992247B CN116992247B CN202311256758.7A CN202311256758A CN116992247B CN 116992247 B CN116992247 B CN 116992247B CN 202311256758 A CN202311256758 A CN 202311256758A CN 116992247 B CN116992247 B CN 116992247B
- Authority
- CN
- China
- Prior art keywords
- data
- gas
- tail gas
- historical
- detected
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 65
- 238000001514 detection method Methods 0.000 title claims abstract description 11
- 239000007789 gas Substances 0.000 claims abstract description 395
- 230000005856 abnormality Effects 0.000 claims abstract description 44
- 238000000034 method Methods 0.000 claims abstract description 44
- 238000013528 artificial neural network Methods 0.000 description 11
- 239000013598 vector Substances 0.000 description 8
- 238000004458 analytical method Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 5
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Chemical compound O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 4
- IJGRMHOSHXDMSA-UHFFFAOYSA-N Atomic nitrogen Chemical compound N#N IJGRMHOSHXDMSA-UHFFFAOYSA-N 0.000 description 2
- CURLTUGMZLYLDI-UHFFFAOYSA-N Carbon dioxide Chemical compound O=C=O CURLTUGMZLYLDI-UHFFFAOYSA-N 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004868 gas analysis Methods 0.000 description 2
- 150000002430 hydrocarbons Chemical class 0.000 description 2
- 239000004215 Carbon black (E152) Substances 0.000 description 1
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 229910002092 carbon dioxide Inorganic materials 0.000 description 1
- 239000001569 carbon dioxide Substances 0.000 description 1
- 229910002091 carbon monoxide Inorganic materials 0.000 description 1
- 238000007621 cluster analysis Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000000446 fuel Substances 0.000 description 1
- -1 hydrocarbon Chemical class 0.000 description 1
- 229930195733 hydrocarbon Natural products 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229910052757 nitrogen Inorganic materials 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N33/00—Investigating or analysing materials by specific methods not covered by groups G01N1/00 - G01N31/00
- G01N33/0004—Gaseous mixtures, e.g. polluted air
- G01N33/0009—General constructional details of gas analysers, e.g. portable test equipment
- G01N33/0062—General constructional details of gas analysers, e.g. portable test equipment concerning the measuring method or the display, e.g. intermittent measurement or digital display
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Life Sciences & Earth Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Chemical & Material Sciences (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Food Science & Technology (AREA)
- Combustion & Propulsion (AREA)
- Medicinal Chemistry (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Testing Of Engines (AREA)
Abstract
The invention relates to the field of data processing, in particular to an abnormal data detection method of a tail gas analyzer, which comprises the steps of obtaining tail gas data and corresponding engine data, and obtaining abnormal indexes of each gas in the tail gas data to be detected according to the content ratio of each gas to other gases in the tail gas data to be detected and the difference of the average content ratio of each gas to other gases in historical tail gas data; clustering historical engine data, obtaining the influence degree of each engine data on each gas according to the difference of each engine data in each type of historical engine data and the difference of each gas in corresponding historical tail gas data, calculating the predicted value of each gas content according to the influence degree, obtaining the abnormality degree of the tail gas data to be detected according to the difference of the predicted value and the actual value and the abnormality index of each gas, judging whether the data of the tail gas analyzer to be detected is abnormal according to the abnormality degree, and realizing intelligent and accurate method.
Description
Technical Field
The application relates to the field of data processing, in particular to an abnormal data detection method of a tail gas analyzer.
Background
The tail gas discharged by motor vehicles such as automobiles contains complex gas components, wherein the complex gas contains harmless gases such as carbon dioxide, nitrogen, water vapor and the like, and hydrocarbon compounds such as hydrocarbon, oxynitride and carbon monoxide and the like, which are harmful gases, and if the content of the harmful gases in the discharged tail gas is high, the environment pollution is caused, so that the gas in the tail gas needs to be accurately analyzed, and whether the tail gas discharge is qualified or not is judged according to the analysis result;
the existing method for analyzing the gas in the tail gas comprises the steps of acquiring tail gas emission data to be detected by using a tail gas analyzer by a worker with abundant experience within a certain period of starting an automobile, and judging whether the tail gas is abnormal according to the data, wherein the data acquired by the tail gas analyzer are influenced by environmental factors, such as the influence of an engine state on the data, so that errors exist in the data acquired by the tail gas analyzer, and the accuracy of tail gas analysis is influenced.
Disclosure of Invention
Aiming at the problems that data acquired by an exhaust gas analyzer are influenced by the state of an engine, so that errors exist in the data acquired by the exhaust gas analyzer and the accuracy of exhaust gas analysis is influenced, the invention provides an abnormal data detection method of the exhaust gas analyzer, which comprises the following steps:
acquiring current engine data and corresponding tail gas data to be detected of an automobile, and acquiring historical engine data and corresponding historical tail gas data of the automobile;
obtaining abnormal indexes of each gas in the tail gas data to be detected according to the content ratio of each gas to other gases in the tail gas data to be detected and the difference between the average content of each gas in the historical tail gas data and the average content of other gases;
clustering the historical engine data, and obtaining the influence degree of each engine data on each gas according to the difference of each engine data in each type of historical engine data and the difference of each gas in the historical tail gas data corresponding to the type of historical engine data;
obtaining target engine data according to the distance between the current engine data and the historical engine data;
according to the influence degree of each engine data in the target engine data on each gas and the content of each gas in the tail gas data corresponding to the target engine data, obtaining a predicted value of the content of each gas in the tail gas data to be detected;
obtaining the abnormality degree of the tail gas data to be detected according to the difference between the predicted value and the actual value of each gas content in the tail gas data to be detected and the abnormality index of each gas;
judging whether the tail gas data to be detected is abnormal or not according to the abnormality degree of the tail gas data to be detected.
The method for acquiring the abnormality degree of the tail gas data to be detected comprises the following steps:
and accumulating the difference value of the predicted value and the actual value of each gas content in the tail gas data to be detected, and taking the product of the accumulated value and the average value of the abnormal indexes of all the gases as the abnormal degree of the tail gas data to be detected.
The method for acquiring the abnormal index of each gas in the tail gas data to be detected comprises the following steps:in (1) the->For the gas in the tail gas data to be detected +.>Is used for the abnormality index of (1),for the gas in the tail gas data to be detected +.>Content of (A),>for gas in historical tail gas data->Content mean value of>Is->Seed gas (s)/(s)>Not equal to->，/>For the +.>Content of seed gas,/->For +.>Average value of seed gas content,/->Is the total number of gas species in the historical tail gas data.
The calculation method of the predicted value of each gas content in the tail gas data to be detected comprises the following steps:
in (1) the->Is a gas->Predicted value of content>For the number of historical engine data contained in the target engine data, +.>Is->Historical engine data>For the total number of categories of engine data +.>Is->Engine data->Is->Engine data versus gas->Is used for controlling the degree of influence of (a),is->Gas +.>Is contained in the composition.
The method for obtaining the target engine data according to the distance between the current engine data and the historical engine data comprises the following steps:
acquiring a distance between current engine data and each historical engine data:
clustering is carried out based on the distance between the current engine data and each historical engine data, so that a plurality of categories are obtained, and all engine data in the category where the current engine is located are used as target engine data.
And the distance between the acquired current engine data and each historical engine data is the Euclidean distance between the current engine data and each historical engine data.
The method for acquiring the influence degree of each engine data on each gas comprises the following steps:in (1) the->Is->Engine data versus gas->Is used for controlling the degree of influence of (a),total category number for clustering historical engine data,/-for the historical engine data>Is->Category (S),>is->Gas +.>Content mean value of>Is->No.>An average of engine data. The method for judging whether the exhaust data to be detected is abnormal or not according to the abnormality degree of the exhaust data to be detected comprises the following steps:
when the abnormality degree of the tail gas data to be detected is larger than the abnormality degree threshold value, the tail gas data to be detected is abnormal.
The beneficial effects of the invention are as follows:
according to the method, to-be-detected tail gas data and historical tail gas data acquired by a tail gas analyzer and corresponding engine data are acquired, and abnormal indexes of each gas in the to-be-detected tail gas data are obtained according to the content ratio of each gas to other gases in the to-be-detected tail gas data and the difference between the content average value of each gas in multiple groups of historical tail gas data and the content ratio of other gases; according to the method, the gas content in the currently collected tail gas data to be detected is compared with the gas content in the historical tail gas data, and the smaller the difference is, the more normal the gas content in the currently collected tail gas data to be detected can be described, so that the gas content is used as an abnormal index of gas, and whether the gas content is abnormal or not is judged preliminarily; according to the method, historical engine data are clustered, and the influence degree of each engine data on each gas is obtained according to the difference of each engine data in each category after clustering and the difference of each tail gas in the tail gas data corresponding to each category; according to the method, each engine data is taken as a target, other engine data are divided into a plurality of categories, other engine data in each category are approximate, and then under the condition that other engine data are approximate, the difference of the targets in each category and the content difference of each gas in each category are compared to obtain the influence degree of each engine data on each gas; the method comprises the steps of obtaining target engine data, obtaining a content predicted value of each gas in exhaust data to be detected according to the influence degree of each engine data on each gas in the target engine data and the content of each gas in the exhaust data corresponding to each engine data, obtaining the abnormality degree of the exhaust data according to the predicted value, the actual content difference and the abnormality index of each gas, and judging whether the exhaust data is abnormal according to the abnormality degree; according to the method, the current engine data is predicted through data similar to the current engine data in the historical engine data, the influence degree of the engine data on the gas is considered, an accurate prediction result can be obtained, the influence of the engine data on the tail gas is combined with the abnormal index of each gas in the tail gas, the abnormal degree of the tail gas data is comprehensively analyzed, and the detection accuracy is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions of the prior art, the drawings which are used in the description of the embodiments or the prior art will be briefly described, it being obvious that the drawings in the description below are only some embodiments of the invention, and that other drawings can be obtained according to these drawings without inventive faculty for a person skilled in the art.
Fig. 1 is a flow chart of an abnormal data detection method of an exhaust gas analyzer according to the present invention.
Description of the embodiments
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
An embodiment of an abnormal data detection method of an exhaust gas analyzer of the present invention, as shown in fig. 1, includes:
step one: acquiring current engine data and corresponding tail gas data to be detected of an automobile, and acquiring historical engine data and corresponding historical tail gas data of the automobile;
the method comprises the steps of acquiring current tail gas data, engine data corresponding to the current tail gas data, historical tail gas data and engine data corresponding to the historical tail gas data by using a tail gas analyzer, and taking the data as a data basis.
The tail gas data to be detected are acquired by using a tail gas analyzer, and the gas in the tail gas data comprises harmful gas, harmless gas and data abnormal state vectors;
the harmful gases are A in a collection way, and comprise a type of harmful gases such as A1, A2, …, aa and the like, wherein A1 is the 1 st harmful gas, A2 is the 2 nd harmful gas, and Aa is the a-th harmful gas;
the harmless gases are collected as B, and comprise B harmless gases such as B1, B2, …, bb and the like, wherein B1 is the 1 st harmless gas, B2 is the 2 nd harmless gas, and Bb is the B nd harmless gas;
a data abnormal state vector, denoted by Z, when z= {0}, i.e., the data abnormal state vector contains only 0, indicating that the tail gas data is normal data; when the first vector value of the data abnormal state vector is-1, the exhaust data is indicated to be abnormal data, and the abnormal reasons corresponding to the abnormal data are marked in the abnormal data according to preset marks, for example, the data abnormal state vector Z= { -1,1,3,7}, the first vector value is-1, namely, the exhaust data is indicated to be abnormal, and the corresponding abnormal reasons are data anomalies caused by the first, third and seventh reasons.
When the exhaust gas analyzer is used to collect data, the state data of the corresponding engine, such as air flow, water temperature, rotation speed, air inlet temperature, fuel pressure, air inlet pressure, etc., are data information which can affect the exhaust gas components, the obtained engine data is represented by a set E, denoted as E { E1, E2, …, eτ }, E1 is first engine data, i.e. air flow, E1 is second engine data, i.e. water temperature, eτ is τ -th engine data, and total τ engine state data, i.e. τ engine data which affects the exhaust gas components, the specific types of the engine data can be set by an operator, each obtained exhaust gas data contains multiple gases, and each engine data is composed of the above different types of engine data.
Step two: obtaining abnormal indexes of each gas in the tail gas data to be detected according to the content ratio of each gas to other gases in the tail gas data to be detected and the difference between the average content of each gas in the historical tail gas data and the average content of other gases;
the method comprises the steps of selecting normal tail gas data in historical tail gas data, comparing the currently acquired tail gas data with the historical tail gas data, and judging whether the current tail gas data is abnormal data or not according to comparison differences;
in this embodiment, the normal historical tail gas data, that is, a plurality of historical tail gas data with a data abnormal state vector of 0, in the historical tail gas data is obtained first, and subsequent analysis is performed based on the normal historical tail gas data;
the method for acquiring the abnormal index of each gas in the tail gas data to be detected comprises the following steps:
for the a-th gas (gas a) in the tail gas data to be detected, the abnormal index calculation method comprises the following steps:in (1) the->To be treatedDetecting gas +.>Is used for the abnormality index of (1),for the gas in the tail gas data to be detected +.>Content of (A),>for gas in historical tail gas data->Content mean value of>Is->Seed gas (s)/(s)>Not equal to->，/>For the +.>Content of seed gas,/->For +.>Average value of seed gas content,/->The total number of gas types in the historical tail gas data;
the formula is for the object to be treatedDetecting and analyzing the abnormality of each gas in the exhaust gas data, and taking the gas as a target gas, for example, taking the gas in a as the target gas,the difference of the average value of the content of the target gas in the tail gas data to be detected and the content of the target gas in the historical tail gas data is obtained;the method comprises the steps of judging the abnormality of the content of target gas according to the difference between the ratio of the content of the target gas in the current tail gas data to be detected and the ratio of the average content of the target gas in the historical tail gas data to the average content of other gases, wherein the smaller the difference between the ratio of the content of the target gas to the content of the other gases and the average ratio of the content of the target gas to the content of the other gases when the historical tail gas data is normal is, the more normal the content of the target gas is indicated, setting an abnormality index threshold value of 0.4, namely, when the obtained abnormality index of the target gas is larger than 0.4, the target gas is considered to be abnormal and needs to be subjected to subsequent analysis, otherwise, the target gas is normal, and according to the step, taking each gas as the target gas, the abnormality index of each gas can be obtained, and when at least one gas is abnormal in the tail gas data to be detected, the tail gas data is considered to be abnormal and needs to be subjected to subsequent treatment; otherwise, the tail gas data to be detected are considered to be normal.
The method comprises the steps of obtaining current engine data of an automobile and corresponding tail gas data to be detected, and obtaining the difference of the corresponding ratio of the content of each gas in historical engine data of the automobile to the content of other gases in the corresponding historical tail gas data and the corresponding ratio of the content of each gas to the corresponding ratio of the historical data, and the difference of the content of each gas to the corresponding content of the historical data to preliminarily judge whether the tail gas data to be detected is abnormal or not.
Step three: clustering the historical engine data, and obtaining the influence degree of each engine data on each gas according to the difference of each engine data in each type of historical engine data and the difference of each gas in the historical tail gas data corresponding to the type of historical engine data;
the method comprises the steps of carrying out cluster analysis on historical engine data corresponding to each piece of historical tail gas data, and further obtaining the influence degree of each piece of engine data on each piece of gas.
The method for acquiring the influence degree of each engine data on each gas comprises the following steps:
(1) Acquiring a plurality of historical engine data corresponding to the historical tail gas data, namely each E corresponding to each historical tail gas data;
(2) DBSCAN clustering is carried out based on Euclidean distance among the plurality of historical tail gas data, the plurality of historical tail gas data are clustered into a plurality of categories, and each category comprises a plurality of historical tail gas data;
it should be noted that, before clustering, one engine data needs to be selected from a plurality of historical engine data as target data; the Euclidean distance among the plurality of historical tail gas data is calculated when the Euclidean distance among other data except the target data is calculated, the reason for this is that, similar to a control variable method, the target data is ignored first to analyze the target data, clustering is carried out based on the distance among the other data except the target data, the other data in each class after clustering is similar, and the difference of the target data is analyzed under the condition that the other data are similar;
the following are illustrated:
for example, there is historical exhaust data E1 (e1=20, e2=10, e3=5), E2 (e1=20, e2=15, e3=10), where E1, E2, E3 represent three engine states of air flow, water temperature, rotational speed, respectively; if the target data is set to be E1, the Euclidean distance between the historical tail gas state data E1 and E2The method comprises the following steps:in this way, the Euclidean distance between the historical tail gas data is calculatedSeparating, and then clustering based on Euclidean distance to obtain a plurality of categories;
(3) The influence degree of each engine data on each gas is obtained, and the calculation method comprises the following steps:in (1) the->Is->Engine data versus gas->Degree of influence of->Is->Category (S),>total category number for clustering historical engine data,/-for the historical engine data>Is->Gas +.>Content mean value of>Is->No.>The average value of the engine data is used,/>is->Gas +.>Content mean value of>Is->No.>An average of seed engine data;
the formula is that in the engine data E { E1, E2, …, eτ } obtained in the first step, the x-th engine data x is selected to have a value smaller than or equal to τ, the x-th engine data is used as target data, the clustering in the step (2) is performed, the calculation is performed according to the obtained clustering result, the meaning of the j-1 th category means that other categories except the j-th category are used, the data in each category and the data of other categories are analyzed,the average value of target data in each type of similar historical engine data after clustering is +.>Target data mean +.>Reflecting the difference in target data in the case where other engine data are similar; />Is the j-th calendar after clusteringAverage value of gas a content in historical exhaust gas data corresponding to historical engine data, and other classes +.>The difference of the average value of the content of the gas a in the historical tail gas data corresponding to the engine data, namely the average value difference of each gas in the historical tail gas data corresponding to each similar engine data reflects the data difference of each gas; the smaller the ratio of the two is, the more the engine data affects the gas, and conversely, the smaller the engine data affects the gas.
In the method according to the present step, the degree of influence of each engine data on each gas can be calculated by using each engine data as target data.
Step four: obtaining target engine data according to the distance between the current engine data and the historical engine data; according to the influence degree of each engine data in the target engine data on each gas and the content of each gas in the tail gas data corresponding to the target engine data, obtaining a predicted value of the content of each gas in the tail gas data to be detected;
the method comprises the steps of clustering current engine data and historical engine data, analyzing the category of the current engine data to obtain historical tail gas data corresponding to the category of the current engine data, and predicting the influence degree of each gas in the historical tail gas data and the content of each gas according to each engine data in the category of the current engine data.
The method for obtaining the target engine data according to the distance between the current engine data and the historical engine data comprises the following steps:
(1) The method for acquiring the distance between the current engine data and each historical engine data, namely the Euclidean distance, comprises the following steps:
obtaining the difference value of each engine data in the current engine data and the corresponding engine data of the same category in each historical engine data, and accumulating the square of each difference value to obtain a value as the current engineThe distance between the data and each historical engine data is calculated by the following formula:in the formula>、、/>、/>1 st, 2 nd, 3 rd, and/or +/in the current engine data>Engine data->、/>、/>、/>1 st, 2 nd, 3 rd, and/or +.>And the engine data reflects the distance between the factors affecting the tail gas components and the data of a certain historical tail gas affecting factor, and when the required D is smaller, the automobile tail gas components corresponding to the current engine data are indicated, and the historical data tail gas components corresponding to the historical engine data corresponding to the current engine data are more similar.
(2) Clustering based on the distance between the current engine data and each historical engine data to obtain a plurality of categories, and taking all engine data in the category of the current engine as target engine data;
it should be noted that, the engine data in the category of the current engine data is selected as the target engine data, because the clustering is performed based on the distance between the current engine data and the historical engine data, the engine state data in the category of the current engine state is the most similar and close to the current engine data, so that the current engine data and the similar historical engine data are selected as the target engine data for subsequent analysis, the clustering algorithm uses a DBSCAN clustering algorithm, the clustering distance is D, and the tail gas component in the historical engine data possibly corresponding to the current engine data E is obtained through clustering;
the method for acquiring the predicted value of each gas content in the tail gas data to be detected comprises the following steps:in (1) the->Is a gas->Predicted value of content>For the number of historical engine data contained in the target engine data, +.>Is->Historical engine data>For the total number of categories of engine data +.>Is->Hair-growing deviceMotivation data->Is->Engine data versus gas->Degree of influence of->Is->Gas +.>Is contained in the composition;
in the formula, by acquiring historical engine data similar to the current engine data and historical tail gas data corresponding to the similar historical engine data, according to the influence degree of each engine data on each gas in the historical engine dataAnd the content of each gas +.>Dividing by->The average content of each gas in the historical engine data similar to the current engine data is obtained, the influence of the engine on the tail gas data is analyzed, and the predicted value of each gas in the current tail gas data to be detected can be accurately estimated.
It should be noted that, the predicted value of each gas in the current exhaust data to be detected may also be obtained by combining with a neural network:
acquiring current engine data and corresponding tail gas data to be detected of an automobile, and acquiring historical engine data of the automobile and corresponding historical engine data corresponding to the historical tail gas data, wherein m engine data are used in total, namely m engine data capable of affecting tail gas components; inputting the acquired m engine data into a neural network, and predicting tail gas components corresponding to the m engine data by using the neural network to obtain a content predicted value of each gas of the tail gas data corresponding to the current engine data;
the method comprises the following steps:
the neural network structure uses a ResNet neural network, uses a large amount of historical tail gas data and a corresponding large amount of historical engine state data, trains the neural network, takes each historical engine state data as input, and outputs the corresponding historical tail gas data;
the neural network has a loss function of：
In the formula (i),representing the mean square error loss function, ">The difference between the predicted value and the actual value of the gas content in the obtained tail gas data is represented, and the calculation method comprises the following steps:
in the formula>Representing the predicted content value of the i-th gas in the exhaust gas data to be detected obtained by the above analysis method,/for>Indicating the actual content value of the gas,when ask +.>The smaller the neural network is, the better the neural network effect is correspondingly obtained, and when the loss function converges, the neural network training is completed;
and inputting the current engine data corresponding to the tail gas data to be detected into a neural network to obtain the predicted content value of each gas in the tail gas data to be detected.
It should be noted that, according to the method of this step, a predicted value of each gas content in the exhaust gas to be detected may be obtained.
Step five: obtaining the abnormality degree of the tail gas data to be detected according to the difference between the predicted value and the actual value of each gas content in the tail gas data to be detected and the abnormality index of each gas; judging whether the tail gas data to be detected is abnormal or not according to the abnormality degree of the tail gas data to be detected;
the method comprises the steps of calculating the abnormality degree of the tail gas data to be detected according to the difference between the actual content of each gas in the tail gas data to be detected, which is acquired by a current tail gas analyzer, and the predicted content of the gas, and the abnormality index of each gas;
the method for acquiring the abnormality degree of the tail gas data to be detected comprises the following steps:
accumulating the difference between the predicted value and the actual value of each gas content in the tail gas data to be detected, and adding the accumulated valueThe product of the average value of the abnormal indexes of all gases is used as the abnormal degree of the tail gas data to be detected, and the formula is as follows:in the formula>For the degree of abnormality of the exhaust gas data to be detected, +.>For all gases in the tail gas data to be detectedWhen the difference between the obtained exhaust gas data and the historical exhaust gas data is smaller, namely the calculated Q is smaller, the difference between the obtained exhaust gas data and the exhaust gas data predicted to be obtained according to the engine data E affecting the exhaust gas components is smaller, namely the calculated R is smaller, the abnormality V of the obtained data is indicated to be smaller, wherein>For the accumulated value of the difference between the predicted value and the actual value of each gas content in the exhaust gas data to be detected: />In the formula>For the abnormal index of the exhaust data to be detected, n represents the total number of gas types in the exhaust data to be detected, i represents the i-th gas, +.>Represents the content of the i-th gas, +.>The predicted value of the i-th gas, that is, when the difference between the predicted value of the gas and the actual content of the gas is larger, is the greater the degree of abnormality of the acquired exhaust gas data to be detected.
The specific method for judging whether the exhaust data to be detected is abnormal or not according to the abnormality degree of the exhaust data to be detected comprises the following steps:
when the abnormality degree is greater than an abnormality degree threshold, the abnormality degree threshold is set to be 0.6 in the invention, and when the abnormality degree of the exhaust data to be detected is greater than the abnormality degree threshold, the abnormality of the data of the exhaust analyzer to be detected is detected
When the difference between the gas content and the normal gas content is larger than 0.6, the difference is considered to be larger, the difference is taken as abnormal gas data, and when the abnormal degree of the tail gas data to be detected is detected
When the gas content is smaller than or equal to 0.6, the difference between the gas content and the normal gas content is considered to be smaller, the gas content is used as normal gas data, and the specific abnormality degree threshold can be set according to the requirement of an operator on the detection precision.
The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.
Claims (7)
1. An abnormal data detection method of an exhaust gas analyzer, the method comprising:
acquiring current engine data and corresponding tail gas data to be detected of an automobile, and acquiring historical engine data and corresponding historical tail gas data of the automobile;
obtaining abnormal indexes of each gas in the tail gas data to be detected according to the content ratio of each gas to other gases in the tail gas data to be detected and the difference between the average content of each gas in the historical tail gas data and the average content of other gases;
clustering the historical engine data, and obtaining the influence degree of each engine data on each gas according to the difference of each engine data in each type of historical engine data and the difference of each gas in the historical tail gas data corresponding to the type of historical engine data;
obtaining target engine data according to the distance between the current engine data and the historical engine data;
according to the influence degree of each engine data in the target engine data on each gas and the content of each gas in the tail gas data corresponding to the target engine data, obtaining a predicted value of the content of each gas in the tail gas data to be detected;
obtaining the abnormality degree of the tail gas data to be detected according to the difference between the predicted value and the actual value of each gas content in the tail gas data to be detected and the abnormality index of each gas;
judging whether the tail gas data to be detected is abnormal or not according to the abnormality degree of the tail gas data to be detected;
the method for acquiring the abnormal index of each gas in the tail gas data to be detected comprises the following steps:in (1) the->For the gas in the tail gas data to be detected +.>Abnormal index of->For the gas in the tail gas data to be detected +.>Content of (A),>for gas in historical tail gas data->Content mean value of>Is->Seed gas (s)/(s)>Not equal to->，/>For the +.>Content of seed gas,/->For +.>The average value of the content of the seed gas,is the total number of gas species in the historical tail gas data.
2. The method for detecting abnormal data of an exhaust gas analyzer according to claim 1, wherein the method for obtaining the degree of abnormality of the exhaust gas data to be detected is as follows:
and accumulating the difference value of the predicted value and the actual value of each gas content in the tail gas data to be detected, and taking the product of the accumulated value and the average value of the abnormal indexes of all the gases as the abnormal degree of the tail gas data to be detected.
3. The method for detecting abnormal data of an exhaust gas analyzer according to claim 1, wherein the method for calculating the predicted value of each gas content in the exhaust gas data to be detected is as follows:in (1) the->Is a gas->Predicted value of content>For the number of historical engine data contained in the target engine data, +.>Is the firstHistorical onsetMachine data->For the total number of categories of engine data +.>Is->Engine data->Is->Engine data versus gas->Degree of influence of->Is->Gas +.>Is contained in the composition.
4. The method for detecting abnormal data of an exhaust gas analyzer according to claim 1, wherein the method for obtaining target engine data according to a distance between current engine data and historical engine data comprises:
acquiring a distance between current engine data and each historical engine data:
clustering is carried out based on the distance between the current engine data and each historical engine data, so that a plurality of categories are obtained, and all engine data in the category where the current engine is located are used as target engine data.
5. The abnormal data detection method of an exhaust gas analyzer according to claim 4, wherein the distance between the current engine data and each of the historical engine data is a euclidean distance therebetween.
6. The method for detecting abnormal data of an exhaust gas analyzer according to claim 1, wherein the method for acquiring the degree of influence of each engine data on each gas is as follows:in (1) the->Is->Engine data versus gas->Degree of influence of->Total category number for clustering historical engine data,/-for the historical engine data>Is->Category (S),>is->Gas +.>Is used for the content average value of (1),is->No.>An average of engine data.
7. The method for detecting abnormal data of an exhaust gas analyzer according to claim 1, wherein the method for judging whether the exhaust gas data to be detected is abnormal according to the degree of abnormality of the exhaust gas data to be detected is as follows:
when the abnormality degree of the tail gas data to be detected is larger than the abnormality degree threshold value, the tail gas data to be detected is abnormal.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311256758.7A CN116992247B (en) | 2023-09-27 | 2023-09-27 | Abnormal data detection method of tail gas analyzer |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311256758.7A CN116992247B (en) | 2023-09-27 | 2023-09-27 | Abnormal data detection method of tail gas analyzer |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116992247A CN116992247A (en) | 2023-11-03 |
| CN116992247B true CN116992247B (en) | 2023-12-22 |
Family
ID=88530627
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311256758.7A Active CN116992247B (en) | 2023-09-27 | 2023-09-27 | Abnormal data detection method of tail gas analyzer |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116992247B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN118243775A (en) * | 2024-05-21 | 2024-06-25 | 中用科技有限公司 | Semiconductor AMC dynamic content detection system and method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107577910A (en) * | 2017-08-24 | 2018-01-12 | 中国科学技术大学 | A kind of Vehicular exhaust retrieving concentration method based on profound neutral net |
| CN109165690A (en) * | 2018-09-03 | 2019-01-08 | 安徽优思天成智能科技有限公司 | A kind of method and system of city road network tail gas pollution forecast of distribution |
| CN110941648A (en) * | 2019-11-19 | 2020-03-31 | 广州市自来水有限公司 | Abnormal data identification method, system and storage medium based on cluster analysis |
| CN111896685A (en) * | 2020-07-22 | 2020-11-06 | 多伦汽车检测集团有限公司 | Automobile exhaust remote monitoring system |
| CN112949930A (en) * | 2021-03-17 | 2021-06-11 | 中国科学院合肥物质科学研究院 | PA-LSTM network-based road motor vehicle exhaust high-emission early warning method |
| CN113222209A (en) * | 2021-03-25 | 2021-08-06 | 中国科学技术大学先进技术研究院 | Regional tail gas migration prediction method and system based on domain adaptation and storage medium |
| CN116089846A (en) * | 2023-04-03 | 2023-05-09 | 北京智蚁杨帆科技有限公司 | New energy settlement data anomaly detection and early warning method based on data clustering |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11562182B2 (en) * | 2021-02-04 | 2023-01-24 | Chengdu Qinchuan Iot Technology Co., Ltd. | Methods and systems for detecting detection devices located at energy metering points of natural gas |
-
2023
- 2023-09-27 CN CN202311256758.7A patent/CN116992247B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107577910A (en) * | 2017-08-24 | 2018-01-12 | 中国科学技术大学 | A kind of Vehicular exhaust retrieving concentration method based on profound neutral net |
| CN109165690A (en) * | 2018-09-03 | 2019-01-08 | 安徽优思天成智能科技有限公司 | A kind of method and system of city road network tail gas pollution forecast of distribution |
| CN110941648A (en) * | 2019-11-19 | 2020-03-31 | 广州市自来水有限公司 | Abnormal data identification method, system and storage medium based on cluster analysis |
| CN111896685A (en) * | 2020-07-22 | 2020-11-06 | 多伦汽车检测集团有限公司 | Automobile exhaust remote monitoring system |
| CN112949930A (en) * | 2021-03-17 | 2021-06-11 | 中国科学院合肥物质科学研究院 | PA-LSTM network-based road motor vehicle exhaust high-emission early warning method |
| CN113222209A (en) * | 2021-03-25 | 2021-08-06 | 中国科学技术大学先进技术研究院 | Regional tail gas migration prediction method and system based on domain adaptation and storage medium |
| CN116089846A (en) * | 2023-04-03 | 2023-05-09 | 北京智蚁杨帆科技有限公司 | New energy settlement data anomaly detection and early warning method based on data clustering |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116992247A (en) | 2023-11-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN116992247B (en) | Abnormal data detection method of tail gas analyzer | |
| CN111444953B (en) | Sensor fault monitoring method based on improved particle swarm optimization algorithm | |
| EP2690423B1 (en) | Vehicle data analysis apparatus, vehicle data analysis method, and defect diagnosis apparatus | |
| CN110826772A (en) | Vehicle exhaust emission prediction method, device, equipment and storage medium | |
| CN113187638B (en) | Method for diagnosing high sulfur content in fuel oil | |
| CN111860701B (en) | Denitration system working condition discrimination preprocessing method based on clustering method | |
| CN113239963B (en) | Method, device, equipment, vehicle and storage medium for processing vehicle data | |
| CN112486962A (en) | Extraction and combination short segment calculation heavy-duty diesel vehicle NOxMethod of discharging | |
| CN111289697A (en) | CEMS running state monitoring and data false identification method and system | |
| CN110827444B (en) | Heavy vehicle emission factor obtaining method suitable for OBD remote emission monitoring data | |
| CN115688609A (en) | Intelligent thrust prediction and real-time early warning method for aircraft engine | |
| CN115392116A (en) | Heavy diesel truck excess emission judgment method based on OBD data restoration | |
| CN116992240B (en) | Vehicle emission remote evaluation method, device and storage medium | |
| CN110826600B (en) | Engine surge prediction method based on adaptive resonance network online incremental learning | |
| CN108628279A (en) | A kind of analysis method based on cloud computing automobile abnormal data stream | |
| CN108268678A (en) | Driving behavior analysis method, apparatus and system | |
| CN114662060B (en) | Vehicle-mounted nitrogen oxide sensor concentration measurement value correction method based on machine learning | |
| CN116432397A (en) | Rail transit fault early warning method based on data model | |
| CN114060132B (en) | NO based on emission remote monitoring x Sensor cheating discrimination method | |
| Brandt et al. | Three-way catalyst diagnostics for advanced emissions control systems | |
| CN114659800A (en) | RDE simulation test method, system, storage medium and electronic equipment | |
| CN112862180A (en) | Denitration system inlet NOx concentration prediction method | |
| CN113448301A (en) | Fault diagnosis method and system based on OBD system | |
| CN114112420A (en) | Vehicle nitrogen oxide emission testing method and related device | |
| CN114961949B (en) | Fault diagnosis method, device, equipment and medium of SCR (selective catalytic reduction) system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |