CN117668470B - Intelligent detection method and system for air inlet and outlet faults of engine - Google Patents
Intelligent detection method and system for air inlet and outlet faults of engine Download PDFInfo
- Publication number
- CN117668470B CN117668470B CN202410131487.0A CN202410131487A CN117668470B CN 117668470 B CN117668470 B CN 117668470B CN 202410131487 A CN202410131487 A CN 202410131487A CN 117668470 B CN117668470 B CN 117668470B
- Authority
- CN
- China
- Prior art keywords
- data
- exhaust
- intake
- air
- similarity
- 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
- 238000001514 detection method Methods 0.000 title claims abstract description 65
- 238000000034 method Methods 0.000 claims abstract description 44
- 230000008859 change Effects 0.000 claims abstract description 41
- 238000003745 diagnosis Methods 0.000 claims abstract description 8
- 238000012216 screening Methods 0.000 claims abstract description 6
- 230000006870 function Effects 0.000 claims description 22
- 238000004364 calculation method Methods 0.000 claims description 13
- 238000010606 normalization Methods 0.000 claims description 11
- 238000004590 computer program Methods 0.000 claims description 6
- 238000004458 analytical method Methods 0.000 abstract description 8
- 238000010219 correlation analysis Methods 0.000 abstract description 5
- 238000012545 processing Methods 0.000 abstract description 2
- 230000002159 abnormal effect Effects 0.000 description 6
- 238000004422 calculation algorithm Methods 0.000 description 6
- 239000007789 gas Substances 0.000 description 6
- 238000012544 monitoring process Methods 0.000 description 5
- 230000000694 effects Effects 0.000 description 4
- 230000005856 abnormality Effects 0.000 description 3
- 239000012530 fluid Substances 0.000 description 3
- 230000002123 temporal effect Effects 0.000 description 3
- 230000003247 decreasing effect Effects 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 239000000428 dust Substances 0.000 description 1
- 230000007613 environmental effect Effects 0.000 description 1
- 239000012535 impurity Substances 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01M—TESTING STATIC OR DYNAMIC BALANCE OF MACHINES OR STRUCTURES; TESTING OF STRUCTURES OR APPARATUS, NOT OTHERWISE PROVIDED FOR
- G01M15/00—Testing of engines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- 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)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Combined Controls Of Internal Combustion Engines (AREA)
Abstract
The invention relates to the technical field of data processing, in particular to an intelligent detection method and system for air inlet and outlet faults of an engine, comprising the following steps: the method comprises the steps of collecting data of an engine during operation, obtaining an original data set, obtaining trend characteristic values of each data in the original data set, obtaining data similarity of the same type, data similarity at the same time and intake-exhaust ratio similarity of each intake flow data, obtaining noise possibility of each intake flow data, screening out a plurality of normal intake flow data from all the intake flow data, transmitting the normal intake flow data to a factory diagnosis system of a manufacturer of the engine, and outputting fault detection results. According to the embodiment of the invention, the noise data is accurately detected through the change trend analysis of the same type of data at different moments and the correlation analysis of the different types of data at the same moment, so that the accuracy of detecting the air inlet and outlet faults of the engine is improved.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to an intelligent detection method and system for air inlet and outlet faults of an engine.
Background
An engine is an important energy conversion device with a complete intake and exhaust system for supplying fresh air thereto and exhausting combusted exhaust gases. In order to ensure that the engine can work normally, an air inlet and outlet system can monitor some system data in the air inlet and outlet process in real time and diagnose and early warn abnormal conditions in time. The intake air flow is one of the key monitoring data of the system, and represents the flow of fresh air sucked by the engine from outside, and the sufficient intake air flow can ensure that the engine can use enough oxygen, so that the engine can provide sufficient power for the machine.
In the actual monitoring process, dust and impurities exist in the air, so that the accuracy of a sensor can be influenced, a large number of electric devices exist in an engine cabin, electromagnetic interference can be generated when the electric devices are operated together, unreal noise data can be generated in collected air inlet flow data, the data can influence the normal monitoring of an air inlet and outlet system, and the system can send early warning by mistake when serious. In the existing data denoising algorithm, only the difference between each data and the surrounding part of the data is usually analyzed to detect whether each data is noise or not, so that the final denoising effect is possibly poor, noise interference still exists, and the accuracy of detecting the air inlet and outlet faults of the engine is reduced.
Disclosure of Invention
The invention provides an intelligent detection method and system for an air inlet and outlet fault of an engine, which are used for solving the existing problems.
The invention discloses an intelligent detection method and system for air inlet and outlet faults of an engine, and the intelligent detection method and system adopts the following technical scheme:
the embodiment of the invention provides an intelligent detection method for an engine air intake and exhaust fault, which comprises the following steps:
collecting data of an engine during operation to obtain an original data set; the original data set includesData at each time, the type of the data at each time including an intake air flow rate, an exhaust air flow rate, an intake air temperature, an exhaust air temperature, an intake air pressure, and an exhaust air pressure; the original data set is->The intake air flow data at the moment is recorded as target data; said->Belonging to->The method comprises the steps of carrying out a first treatment on the surface of the In the original data setAccording to the difference between each data and the data at the adjacent time of each data, the trend characteristic value of each data is obtained;
obtaining the same type data similarity of the target data according to the trend characteristic value difference of the target data and all the intake flow data in the original data set respectively;
according to the target data and the original data setThe trend characteristic value difference of the exhaust flow data, the intake pressure data and the intake temperature data at the moment is used for obtaining the data similarity of the target data at the moment;
according to the first set of raw dataTrend characteristic value differences among the intake air flow data, the exhaust air flow data, the intake air pressure data, the exhaust air pressure data, the intake air temperature data and the exhaust air temperature data at moment are obtained to obtain the intake and exhaust ratio similarity of the target data;
obtaining the noise possibility of the target data according to the trend characteristic value, the data similarity of the same type, the data similarity at the same time and the air inlet and outlet proportion similarity of the target data;
screening a plurality of normal air inlet flow data from all the air inlet flow data in the original data set according to the noise possibility of each air inlet flow data in the original data set; and transmitting all the normal air inlet flow data to a factory diagnosis system of the manufacturer of the engine, and outputting a fault detection result.
Further, the trend characteristic value of each data is obtained according to the difference between each data and the data at the adjacent time of each data in all the data of the same type in the original data set, and the specific steps include:
centralizing the original dataTime to->The intake air flow data between the moments is recorded as reference data; said->A preset time threshold value;
calculating all the reference data by using a first derivative method to obtain local extreme points in all the reference data;
in all the reference data, according to the difference of adjacent reference data and the number of local extreme points, obtaining a trend characteristic value of the target data;
and respectively obtaining trend characteristic values of each exhaust flow data, each intake air temperature data, each exhaust temperature data, each intake air pressure data and each exhaust pressure data in all data of the same type in the original data set according to the acquisition mode of the trend characteristic values of the target data.
Further, in all the reference data, according to the difference between the adjacent reference data and the number of local extremum points, a specific calculation formula corresponding to the trend characteristic value of the target data is obtained as follows:
wherein the method comprises the steps ofTrend characteristic value for target data, +.>For the number of reference data>、/>And +.>Respectively +.>Person, th->Person and->Reference data->For the number of local extremal points in all reference data, < >>As absolute function>Is a linear normalization function.
Further, the method for obtaining the data similarity of the same type of the target data according to the trend characteristic value difference of the target data and all the intake flow data in the original data set respectively comprises the following specific steps:
calculating the difference value of the trend characteristic value of each air inlet flow data in the target data and the original data set, calculating the average value of the difference values of the trend characteristic values of all the air inlet flow data in the target data and the original data set, and recording the normalized value of the reciprocal of the average value as the data similarity of the same type of the target data.
Further, the target data are respectively integrated with the original data setThe specific calculation formula corresponding to the similarity of the data at the same time when the target data is obtained is as follows:
wherein the method comprises the steps ofTemporal data similarity for target data, +.>Trend characteristic value for target data, +.>For the +.>Trend characteristic value of exhaust gas flow data at time, +.>For the +.>Intake pressure data at time ∈>Trend characteristic value of>For the +.>Intake air temperature data +.>Trend characteristic value of>As absolute function>Is a linear normalization function.
Further, according to the first set of raw dataIntake air flow data, exhaust air flow data, intake air pressure data, exhaust air pressure data, and intake air temperature data at the moment of timeAnd trend eigenvalue difference between the exhaust temperature data to obtain the similarity of the intake and exhaust proportion of the target data, comprising the following specific steps:
centralizing the target data with the original dataThe ratio of trend characteristic values of the exhaust flow data at the moment is recorded as the change consistency of the intake and exhaust flow;
centralizing the original dataThe ratio of the trend characteristic value of the intake pressure data to the exhaust pressure data at the moment is recorded as the change consistency of the intake pressure and the exhaust pressure;
centralizing the original dataThe ratio of the trend characteristic value of the intake air temperature data to the trend characteristic value of the exhaust air temperature data at the moment is recorded as the change consistency of the intake air temperature and the exhaust air temperature;
and obtaining the similarity of the air inlet and outlet proportion of the target data according to the air inlet and outlet flow, the air inlet and outlet pressure and the air inlet and outlet temperature.
Further, according to the consistency of the changes of the intake and exhaust flow, the intake and exhaust pressure and the intake and exhaust temperature, a specific calculation formula corresponding to the similarity of the intake and exhaust proportions of the target data is obtained as follows:
wherein the method comprises the steps ofIntake and exhaust ratio similarity of target data, +.>Trend characteristic value for target data, +.>Is original asData set +.>Trend characteristic value of exhaust gas flow data at time, +.>For the +.>Intake pressure data at time ∈>Trend characteristic value of>For the +.>Exhaust pressure data ∈time>Trend characteristic value of>For the +.>Intake air temperature data +.>Trend characteristic value of>For the +.>Exhaust temperature data->Trend characteristic value of>As absolute function>As a linear normalization function>For the consistency of the air intake and exhaust flow rate>For the consistency of the changes of the intake and exhaust pressures +.>The temperature of the air inlet and the air outlet is consistent in change.
Further, the method for obtaining the noise possibility of the target data according to the trend characteristic value, the similarity of the same type of data, the similarity of the data at the same time and the similarity of the air inlet and outlet proportion of the target data comprises the following specific steps:
calculating the product of the similarity of the same type of the target data, the similarity of the data at the same time and the similarity of the air inlet and outlet proportion, calculating the ratio of the trend characteristic value of the target data to the product, and recording the normalized value of the ratio as the noise possibility of the target data.
Further, the step of screening out a plurality of normal intake air flow data from all the intake air flow data in the original data set according to the noise possibility of each intake air flow data in the original data set includes the following specific steps:
and (3) recording the air inlet flow data with the noise possibility larger than a preset judging threshold value as normal air inlet flow data in all the air inlet flow data in the original data set.
The invention also provides an intelligent detection system for the air intake and exhaust faults of the engine, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the computer program stored in the memory so as to realize the steps of the intelligent detection method for the air intake and exhaust faults of the engine.
The technical scheme of the invention has the beneficial effects that:
in the embodiment of the invention, the data of the engine during operation is collected to obtain the original data set, the trend characteristic value of each data in the original data set is obtained, the noise data is detected based on the trend characteristic value, and the characteristic difference between the noise data and the normal data is increased, so that the noise detection is more accurate. The method comprises the steps of obtaining the similarity of the same type of data, the similarity of the same time data and the similarity of the ratio of air intake to exhaust of each air intake flow data, so that the noise possibility of each air intake flow data is obtained, and obtaining the accurate noise possibility through the trend characteristic value difference of the same type of data at different times and the trend characteristic value difference of the different types of data at the same time, so that the accuracy of noise detection is improved. And the fault detection result is output by using the reliable normal data for analysis, so that the obtained fault detection result is more reliable. According to the embodiment of the invention, the noise data is accurately detected through the change trend analysis of the same type of data at different moments and the correlation analysis of the different types of data at the same moment, so that the influence of the noise data on the detection of the air intake and exhaust faults of the engine is reduced, and the accuracy of the detection of the air intake and exhaust faults of the engine is improved.
Drawings
In order to more clearly illustrate the embodiments of the invention or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, it being obvious that the drawings in the following description are only some embodiments of the invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of steps of an intelligent detection method for engine air intake and exhaust faults.
Detailed Description
In order to further describe the technical means and effects adopted by the invention to achieve the preset aim, the following is a detailed description of specific implementation, structure, characteristics and effects of the intelligent detection method and system for the air intake and exhaust faults of the engine according to the invention, which are provided by the invention, with reference to the accompanying drawings and the preferred embodiment. In the following description, different "one embodiment" or "another embodiment" means that the embodiments are not necessarily the same. Furthermore, the particular features, structures, or characteristics of one or more embodiments may be combined in any suitable manner.
Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.
The embodiment of the invention provides an intelligent detection method and an intelligent detection system for an engine air intake and exhaust fault, which are used for accurately detecting noise data through analysis of variation trend of the same type of data at different moments and correlation analysis among the different types of data at the same moment, reducing the influence of the noise data on the detection of the engine air intake and exhaust fault and improving the accuracy of the detection of the engine air intake and exhaust fault.
The invention provides an intelligent detection method and a system for an engine air intake and exhaust fault, which are specifically described below with reference to the accompanying drawings.
Referring to fig. 1, a flowchart of steps of an intelligent detection method for an intake and exhaust failure of an engine according to an embodiment of the present invention is shown, where the method includes the following steps:
step S001: collecting data of an engine during operation to obtain an original data set; the original data set includesData at each time, the type of the data at each time including an intake air flow rate, an exhaust air flow rate, an intake air temperature, an exhaust air temperature, an intake air pressure, and an exhaust air pressure; the original data set is->The intake air flow data at the moment is recorded as target data; said->Belonging toThe method comprises the steps of carrying out a first treatment on the surface of the And obtaining a trend characteristic value of each data according to the difference between each data and the data at the adjacent time of each data in all the data of the same type in the original data set.
According to the method, the trend characteristic value of each data is obtained by analyzing the trend change of the same type of data in the original data set in time and is used as the parameter basis for the subsequent noise data detection, so that the accuracy of the noise data detection is improved. One possible implementation of obtaining the trend feature value of each data is as follows:
during a period of engine operation, the following data were collected using the air flow sensor, the temperature sensor, and the pressure sensor, respectively: the method comprises the steps of obtaining an original data set by air inlet flow, air outlet flow, air inlet temperature, air outlet temperature, air inlet pressure and air outlet pressure. The original data set comprisesThe data at each moment, and the collection frequency of the data of each type is the same, namely all types of data exist at the same moment. The preset data acquisition frequency of this embodiment is 0.2 seconds, the number of acquisition time is +.>1000, for example, other values may be set in other embodiments, and the present embodiment is not limited thereto.
In this embodiment, intake air flow data is taken as an example, in the actual monitoring process of the intake air flow data, noise data generated due to some interference factors is difficult to avoid, and the noise data has no practical meaning on monitoring of an intake and exhaust system and also has negative effects. And the size of the data is not obviously different from that of the real data, so that the detection of the data is more difficult, which is particularly obvious when the data is analyzed in whole.
However, the time sequence of the noise data is usually obviously characterized, the noise data is often reflected in data abrupt change for a certain period of time, so that the change speed of the noise data before and after is rapid, the change trend is unstable, and the change of the real intake air flow data under the normal physical environment is relatively gentle and stable. Therefore, the trend characteristic value of each air inlet flow data can be obtained through the change trend of the air inlet flow data in a period of time.
The time threshold preset in this embodimentIn the description of this example, other values may be set in other embodiments, and the present example is not limited thereto.
Centralizing the original dataThe intake air flow rate data at the time is recorded as target data. Wherein->Belonging to->。
In the original data set, the firstTime to->Intake air flow data between the moments is noted as reference data.
What needs to be described is: in the original data set, all intake air flow rate data constitute a time series data sequence arranged in time series, and all reference data constitute a time series data sequence segment in the time series data sequence.
And calculating all the reference data by using a first derivative method to obtain local extreme points in all the reference data. The first derivative method is a known technique, and the specific method is not described here.
Thus, the trend characteristic value of the target data can be knownThe calculation formula of (2) is as follows:
wherein the method comprises the steps ofTrend characteristic value for target data, +.>For the number of reference data>、/>And +.>Respectively +.>Person, th->Person and->Reference data->For the number of local extremal points in all reference data, < >>As absolute function>Normalizing the data values to [0,1 ] as a linear normalization function]Within the interval. When->The larger the description is inFirst->Time to->The more the number of transitions of the increasing and decreasing trend of the reference data between the moments, the less stable the trend change of the reference data, the more noisy the target data. When->The larger the instruction in +.>Time to->The larger the difference between adjacent reference data between the moments, the larger the change speed of the reference data, the stronger the noise performance of the target data. When->The larger the instruction in +.>Time to->The less stable the rate of change of the reference data between moments, the more noisy the target data will be. Therefore use->、/>And +.>Normalized value of the product of (2) representing trend characteristic value of the target data,/>The larger the target data is, the more likely the noise is.
According to the mode, the trend characteristic value of each air inlet flow data in the original data set is obtained.
And in the original data set, according to the acquisition mode of the trend characteristic value of the target data, the trend characteristic value of each exhaust flow data is obtained in all the exhaust flow data. And obtaining trend characteristic values of each air inlet temperature data in all air inlet temperature data. And obtaining trend characteristic values of each exhaust temperature data in all the exhaust temperature data. And obtaining trend characteristic values of each air inlet pressure data in all air inlet pressure data. And obtaining trend characteristic values of each exhaust pressure data in all the exhaust pressure data.
What needs to be described is: the formula passes throughTime to->The trend characteristic value of the target data is constructed by referring to the change speed and the change trend of the data between the moments, and the trend characteristic value also represents the noise expression degree of the target data, and the larger the trend characteristic value is, the larger the noise expression degree is. If one data is noise data, the change speed of the data adjacent to the time of the data is larger, the change trend is unstable, and the corresponding noise expression degree is larger, and the trend characteristic value is larger. The difference between noise data and real data can be obviously increased through the constructed trend characteristic value, so that the abnormality detection accuracy of an algorithm is improved, and a foundation is provided for subsequent data similarity analysis.
Step S002: and obtaining the data similarity of the same type of the target data according to the trend characteristic value difference of the target data and all the intake flow data in the original data set.
The step is to obtain the similarity of the same type of data by analyzing the trend characteristic value difference between the same type of data in the original data set, and use the similarity as a judgment parameter of noise data to improve the accuracy of noise data detection. One possible implementation of obtaining the same type of data similarity for each data is as follows:
it is known that in the same physical environment, various data changes of the air intake and exhaust system do not have larger differences. Although there are different data change trends in different time periods for the intake air flow rate data, these changes are all within a certain range, and only the change of the noise data may show a large abnormality. Thus, the mean value of the difference of the trend feature value of each data from all other data can be calculated to represent the similarity of the data change of the target data over the whole time.
Thereby knowing the data similarity of the same type of the target dataThe calculation formula of (2) is as follows:
wherein the method comprises the steps ofFor the same type of data similarity of the target data +.>For the amount of intake air flow data in the original dataset, +.>Trend characteristic value for target data, +.>For the +.>Trend characteristic value of individual intake air flow data, +.>As absolute function>Normalizing the data values to [0,1 ] as a linear normalization function]Within the interval.Mean value representing difference of trend characteristic values of the target data from all of the intake air flow rate data, respectively, the larger the mean value, the more abnormal the trend characteristic value of the target data is described, therefore +.>Normalized value of reciprocal of (2) representing the same type of data similarity of the target data,/2>The larger the target data is, the more trusted it is, and the less likely it is noise.
What needs to be described is: the formula characterizes the similarity of the target data in all the air inflow data by the mean value of the differences of the trend characteristic values of the target data and all other data in the air inflow data, wherein the larger the difference of the trend characteristic values is, the smaller the similarity is, the smaller the difference of the trend characteristic values is, and the larger the similarity is. The similarity calculated by the method can better accord with the overall characteristics of the data, and the situation of local optimization is avoided.
Step S003: according to the target data and the original data setAnd obtaining the trend characteristic value difference of the exhaust flow data, the intake pressure data and the intake temperature data at the moment, and obtaining the similarity of the target data at the moment.
The step is to obtain the similarity of the data at the same time by analyzing the trend characteristic value difference between different types of data in the original data set at the same time, and use the similarity as a judgment parameter of noise data to improve the accuracy of noise data detection. One possible implementation of the temporal data similarity for each data is as follows:
it is known that some of the various data monitored by the intake and exhaust systems may change simultaneously with changes in intake air flow. The air inlet and outlet system is a closed system, and the mass entering the system is equal to the mass leaving the system according to the law of conservation of mass, so that the air inlet flow and the air outlet flow at the same time are approximate, and the change trend of the air inlet flow and the air outlet flow is consistent. From the maston equation, it is known that when the flow changes, the pressure and temperature changes can be regarded as logarithmic changes approximately.
What needs to be described is: the maston equation is a well known technique, and a specific method is not described here. The maston equation is one of the basic equations describing fluid flow in fluid mechanics and describes the relationship between mass flow, pressure, and temperature of a fluid, where mass flow is proportional to pressure, and proportional to temperature and velocity.
Thereby knowing the data similarity of the target data at the same timeThe calculation formula of (2) is as follows:
wherein the method comprises the steps ofTemporal data similarity for target data, +.>Trend characteristic value for target data, +.>For the +.>Trend characteristic value of exhaust gas flow data at time, +.>For the +.>Intake pressure data at time ∈>Trend characteristic value of>For the +.>Intake air temperature data +.>Trend characteristic value of>As absolute function>Normalizing the data values to [0,1 ] as a linear normalization function]Within the interval. When->The larger the difference between the intake air flow data and trend characteristic values of the exhaust air flow, the intake air pressure and the intake air temperature data respectively at the moment, the more abnormal the target data is, and the trend of the intake air flow data, the intake air pressure and the intake air temperature data respectively, which are known to be in an exponential-logarithmic relationship according to the deduction result of a Maston equation, therefore, the calculation of +_>Respectively and->And->The difference of (2) is thus->Normalized value of reciprocal of (2), representing the temporal data similarity of the target data,/v>The smaller the target data, the more trusted it is, and the less likely it is to be noise.
What needs to be described is: the formula is used for constructing a similarity model of the air inlet flow by comparing the differences of trend characteristic values of various types of data at the same moment, so that whether the change of the air inlet flow data is normal or not can be analyzed from a plurality of data dimensions, and the reliability of an algorithm is improved.
Step S004: according to the first set of raw dataAnd obtaining the similarity of the intake and exhaust proportion of the target data by trend characteristic value differences among the intake air flow data, the exhaust air flow data, the intake air pressure data, the exhaust air pressure data, the intake air temperature data and the exhaust air temperature data at the moment.
According to the method, the proportion difference of trend characteristic values between the corresponding types of data of the air inlet and the air outlet in the original data set is analyzed, the similarity of the air inlet and the air outlet proportion is obtained and is used as a judgment parameter of noise data, and the accuracy of noise data detection is improved. One possible implementation way of obtaining the intake-exhaust ratio similarity of each data is as follows:
according to the working principle of the engine and the structure of the air intake and exhaust system, the change relation between air intake data and exhaust data can be analyzed. The known air inlet flow and the exhaust flow are kept, the change trend between the air inlet flow and the exhaust flow tends to be consistent, the air inlet pressure tends to be higher than the exhaust pressure, the reason is that part of energy is lost in the engine, however, the lost energy is usually smaller and does not change greatly, the change trend of the air inlet pressure and the exhaust pressure also tends to be consistent, the air inlet temperature is much lower than the exhaust temperature, the reason is that a large amount of heat energy is generated in the combustion process, however, the change trend relationship of the air inlet flow and the exhaust flow, the air inlet pressure and the exhaust pressure is known, and the change trend of the air inlet temperature and the exhaust temperature can be obtained according to the relationship of the flow and the temperature in a Marston equation and the relationship of the pressure and the temperature in an ideal gas state equation.
Thus, the similarity of the air inlet and outlet ratios of the target data can be knownThe calculation formula of (2) is as follows:
wherein the method comprises the steps ofIntake and exhaust ratio similarity of target data, +.>Trend characteristic value for target data, +.>For the +.>Trend characteristic value of exhaust gas flow data at time, +.>For the +.>Intake pressure data at time ∈>Trend characteristic value of>For the +.>Exhaust pressure data ∈time>Trend characteristic value of>For the +.>Intake air temperature data +.>Trend characteristic value of>For the +.>Exhaust temperature data->Trend characteristic value of>As absolute function>Normalizing the data values to [0,1 ] as a linear normalization function]Within the interval. />For the consistency of the air intake and exhaust flow rate>For the consistency of the variation of the intake and exhaust pressures,for the uniformity of the temperature change of the intake and exhaust, when +.>And->The larger the difference between the intake and exhaust flow rates and the consistency of the intake and exhaust pressure and temperature changes, the more abnormal the target data, soNormalized value of reciprocal of (2) representing intake-exhaust ratio similarity of target data,/v->The smaller the target data, the more trusted it is, and the less likely it is to be noise.
What needs to be described is: the formula judges whether the change of the intake air flow data is normal or not by comparing the change consistency between the intake and exhaust data at the same time, and when abnormal mutation occurs to the data, the change consistency of the intake and exhaust data can be obviously increased or decreased, so that the abnormal degree of the data can be better reflected.
Step S005: and obtaining the noise possibility of the target data according to the trend characteristic value, the similarity of the same type of data, the similarity of the data at the same time and the similarity of the air inlet and outlet proportions of the target data.
According to the method, the noise possibility of each data is obtained through the trend characteristic value, the data similarity of the same type, the data similarity at the same time and the air inlet and outlet proportion similarity of each data in the obtained original data set, so that the noise data is accurately screened out, the data provided during subsequent fault detection is more credible, and the fault detection accuracy is guaranteed. One possible implementation of obtaining the noise probability of each data is as follows:
thereby knowing the noise probability of the target dataThe calculation formula of (2) is as follows:
wherein the method comprises the steps ofNoise probability for target data, +.>Is the object ofTrend characteristic value of data->For the same type of data similarity of the target data +.>Temporal data similarity for target data, +.>Intake and exhaust ratio similarity of target data, +.>As absolute function>Normalizing the data values to [0,1 ] as a linear normalization function]Within the interval.
What needs to be described is:the larger the target data is, the more likely it is noise, and +.>、/>And +.>The smaller the target data is, the more reliable it is, the less likely it is noise, thus using +.>Is indicative of the noise probability of the target data,/->The larger the target data is, the greater the likelihood that the target data is noise. Thereby, through analysis of the time variation trend of the same type of data and correlation analysis among different types of data, the accurate possibility that each data is noise is obtained, so that the noiseThe detection is more accurate.
In the above manner, the noise probability of each intake air flow rate data in the raw data set is obtained.
Step S006: screening a plurality of normal air inlet flow data from all the air inlet flow data in the original data set according to the noise possibility of each air inlet flow data in the original data set; and transmitting all the normal air inlet flow data to a factory diagnosis system of the manufacturer of the engine, and outputting a fault detection result.
According to the method, the normal data are screened out according to the noise possibility of each data, and the normal data are used for fault detection, so that the accuracy of fault detection is guaranteed. One possible implementation of obtaining the fault detection result is as follows:
the preset determination threshold value in this embodiment is 0.85, which is described as an example, and other values may be set in other embodiments, which is not limited in this embodiment.
And (3) recording the air inlet flow data with the noise possibility larger than a preset judging threshold value as normal air inlet flow data in all the air inlet flow data in the original data set.
And transmitting all the normal air inlet flow data to a factory diagnosis system of the manufacturer of the engine, and outputting a fault detection result.
What needs to be described is: in all the normal air intake flow data, the air intake flow data with the noise possibility smaller than or equal to a preset judgment threshold value is absent, so that the missing air intake flow data is filled by an interpolation algorithm before the fault detection of the original plant diagnosis system is carried out, and the complete normal air intake flow data is obtained. The interpolation algorithm is a well-known technique, and a specific method is not described herein. The engine manufacturer's factory diagnostic system provides its own factory diagnostic system and software for the engine manufacturer, which systems typically have greater accuracy and more detailed fault diagnosis functionality.
The fault detection result obtained by the method reflects the fault detection of the engine air inlet. For fault detection of engine exhaust, the embodiment screens out all normal exhaust flow data from all exhaust flow data in the original data set according to the acquisition process of normal intake flow data, and outputs a fault detection result according to all normal exhaust flow data in the above manner, which represents fault detection of engine exhaust.
What needs to be described is: the trend characteristic value of the exhaust flow data is already obtained, so that the data similarity of the same type of the exhaust flow data can be obtained according to the obtaining mode of the data similarity of the same type of the target data, and the trend characteristic value of the exhaust flow data is the firstThe acquisition of the data similarity at the same time as the exhaust flow data at the moment is only required to be +.>Is ∈r in the calculation formula of (a)>And->Modified to->Andand (3) obtaining the product. First->The similarity of the intake and exhaust ratio of the exhaust flow data at the moment is +.>Thus, the noise possibility of the exhaust flow data is obtained, and all the normal exhaust flow data are screened out.
In intake air flow data collected by an intake and exhaust system of an engine, there are both real data and unreal noise data generated due to environmental interference or equipment abnormality. In order to increase the feature difference between the noise data and the real data and increase the accuracy of anomaly detection, the present embodiment constructs a unique trend feature value to increase the noise performance of the noise data in its entirety. The problem of local optimum is easily generated by noise analysis of the data only through local expression of single type data, so that the embodiment analyzes the expression of the air inlet flow data in the whole time and multi-dimensional data to improve the accuracy of the algorithm.
The present invention has been completed.
In summary, in the embodiment of the present invention, data of an engine during operation is collected to obtain an original data set, and a trend feature value of each data in the original data set is obtained. Will be the firstAnd recording the intake air flow data at the moment as target data, and obtaining the data similarity of the same type of the target data according to the trend characteristic value difference of the target data and all the intake air flow data respectively. According to->The method comprises the steps of obtaining trend characteristic value differences among intake air flow data, exhaust air flow data, intake air pressure data, exhaust air pressure data, intake air temperature data and exhaust air temperature data at moment, obtaining the similarity of the data at the moment and the similarity of the intake and exhaust ratio respectively, obtaining noise possibility of the target data, screening out a plurality of normal intake air flow data from all the intake air flow data, transmitting the normal intake air flow data to an original factory diagnosis system of a manufacturer of an engine, and outputting a fault detection result. According to the embodiment of the invention, the noise data is accurately detected through the change trend analysis of the same type of data at different moments and the correlation analysis of the different types of data at the same moment, so that the influence of the noise data on the detection of the air intake and exhaust faults of the engine is reduced, and the accuracy of the detection of the air intake and exhaust faults of the engine is improved.
The invention also provides an intelligent detection system for the air intake and exhaust faults of the engine, which comprises a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the processor executes the computer program stored in the memory so as to realize the steps of the intelligent detection method for the air intake and exhaust faults of the engine.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the invention, but any modifications, equivalent substitutions, improvements, etc. within the principles of the present invention should be included in the scope of the present invention.
Claims (9)
1. An intelligent detection method for an engine air intake and exhaust fault is characterized by comprising the following steps:
collecting data of an engine during operation to obtain an original data set; the original data set includesData at each time, the type of the data at each time including an intake air flow rate, an exhaust air flow rate, an intake air temperature, an exhaust air temperature, an intake air pressure, and an exhaust air pressure; the original data set is->The intake air flow data at the moment is recorded as target data; said->Belonging to->The method comprises the steps of carrying out a first treatment on the surface of the In all data of the same type in the original data set, according to the difference between each data and the data at the adjacent time of each data, obtaining the trend characteristic value of each data;
obtaining the same type data similarity of the target data according to the trend characteristic value difference of the target data and all the intake flow data in the original data set respectively;
according to the target data and the original data setExhaust flow data, intake pressure data, and intake temperature data at timeThe trend characteristic value difference is used for obtaining the similarity of the data at the same time of the target data;
according to the first set of raw dataTrend characteristic value differences among the intake air flow data, the exhaust air flow data, the intake air pressure data, the exhaust air pressure data, the intake air temperature data and the exhaust air temperature data at moment are obtained to obtain the intake and exhaust ratio similarity of the target data;
obtaining the noise possibility of the target data according to the trend characteristic value, the data similarity of the same type, the data similarity at the same time and the air inlet and outlet proportion similarity of the target data;
screening a plurality of normal air inlet flow data from all the air inlet flow data in the original data set according to the noise possibility of each air inlet flow data in the original data set; transmitting all normal air inflow data to a factory diagnosis system of a manufacturer of the engine, and outputting a fault detection result;
according to the noise possibility of each air inlet flow data in the original data set, a plurality of normal air inlet flow data are screened out from all the air inlet flow data in the original data set, and the method comprises the following specific steps:
and (3) recording the air inlet flow data with the noise possibility larger than a preset judging threshold value as normal air inlet flow data in all the air inlet flow data in the original data set.
2. The intelligent detection method for engine intake and exhaust faults according to claim 1, wherein the trend characteristic value of each data is obtained according to the difference between each data and the data at the adjacent time of each data in all data of the same type in the original data set, and the method comprises the following specific steps:
centralizing the original dataTime to->The intake air flow data between the moments is recorded as reference data; said->A preset time threshold value;
calculating all the reference data by using a first derivative method to obtain local extreme points in all the reference data;
in all the reference data, according to the difference of adjacent reference data and the number of local extreme points, obtaining a trend characteristic value of the target data;
and respectively obtaining trend characteristic values of each exhaust flow data, each intake air temperature data, each exhaust temperature data, each intake air pressure data and each exhaust pressure data in all data of the same type in the original data set according to the acquisition mode of the trend characteristic values of the target data.
3. The intelligent detection method for engine air intake and exhaust faults according to claim 2, wherein the specific calculation formula corresponding to the trend characteristic value of the target data is obtained in all the reference data according to the difference of adjacent reference data and the number of local extreme points, and is as follows:
wherein the method comprises the steps ofTrend characteristic value for target data, +.>For the number of reference data>、/>And +.>Respectively +.>Person, th->Person and->Reference data->For the number of local extremal points in all reference data, < >>As absolute function>Is a linear normalization function.
4. The intelligent detection method for engine intake and exhaust faults according to claim 1, wherein the obtaining the data similarity of the same type of the target data according to the trend characteristic value difference of the target data and all intake flow data in the original data set respectively comprises the following specific steps:
calculating the difference value of the trend characteristic value of each air inlet flow data in the target data and the original data set, calculating the average value of the difference values of the trend characteristic values of all the air inlet flow data in the target data and the original data set, and recording the normalized value of the reciprocal of the average value as the data similarity of the same type of the target data.
5. The intelligent detection method for engine air intake and exhaust faults according to claim 1, wherein the intelligent detection method is based on targetsData is respectively with the first data setThe specific calculation formula corresponding to the similarity of the data at the same time when the target data is obtained is as follows:
wherein the method comprises the steps ofTemporal data similarity for target data, +.>Trend characteristic value for target data, +.>For the +.>Trend characteristic value of exhaust gas flow data at time, +.>For the +.>Intake pressure data at timeTrend characteristic value of>For the +.>Number of intake air temperatures at timeAccording to->Trend characteristic value of>As absolute function>Is a linear normalization function.
6. The intelligent detection method for engine air intake and exhaust faults according to claim 1, wherein the method is characterized in that according to the first data setThe trend characteristic value difference among the intake air flow data, the exhaust air flow data, the intake air pressure data, the exhaust air pressure data, the intake air temperature data and the exhaust air temperature data at the moment is used for obtaining the intake and exhaust ratio similarity of target data, and the method comprises the following specific steps of:
centralizing the target data with the original dataThe ratio of trend characteristic values of the exhaust flow data at the moment is recorded as the change consistency of the intake and exhaust flow;
centralizing the original dataThe ratio of the trend characteristic value of the intake pressure data to the exhaust pressure data at the moment is recorded as the change consistency of the intake pressure and the exhaust pressure;
centralizing the original dataThe ratio of the trend characteristic value of the intake air temperature data to the trend characteristic value of the exhaust air temperature data at the moment is recorded as the change consistency of the intake air temperature and the exhaust air temperature;
and obtaining the similarity of the air inlet and outlet proportion of the target data according to the air inlet and outlet flow, the air inlet and outlet pressure and the air inlet and outlet temperature.
7. The intelligent detection method for engine air intake and exhaust faults according to claim 6, wherein the specific calculation formula corresponding to the air intake and exhaust proportion similarity of the target data is obtained according to the air intake and exhaust flow, air intake and exhaust pressure and air intake and exhaust temperature change consistency, and is as follows:
wherein the method comprises the steps ofIntake and exhaust ratio similarity of target data, +.>Trend characteristic value for target data, +.>For the +.>Trend characteristic value of exhaust gas flow data at time, +.>For the +.>Intake pressure data at timeTrend characteristic value of>For the +.>Exhaust pressure data ∈time>Trend characteristic value of>For the +.>Intake air temperature data +.>Trend characteristic value of>For the +.>Exhaust temperature data->Trend characteristic value of>As absolute function>As a linear normalization function>For the consistency of the air intake and exhaust flow rate>For the consistency of the changes of the intake and exhaust pressures +.>To make the change of air inlet and outlet temperature uniformSex.
8. The intelligent detection method for engine air intake and exhaust faults according to claim 1, wherein the noise possibility of the target data is obtained according to trend characteristic values, data similarity of the same type, data similarity at the same time and air intake and exhaust proportion similarity of the target data, and the method comprises the following specific steps:
calculating the product of the similarity of the same type of the target data, the similarity of the data at the same time and the similarity of the air inlet and outlet proportion, calculating the ratio of the trend characteristic value of the target data to the product, and recording the normalized value of the ratio as the noise possibility of the target data.
9. An intelligent detection system for engine air intake and exhaust faults, comprising a memory, a processor and a computer program stored on the memory and capable of running on the processor, wherein the computer program when executed by the processor realizes the steps of the intelligent detection method for engine air intake and exhaust faults according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410131487.0A CN117668470B (en) | 2024-01-31 | 2024-01-31 | Intelligent detection method and system for air inlet and outlet faults of engine |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410131487.0A CN117668470B (en) | 2024-01-31 | 2024-01-31 | Intelligent detection method and system for air inlet and outlet faults of engine |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN117668470A CN117668470A (en) | 2024-03-08 |
| CN117668470B true CN117668470B (en) | 2024-04-12 |
Family
ID=90064492
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410131487.0A Active CN117668470B (en) | 2024-01-31 | 2024-01-31 | Intelligent detection method and system for air inlet and outlet faults of engine |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117668470B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113266461A (en) * | 2021-06-08 | 2021-08-17 | 湖南道依茨动力有限公司 | Fault detection method, control device, turbocharger and engine system |
| CN115144186A (en) * | 2022-05-25 | 2022-10-04 | 西北工业大学 | Gas turbine engine gas path fault continuous high-precision diagnosis method |
| CN116451484A (en) * | 2023-04-20 | 2023-07-18 | 中国航空发动机研究院 | Engine gas circuit fault simulation method and device, electronic equipment and storage medium |
| CN117407827A (en) * | 2023-12-15 | 2024-01-16 | 湖南辉达净化工程有限公司 | Abnormal operation data detection method for purification engineering waste gas purification equipment |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7702447B2 (en) * | 2006-12-18 | 2010-04-20 | United Technologies Corporation | Method and system for identifying gas turbine engine faults |
| WO2016060994A1 (en) * | 2014-10-16 | 2016-04-21 | Tula Technology, Inc. | Engine error detection system |
-
2024
- 2024-01-31 CN CN202410131487.0A patent/CN117668470B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113266461A (en) * | 2021-06-08 | 2021-08-17 | 湖南道依茨动力有限公司 | Fault detection method, control device, turbocharger and engine system |
| CN115144186A (en) * | 2022-05-25 | 2022-10-04 | 西北工业大学 | Gas turbine engine gas path fault continuous high-precision diagnosis method |
| CN116451484A (en) * | 2023-04-20 | 2023-07-18 | 中国航空发动机研究院 | Engine gas circuit fault simulation method and device, electronic equipment and storage medium |
| CN117407827A (en) * | 2023-12-15 | 2024-01-16 | 湖南辉达净化工程有限公司 | Abnormal operation data detection method for purification engineering waste gas purification equipment |
Non-Patent Citations (2)
| Title |
|---|
| 基于EMD算法的发动机故障监测研究;范晓梅;许勇;李炎;;内燃机学报;20090525(第03期);全文 * |
| 基于信息融合的发动机全系统故障诊断研究;沈寿林, 郑海起, 张英堂;车用发动机;20030430(第02期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN117668470A (en) | 2024-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111275288B (en) | XGBoost-based multidimensional data anomaly detection method and device | |
| CN109857094B (en) | Two-stage Kalman filtering algorithm-based aeroengine fault diagnosis method | |
| CN110608187B (en) | Axial-flow compressor stall surge prediction device based on frequency characteristic change | |
| CN117708748B (en) | Operation monitoring system and method for centrifugal fan | |
| US20190242319A1 (en) | Control device, control method, and control program | |
| CN108446529B (en) | Organic Rankine cycle system fault detection method based on generalized mutual entropy-DPCA algorithm | |
| CN111637045B (en) | Fault diagnosis method for air compressor of ocean platform | |
| CN109710983B (en) | Diesel engine cylinder layered fault diagnosis method based on key performance indexes | |
| CN110472335B (en) | Sensor fault diagnosis threshold value determination method based on particle swarm optimization algorithm | |
| WO2023207190A1 (en) | Fault early warning system and method for fossil fuel power plant, electronic device, and storage medium | |
| CN112052871B (en) | Rocket engine health diagnosis method and system based on support vector machine | |
| CN113469061A (en) | Gas turbine fault detection method based on principal component analysis method | |
| CN112443479A (en) | Compressor fault diagnosis method and device, storage medium and electronic equipment | |
| CN117195007A (en) | Heat exchanger performance prediction method and system | |
| CN107607342B (en) | Healthy energy efficiency detection method for air conditioner room equipment group | |
| CN117668470B (en) | Intelligent detection method and system for air inlet and outlet faults of engine | |
| CN117807538A (en) | Fault detection method based on dynamic threshold and trend prediction | |
| CN117454145A (en) | Aeroengine fault diagnosis method based on multi-mode sensor | |
| CN116577104A (en) | Engine fault detection method and equipment | |
| KR102212022B1 (en) | Method of automatically determining condition of hydro turbine in hydroelectric power plant and system for the same | |
| CN110146267B (en) | Method for detecting flutter of valve plate of totally-enclosed refrigeration compressor | |
| Hongkun et al. | Intelligent identification of rotating stall for centrifugal compressor based on pressure pulsation signals and SDKAE network | |
| CN113266449A (en) | Method and system for predicting air leakage situation in front of aftertreatment system | |
| Qin et al. | Intelligent Edge Gearbox Faults Diagnosis System via Multiscale Depthwise Separable Convolution Network | |
| CN118410364A (en) | Fault detection method of aeroengine control system sensor based on self-adaptive KPCA |
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 |