CN113915015A - Engine running state standard determining method, early warning method and device and vehicle - Google Patents
Engine running state standard determining method, early warning method and device and vehicle Download PDFInfo
- Publication number
- CN113915015A CN113915015A CN202111168871.0A CN202111168871A CN113915015A CN 113915015 A CN113915015 A CN 113915015A CN 202111168871 A CN202111168871 A CN 202111168871A CN 113915015 A CN113915015 A CN 113915015A
- Authority
- CN
- China
- Prior art keywords
- data
- matrix
- value
- characteristic
- vector
- 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.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 46
- 239000011159 matrix material Substances 0.000 claims abstract description 133
- 239000013598 vector Substances 0.000 claims abstract description 94
- 238000012549 training Methods 0.000 claims abstract description 54
- 238000012544 monitoring process Methods 0.000 claims abstract description 17
- 238000005070 sampling Methods 0.000 claims description 19
- 238000004364 calculation method Methods 0.000 claims description 9
- 238000000354 decomposition reaction Methods 0.000 claims description 5
- 238000013507 mapping Methods 0.000 claims description 4
- 230000000694 effects Effects 0.000 abstract description 5
- 239000000306 component Substances 0.000 description 34
- 238000000605 extraction Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 1
- 239000008358 core component Substances 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 238000012847 principal component analysis method Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000008707 rearrangement Effects 0.000 description 1
- 230000035945 sensitivity Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Images
Classifications
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
- F02D—CONTROLLING COMBUSTION ENGINES
- F02D41/00—Electrical control of supply of combustible mixture or its constituents
- F02D41/22—Safety or indicating devices for abnormal conditions
-
- F—MECHANICAL ENGINEERING; LIGHTING; HEATING; WEAPONS; BLASTING
- F02—COMBUSTION ENGINES; HOT-GAS OR COMBUSTION-PRODUCT ENGINE PLANTS
- F02D—CONTROLLING COMBUSTION ENGINES
- F02D41/00—Electrical control of supply of combustible mixture or its constituents
- F02D41/02—Circuit arrangements for generating control signals
- F02D41/14—Introducing closed-loop corrections
- F02D41/1401—Introducing closed-loop corrections characterised by the control or regulation method
- F02D41/1405—Neural network control
-
- 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)
- Chemical & Material Sciences (AREA)
- Combustion & Propulsion (AREA)
- Mechanical Engineering (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Combined Controls Of Internal Combustion Engines (AREA)
Abstract
The embodiment of the invention discloses an engine running state standard determining method, an early warning device and a vehicle. The engine operating condition criterion determining method includes: acquiring original training data, wherein the original training data comprises normal data and fault data; acquiring a characteristic data matrix and a space base vector, and acquiring a projection value matrix of the characteristic data matrix on the space base vector according to original training data; determining the weight of each vector in the projection value matrix through the projection value matrix; arranging the weights from large to small, and acquiring vectors in the first r projection value matrixes as principal components; calculating a limit value corresponding to each principal component; and selecting a limit value of a preset percentage as a control limit value. The embodiment of the invention improves the fault monitoring accuracy of the running state of the engine, advances the time for finding the running fault of the engine, and has more ideal fault monitoring effect on the running state of the engine.
Description
Technical Field
The embodiment of the invention relates to an engine monitoring technology, in particular to an engine running state standard determining method, an early warning method and device and a vehicle.
Background
The engine is a core component of the vehicle, and the running state and reliability of the engine directly affect the reliability of the vehicle. Therefore, whether the running state of the engine is normal or not is judged by monitoring the running characteristics of the engine, so that the fault of the engine can be found in time, and the reliability of the vehicle is ensured.
In the prior art, after the characteristic variables of the engine are collected, thresholds are respectively set for the monitored characteristic variables, and whether each characteristic variable is abnormal or not is respectively judged. It is difficult to effectively utilize the multivariable advantage of the collected variables, and thus the fault monitoring effect on the engine operating state is not ideal.
Disclosure of Invention
The invention provides a method for determining the standard of an engine running state, a method and a device for early warning and a vehicle, which are used for improving the fault monitoring accuracy of the engine running state, advancing the time for finding the running fault of the engine and having more ideal fault monitoring effect on the engine running state.
In a first aspect, an embodiment of the present invention provides an engine operating condition criterion determining method, including:
acquiring original training data, wherein the original training data comprises normal data and fault data;
acquiring a characteristic data matrix and a space basis vector, and acquiring a projection value matrix of the characteristic data matrix on the space basis vector according to the original training data;
determining the weight of each vector in the projection value matrix through the projection value matrix;
arranging the weights from large to small, and acquiring vectors in the first r projection value matrixes as principal components;
calculating a limit value corresponding to each principal component;
and selecting the limit value of the preset percentage as a control limit value.
Optionally, obtaining a feature data matrix and a space basis vector, and obtaining a projection value matrix of the feature data matrix on the space basis vector according to the original training data includes:
extracting each sampling frequency characteristic variable of the original training data;
weighting each sampling frequency characteristic variable to obtain a characteristic data matrix;
performing eigenvalue decomposition on the characteristic data matrix to obtain a space basis vector;
and mapping the characteristic data matrix to the space base vector to obtain a projection value matrix of the characteristic data matrix on the space base vector.
Optionally, extracting each sampling frequency feature variable of the original training data includes:
and extracting each sampling frequency characteristic variable of the original training data through a sliding window.
Optionally, the determining, by the projection value matrix, a weight of each vector in the projection value matrix includes:
determining the contribution degree of the original training data through the projection value matrix;
and determining the weight value of each vector in the projection value matrix according to the contribution degree of the original training data.
Optionally, determining the contribution degree of the original training data through the projection value matrix includes:
and calculating the classification contribution degree of each vector in a normal characteristic data projection value matrix obtained from the normal data and a fault characteristic data projection value matrix obtained from the fault data to the normal data and the fault data.
Optionally, determining the weight of each vector in the projection value matrix according to the contribution of the original training data includes:
and determining the weight value of each vector in the projection value matrix according to the classification contribution degree of the normal data and the classification contribution degree of the fault data.
Optionally, calculating the limit corresponding to each principal component includes:
establishing a principal component vector matrix by using the principal components;
and calculating the limit value corresponding to the principal component in each principal component vector matrix.
In a second aspect, an embodiment of the present invention further provides an engine operating state early warning method, including:
acquiring original data;
extracting data characteristic sample points in the original data;
standardizing the data characteristic sample points to obtain standardized data characteristic sample points;
calculating a statistical quantity value according to the standardized data characteristic sample points;
and comparing the statistical value with a control limit value, and if the statistical value exceeds the control limit value, carrying out early warning.
Optionally, the raw data is in a matrix form.
Optionally, the data feature sample points include one or more of an average frequency, a root mean square frequency, a standard deviation frequency, a kurtosis frequency, a variance, a mean square value, a root mean square value, a skewness, a kurtosis, a waveform indicator, a margin indicator, a pulse indicator, a peak indicator, or a kurtosis indicator of the original data.
In a third aspect, an embodiment of the present invention further provides an engine operating condition criterion determining apparatus, including:
the system comprises an original training data acquisition module, a failure data acquisition module and a failure data acquisition module, wherein the original training data acquisition module is used for acquiring original training data which comprises normal data and failure data;
the projection value matrix acquisition module is used for acquiring a characteristic data matrix and a space basis vector and acquiring a projection value matrix of the characteristic data matrix on the space basis vector according to the original training data;
the weight determination module is used for determining the weight of each vector in the projection value matrix through the projection value matrix;
the principal component determining module is used for arranging the weights from large to small and acquiring vectors in the first r projection value matrixes as principal components;
the limit value calculation module is used for calculating the limit value corresponding to each principal component;
and the control limit value determining module is used for selecting the limit value of a preset percentage as the control limit value.
In a fourth aspect, an embodiment of the present invention further provides an engine operating state early warning device, including:
the original data acquisition module is used for acquiring original data;
the sample point acquisition module is used for extracting data characteristic sample points in the original data;
the standardization module is used for standardizing the data characteristic sample points to obtain standardized data characteristic sample points;
the statistic value calculating module is used for calculating a statistic value according to the standardized data feature sample points;
and the early warning monitoring module is used for comparing the statistical value with a control limit value, and if the statistical value exceeds the control limit value, early warning is carried out.
In a fifth aspect, an embodiment of the present invention further provides a vehicle, including:
one or more processors;
storage means for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement the engine operating state criterion determining method of any one of the above and/or the engine operating state warning method of any one of the above.
The embodiment of the invention provides a method for determining the standard of an engine running state, which comprises the steps of firstly acquiring original training data, wherein the original training data comprises normal data and fault data; then acquiring a characteristic data matrix and a space base vector, and acquiring a projection value matrix of the characteristic data matrix on the space base vector according to the original training data; determining the weight of each vector in the projection value matrix through the projection value matrix; arranging the weights from large to small, and acquiring vectors in the first r projection value matrixes as principal components; calculating a limit value corresponding to each principal component; and selecting a limit value of a preset percentage as a control limit value. By determining the control limit value, a foundation is provided for determining the running state of the engine according to the control limit value and carrying out early warning according to the running state of the engine. By multivariate cooperative analysis of the engine operating state, the engine operating state reflected by a plurality of variables as a whole can be obtained. Therefore, the fault monitoring accuracy of the running state of the engine is improved, the time for finding the running fault of the engine is advanced, and the fault monitoring effect on the running state of the engine is more ideal.
Drawings
FIG. 1 is a flow chart of a method for determining an engine operating condition criterion according to an embodiment of the present invention;
FIG. 2 is a flowchart of an engine operating condition warning method according to an embodiment of the present invention;
fig. 3 is a schematic structural diagram of an engine operating condition criterion determining apparatus according to an embodiment of the present invention;
fig. 4 is a schematic structural diagram of an engine operating state early warning device provided in an embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Fig. 1 is a flowchart of an engine operating condition criterion determining method according to an embodiment of the present invention, referring to fig. 1. The embodiment of the invention provides a method for determining the standard of an engine running state, which comprises the following steps:
s11: raw training data is acquired, the raw training data including normal data and fault data.
The original training data is data obtained under the conditions of fixed rotating speed and fixed torque. The collected data X can be divided into data of the engine running state(n1Number of sample points, m1Representing number of variables), vibration data of key parts of engine(n2Number of sample points, m2Representing the number of variables). Similarly, the fault data collected by us are
Wherein
In the general case of n1≠n2、m1≠m2In order to use various types of variable modeling analysis, feature extraction needs to be performed on the data to eliminate the influence of the sampling frequency on the data.
S12: and acquiring a characteristic data matrix and a space basis vector, and acquiring a projection value matrix of the characteristic data matrix on the space basis vector according to the original training data.
Wherein the step may comprise the sub-steps of:
s121: extracting each sampling frequency characteristic variable of the original training data;
wherein, each sampling frequency characteristic variable of the original training data can be extracted through a sliding window. Form-to-data using sliding windowPerforming feature extraction, hereThe size of the sliding window is reasonably determined according to the actual sampling frequency so as to ensure that the extracted features meet the modeling requirements. Wherein a and b are constants.
Part of the data on the engine running stateAnd extracting the characteristics of the mean value and the variance. For engine vibration dataAnd extracting the time domain characteristics such as variance, mean square value, root mean square value, skewness, kurtosis, waveform index, margin index, pulse index, peak index and kurtosis index, and the frequency domain characteristics such as average frequency, root mean square frequency, standard deviation frequency and kurtosis frequency, wherein the calculation mode of the specific characteristics can be determined according to actual needs.
Here, the frequency refers to the frequency of the acquired raw data. The average frequency means an average value of frequencies. Root mean square frequency refers to the root mean square of the frequency. The standard deviation frequency refers to the standard deviation of the frequency. Kurtosis frequency refers to the kurtosis or kurtosis coefficient of a frequency. Variance refers to the variance of frequency. The mean square value refers to the mean square value of the frequency. The root mean square value refers to the root mean square value of the frequency. Skewness is a skewness coefficient of frequency, and indicates the direction and degree of skewness of the frequency distribution. Kurtosis is the kurtosis of frequency, and is a numerical statistic that reflects the characteristics of frequency distribution. The waveform index reflects the degree of waveform fluctuation of the frequency. The margin index reflects the error tolerance range of the frequency. The pulse index reflects the number of pulses in the frequency per unit time. The peak indicator reflects the peak magnitude of the frequency. The kurtosis index reflects the steepness of the pulses in frequency.
Obtaining a multi-feature matrix after feature extractionWherein n is3Number of sample points representing multi-feature matrix, 2m1+15m2F is a constant for the number of feature variables extracted. A multi-feature matrix for extracting fault data in the same way, namely a feature data matrix of
S122: weighting each sampling frequency characteristic variable to obtain a characteristic data matrix;
the method for determining the weight matrix of the characteristic variables by depending on the contribution degree of each characteristic variable to data classification comprises the following specific steps:
first, the multi-feature matrix of normal dataObtain mean and variance for normalization, and apply dataAndnormalization is performed. For the convenience of description will beIs marked as X1,Is marked as X2。
Wherein,is X1In the form after the standardization of the standard,is X1Column mean vector of (1), Σ-1Is X1I is a row vector whose elements are all 1, the subscript of I represents its dimension, T is the score vector, and n is a constant.
and then, calculating the weight of each characteristic variable by utilizing a Relieff algorithm, and obtaining a weight matrix W.
In order to add the weight values of the feature variables to 1, the weight values in equation (6) need to be processed, and the processing form is shown in equation (7.1). The weight matrix W is updated to the updated weight matrix W.
Will be provided withTo carry outWeighting to obtain a weighted data feature matrix ofThe calculation is as shown in equation (8) below.
S123: performing eigenvalue decomposition on the characteristic data matrix to obtain a space basis vector;
wherein, can be calculated and decomposedCovariance matrix ofA set of spatial basis vectors is obtained.
To pairDecomposing the characteristic value to obtain a group of characteristic vectors pi,i=1,…,(2m1+15m2)。
S124: and mapping the characteristic data matrix to a space base vector to obtain a projection value matrix of the characteristic data matrix on the space base vector.
Wherein the normalized normal data can be comparedProjecting the normal data onto a base vector, namely a space base vector P to obtain a projection value matrix T of the normal data on the base vector1。
Fault data after standardizationProjecting the data to a base vector P to obtain a projection value matrix T of the fault data on the base vector2。
S13: determining the weight of each vector in the projection value matrix through the projection value matrix;
wherein the step may comprise the sub-steps of:
s131: determining the contribution degree of the original training data through a projection value matrix;
among them, S1311: and calculating the classification contribution degree of each vector in the normal characteristic data projection value matrix obtained from the normal data and the fault characteristic data projection value matrix obtained from the fault data to the normal data and the fault data.
S132: and determining the weight of each vector in the projection value matrix according to the contribution degree of the original training data.
Among them, S1321: and determining the weight of each vector in the projection value matrix according to the classification contribution of the normal data and the classification contribution of the fault data.
Wherein the obtained projection value data matrix T can be used1、T2Calculating the weight e of each base vector to the classification result through a Relieff algorithmi(i=1,…(2m1+15m2) And the obtained weight is used for selecting the principal component vector of the monitoring model. The larger the weight, the more important the component.
S14: arranging the weights from large to small, and acquiring vectors in the first r projection value matrixes as principal components;
the selection rule is that the first r principal components with larger weight are selected for establishing the model, and the size of r can be determined according to actual needs.
The principal component is the column vector with higher weight e in the base vector matrix P, and the principal component vector matrix can be written as Pr=[p1 p2 … pr]。
S15: calculating a limit value corresponding to each principal component;
wherein the step may comprise the sub-steps of:
s151: establishing a principal component vector matrix by using the principal components;
s152: and calculating the limit value corresponding to the principal component in each principal component vector matrix.
Wherein the index can be monitored using T in the principal component analysis method2The form of the statistics. Combining equation (3), equation (4) and equation (8) for sample x at time iiNormalized and weighted in the form ofThe calculation form is shown in equation (15)。T of2Limit value tiSee equation (16) for calculation:
s16: and selecting a limit value of a preset percentage as a control limit value.
Wherein, Λ ═ ΣTSigma, a set of t that can be calculated by calculating the normal sample for the control limiti(i=1,…n3) Taking the value of 98% of the number of the group divided into points as the control limit value T of the modelL。
The embodiment of the invention provides a novel method for determining principal components, which is different from the traditional selection rule directly utilizing the magnitude of characteristic values, and the principal components of a monitoring model are determined according to the contribution degree of each vector to data classification. The multivariate advantage of the collected variables can be fully utilized to establish a monitoring model. The problem of inconsistent acquisition frequency among variables is solved by a mode of extracting features through a sliding window. Meanwhile, a determination method of the weight matrix is introduced and proposed, and the weight of fault reaction among different characteristics is optimized. A novel method for determining the principal components of a monitoring model is provided, and the monitoring capability of the model on faults is improved.
Fig. 2 is a flowchart of an engine operating state early warning method according to an embodiment of the present invention, and refer to fig. 2. The embodiment of the invention also provides an engine running state early warning method, which comprises the following steps:
s21: acquiring original data;
s22: extracting data characteristic sample points in original data;
s23: standardizing the data characteristic sample points to obtain standardized data characteristic sample points;
s24: calculating a statistical quantity value according to the standardized data characteristic sample points;
s25: and comparing the statistic value with the control limit value, and if the statistic value exceeds the control limit value, performing early warning.
The steps S21 to S24 are almost the same as those in the previous embodiment, and thus are not described again. Wherein S25 includes the weighted feature sample at time iT of2Statistic value tiGreater than the control limit TLAnd indicating that the running state of the engine is abnormal. And carrying out early warning on the running state of the engine. The early warning of the running state of the engine prompts a user that the engine is in an abnormal state, so that the user can overhaul the engine in advance, the reliability of a vehicle is improved, and the serious fault of the engine is avoided.
Alternatively, a statistical value of k sample points may be set as required, and exceeding the control limit value indicates that the engine is running in a fault. K can be 10-20.
In other embodiments, the raw data is in the form of a matrix.
The original data is in a matrix form, so that matrix operation is facilitated to be performed according to the original data, and convenience is provided for subsequent calculation.
In other embodiments, the data feature sample points include one or more of an average frequency, a root mean square frequency, a standard deviation frequency, a kurtosis frequency, a variance, a mean square value, a root mean square value, a skewness, a kurtosis, a waveform indicator, a margin indicator, a pulse indicator, a peak indicator, or a kurtosis indicator of the original data.
Here, the frequency refers to the frequency of the acquired raw data. The average frequency means an average value of frequencies. Root mean square frequency refers to the root mean square of the frequency. The standard deviation frequency refers to the standard deviation of the frequency. Kurtosis frequency refers to the kurtosis or kurtosis coefficient of a frequency. Variance refers to the variance of frequency. The mean square value refers to the mean square value of the frequency. The root mean square value refers to the root mean square value of the frequency. Skewness is a skewness coefficient of frequency, and indicates the direction and degree of skewness of the frequency distribution. Kurtosis is the kurtosis of frequency, and is a numerical statistic that reflects the characteristics of frequency distribution. The waveform index reflects the degree of waveform fluctuation of the frequency. The margin index reflects the error tolerance range of the frequency. The pulse index reflects the number of pulses in the frequency per unit time. The peak indicator reflects the peak magnitude of the frequency. The kurtosis index reflects the steepness of the pulses in frequency. The perception sensitivity of the running state of the engine can be further improved by acquiring the data, and the early warning effect is further improved.
Fig. 3 is a schematic structural diagram of an engine operating condition criterion determining apparatus according to an embodiment of the present invention, and refer to fig. 3. The embodiment of the invention also provides a device for determining the operating state standard of an engine, which comprises:
an original training data obtaining module 11, configured to obtain original training data, where the original training data includes normal data and fault data;
the projection value matrix obtaining module 12 is configured to obtain a feature data matrix and a space basis vector, and obtain a projection value matrix of the feature data matrix on the space basis vector according to the original training data;
a weight determination module 13, configured to determine a weight of each vector in the projection value matrix through the projection value matrix;
a principal component determining module 14, configured to arrange the weights from large to small, and obtain vectors in the first r projection value matrices as principal components;
a limit value calculation module 15, configured to calculate a limit value corresponding to each principal component;
and the control limit value determining module 16 is used for selecting a limit value of a preset percentage as a control limit value.
The engine operation state standard determining device provided by the embodiment of the invention can execute the engine operation state standard determining method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the executing method.
In other embodiments, the projection value matrix acquisition module includes:
the sampling frequency characteristic variable extraction module is used for extracting each sampling frequency characteristic variable of the original training data;
the weighting acquisition characteristic data matrix module is used for weighting each sampling frequency characteristic variable to acquire a characteristic data matrix;
the decomposition obtaining space base vector module is used for carrying out eigenvalue decomposition on the characteristic data matrix to obtain a space base vector;
and the projection value matrix module is used for mapping the characteristic data matrix to the space basis vector to obtain a projection value matrix of the characteristic data matrix on the space basis vector.
In other embodiments, the sampling frequency feature variable extraction module comprises:
and the sliding window sampling frequency characteristic variable extraction module is used for extracting each sampling frequency characteristic variable of the original training data through a sliding window.
In other embodiments, the weight determination module includes:
the module for determining the contribution degree of the original training data is used for determining the contribution degree of the original training data through a projection value matrix;
and the module for determining the weight of each vector in the projection value matrix is used for determining the weight of each vector in the projection value matrix according to the contribution degree of the original training data.
In other embodiments, the determining the contribution of the raw training data module includes:
and the normal data and fault data classification contribution degree determining module calculates the classification contribution degree of each vector in a normal characteristic data projection value matrix obtained from the normal data and a fault characteristic data projection value matrix obtained from the fault data to the normal data and the fault data.
In other embodiments, the module for determining the weight of each vector in the matrix of projection values comprises:
and the weight determining module is used for determining the weight of each vector in the projection value matrix according to the classification contribution of the normal data and the classification contribution of the fault data.
In other embodiments, the limit calculation module includes:
the principal component vector matrix establishing module is used for establishing a principal component vector matrix by using principal components;
and the corresponding limit value calculating module is used for calculating the limit value corresponding to the principal component in each principal component vector matrix.
Fig. 4 is a schematic structural diagram of an engine operating state early warning device provided by an embodiment of the present invention, and refer to fig. 4. The embodiment of the invention also provides an engine running state early warning device, which comprises:
an original data obtaining module 21, configured to obtain original data;
a sample point obtaining module 22, configured to extract data feature sample points in the original data;
the standardization module 23 is configured to standardize the data feature sample points to obtain standardized data feature sample points;
the statistical quantity calculating module 24 is used for calculating a statistical quantity according to the standardized data feature sample points;
and the early warning monitoring module 25 is used for comparing the statistical value with the control limit value, and if the statistical value exceeds the control limit value, early warning is carried out.
The engine running state early warning device provided by the embodiment of the invention can execute the engine running state early warning method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.
An embodiment of the present invention further provides a vehicle, including:
one or more processors;
storage means for storing one or more programs;
when executed by one or more processors, the one or more programs cause the one or more processors to implement any of the engine operating state criterion determining methods described above and/or any of the engine operating state warning methods described above.
The vehicle provided by the embodiment of the invention can execute the engine operation state standard determining method provided by any embodiment of the invention and/or the engine operation state early warning method provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the executing method.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious modifications, rearrangements, combinations and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (13)
1. An engine operating condition criterion determining method, comprising:
acquiring original training data, wherein the original training data comprises normal data and fault data;
acquiring a characteristic data matrix and a space basis vector, and acquiring a projection value matrix of the characteristic data matrix on the space basis vector according to the original training data;
determining the weight of each vector in the projection value matrix through the projection value matrix;
arranging the weights from large to small, and acquiring vectors in the first r projection value matrixes as principal components;
calculating a limit value corresponding to each principal component;
and selecting the limit value of the preset percentage as a control limit value.
2. The method of claim 1, wherein obtaining a feature data matrix and a spatial basis vector, and obtaining a projection value matrix of the feature data matrix on the spatial basis vector from the original training data comprises:
extracting each sampling frequency characteristic variable of the original training data;
weighting each sampling frequency characteristic variable to obtain a characteristic data matrix;
performing eigenvalue decomposition on the characteristic data matrix to obtain a space basis vector;
and mapping the characteristic data matrix to the space base vector to obtain a projection value matrix of the characteristic data matrix on the space base vector.
3. The criteria determination method of claim 2 wherein extracting respective sampling frequency characteristic variables of the raw training data comprises:
and extracting each sampling frequency characteristic variable of the original training data through a sliding window.
4. The method of claim 1, wherein the determining the weight value of each vector in the projection value matrix through the projection value matrix comprises:
determining the contribution degree of the original training data through the projection value matrix;
and determining the weight value of each vector in the projection value matrix according to the contribution degree of the original training data.
5. The criteria determination method of claim 4 wherein determining the degree of contribution of the raw training data from the matrix of projection values comprises:
and calculating the classification contribution degree of each vector in a normal characteristic data projection value matrix obtained from the normal data and a fault characteristic data projection value matrix obtained from the fault data to the normal data and the fault data.
6. The criteria determination method of claim 4 wherein determining the weight for each vector in the matrix of projection values based on the contribution of the original training data comprises:
and determining the weight value of each vector in the projection value matrix according to the classification contribution degree of the normal data and the classification contribution degree of the fault data.
7. The method of claim 1, wherein calculating the limit value for each of the principal components comprises:
establishing a principal component vector matrix by using the principal components;
and calculating the limit value corresponding to the principal component in each principal component vector matrix.
8. An engine operating state early warning method is characterized by comprising the following steps:
acquiring original data;
extracting data characteristic sample points in the original data;
standardizing the data characteristic sample points to obtain standardized data characteristic sample points;
calculating a statistical quantity value according to the standardized data characteristic sample points;
and comparing the statistical value with a control limit value, and if the statistical value exceeds the control limit value, carrying out early warning.
9. The warning method of claim 8 wherein the raw data is in the form of a matrix.
10. The warning method of claim 8, wherein the data feature sample points include one or more of a mean frequency, a root mean square frequency, a standard deviation frequency, a kurtosis frequency, a variance, a mean square value, a root mean square value, a skewness, a kurtosis, a waveform indicator, a margin indicator, a pulse indicator, a peak indicator, or a kurtosis indicator of the original data.
11. An engine operating condition criterion determining apparatus, characterized by comprising:
the system comprises an original training data acquisition module, a failure data acquisition module and a failure data acquisition module, wherein the original training data acquisition module is used for acquiring original training data which comprises normal data and failure data;
the projection value matrix acquisition module is used for acquiring a characteristic data matrix and a space basis vector and acquiring a projection value matrix of the characteristic data matrix on the space basis vector according to the original training data;
the weight determination module is used for determining the weight of each vector in the projection value matrix through the projection value matrix;
the principal component determining module is used for arranging the weights from large to small and acquiring vectors in the first r projection value matrixes as principal components;
the limit value calculation module is used for calculating the limit value corresponding to each principal component;
and the control limit value determining module is used for selecting the limit value of a preset percentage as the control limit value.
12. An engine operating condition early warning device, characterized by comprising:
the original data acquisition module is used for acquiring original data;
the sample point acquisition module is used for extracting data characteristic sample points in the original data;
the standardization module is used for standardizing the data characteristic sample points to obtain standardized data characteristic sample points;
the statistic value calculating module is used for calculating a statistic value according to the standardized data feature sample points;
and the early warning monitoring module is used for comparing the statistical value with a control limit value, and if the statistical value exceeds the control limit value, early warning is carried out.
13. A vehicle, characterized by comprising:
one or more processors;
storage means for storing one or more programs;
when executed by the one or more processors, cause the one or more processors to implement the engine operating state criterion determining method of any one of claims 1-7 and/or the engine operating state forewarning method of any one of claims 8-10.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111168871.0A CN113915015A (en) | 2021-09-30 | 2021-09-30 | Engine running state standard determining method, early warning method and device and vehicle |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202111168871.0A CN113915015A (en) | 2021-09-30 | 2021-09-30 | Engine running state standard determining method, early warning method and device and vehicle |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113915015A true CN113915015A (en) | 2022-01-11 |
Family
ID=79238017
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202111168871.0A Pending CN113915015A (en) | 2021-09-30 | 2021-09-30 | Engine running state standard determining method, early warning method and device and vehicle |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113915015A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114689326A (en) * | 2022-03-24 | 2022-07-01 | 潍柴动力股份有限公司 | Engine detection method and detection device |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070124113A1 (en) * | 2005-11-28 | 2007-05-31 | Honeywell International, Inc. | Fault detection system and method using multiway principal component analysis |
CN110850839A (en) * | 2018-08-21 | 2020-02-28 | 上海交通大学 | Real-time monitoring control system for energy network |
-
2021
- 2021-09-30 CN CN202111168871.0A patent/CN113915015A/en active Pending
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20070124113A1 (en) * | 2005-11-28 | 2007-05-31 | Honeywell International, Inc. | Fault detection system and method using multiway principal component analysis |
CN110850839A (en) * | 2018-08-21 | 2020-02-28 | 上海交通大学 | Real-time monitoring control system for energy network |
Non-Patent Citations (2)
Title |
---|
蓝艇等: "变量加权型主元分析算法及其在故障检测中的应用", 化工学报, vol. 68, no. 8, pages 3177 - 3182 * |
陶阳: "基于ReliefF的主元挑选算法在过程监控中的应用", 化工学报, vol. 68, no. 4, pages 1525 - 1532 * |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114689326A (en) * | 2022-03-24 | 2022-07-01 | 潍柴动力股份有限公司 | Engine detection method and detection device |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107632592B (en) | Nonlinear time-varying process fault monitoring method based on efficient recursion kernel principal component analysis | |
CN108062565B (en) | Double-principal element-dynamic core principal element analysis fault diagnosis method based on chemical engineering TE process | |
CN102736546A (en) | State monitoring device of complex electromechanical system for flow industry and method | |
CN108549908B (en) | Chemical process fault detection method based on multi-sampling probability kernel principal component model | |
CN105259895A (en) | Method and monitoring system for detecting and separating micro fault in industrial process | |
CN104914847A (en) | Industrial process fault diagnosis method based on direction kernel partial least square | |
CN114062850B (en) | Double-threshold power grid early fault detection method | |
CN111680725A (en) | Gas sensor array multi-fault isolation algorithm based on reconstruction contribution | |
CN113391244B (en) | VMD-based transformer closing vibration signal characteristic frequency calculation method | |
CN106874676B (en) | State evaluation method for electric energy metering device | |
CN114237046B (en) | Partial discharge pattern recognition method based on SIFT data feature extraction algorithm and BP neural network model | |
CN110245460A (en) | A kind of batch process fault monitoring method based on multistage OICA | |
CN111796233A (en) | Method for evaluating secondary errors of multiple voltage transformers in double-bus connection mode | |
CN111340367B (en) | Oil-shortage equipment state evaluation method and system based on multi-dimensional relevance data analysis | |
CN105137354A (en) | Motor fault detection method based on nerve network | |
CN113915015A (en) | Engine running state standard determining method, early warning method and device and vehicle | |
CN115495924A (en) | MOSFET service life prediction method based on ARIMA model | |
CN117458955A (en) | Operation control method and system of motor | |
CN114263621B (en) | Test method and system for centrifugal pump cavitation fault diagnosis simulation | |
CN110751217A (en) | Equipment energy consumption ratio early warning analysis method based on principal component analysis | |
CN117150244B (en) | Intelligent power distribution cabinet state monitoring method and system based on electrical parameter analysis | |
CN116627116B (en) | Process industry fault positioning method and system and electronic equipment | |
CN108170648A (en) | A kind of nongausian process monitoring method returned based on given data | |
CN116305733B (en) | Quality-related fault detection method based on global and local feature extraction | |
CN107817784B (en) | A kind of procedure failure testing method based on concurrent offset minimum binary |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20220111 |
|
RJ01 | Rejection of invention patent application after publication |