Diesel engine lubrication system fault diagnosis method based on improved grey correlation analysis

Abstract

The invention relates to the field of diesel engine fault diagnosis, and provides a diesel engine lubrication system fault diagnosis method based on improved grey correlation analysis, and the method comprises the steps: S1, collecting the operation data of a marine diesel engine lubrication system in different states, and carrying out the initial value processing, and obtaining the initial value data; s2, calculating reference feature vectors and to-be-detected state feature vectors of different operation states based on the initial valued data; s3, calculating an improved grey correlation degree between the reference feature vector and the to-be-detected state feature vector by using an improved grey correlation analysis method; and S4, sorting the gray correlation degree values, and classifying the to-be-detected state into an operation state corresponding to the minimum improved gray correlation degree based on a sorting result, so as to realize fault diagnosis of the diesel engine lubrication system. According to the method, the problem that the distance feature between the to-be-detected feature vector and the reference feature vector of the lubricating system cannot be fully utilized in traditional grey correlation analysis is solved.

Description

Technical Field

[0001] The invention relates to the technical field of diesel engine fault diagnosis, and in particular to a diesel engine lubrication system fault diagnosis method based on improved grey correlation analysis. Background Art

[0002] The lubrication system of a marine diesel engine is responsible for providing clean lubricating oil with appropriate temperature to the friction parts, maintaining the liquid friction between the two relatively moving parts, reducing the wear of the parts, reducing the power consumed in friction, and taking away the heat and metal chips generated by friction, lubricating the friction parts in the diesel engine, which is of great significance to ensure the normal and reliable operation of the diesel engine and extend the life of the diesel engine. The diesel engine lubrication system mainly includes equipment such as lubricating oil pumps, filters, and coolers. When the diesel engine lubrication system fails, it cannot provide clean lubricating oil with the temperature and pressure that meet the standards, which will lead to the failure to establish a lubricating oil film between the moving parts of the diesel engine, which is extremely harmful to the operation safety of the marine diesel engine. In severe cases, it will lead to serious consequences such as cylinder pulling and shaft seizure. Therefore, it is necessary to diagnose the fault of the marine diesel engine lubrication system, find the abnormal state of the lubrication system, accurately locate the source of the fault, and perform timely maintenance on the diesel engine lubrication system.

[0003] Grey correlation analysis is a method used in grey system theory to analyze the correlation and mutual influence between multiple variables. It evaluates the correlation between variables by comparing and analyzing the consistency of the variables in the process of development and change. In recent years, some scholars have applied grey correlation analysis to the fault diagnosis task of mechanical equipment, which has improved the accuracy of fault diagnosis to a certain extent. However, most existing studies focus on directly applying grey correlation analysis to fault diagnosis tasks. By calculating the grey correlation between the fault mode to be detected and the reference fault mode, the fault mode to be detected is identified. The public document "Automatic diagnosis of electromechanical equipment faults by determining feature weights with grey correlation analysis" (Automation Technology and Applications, 2021) applies grey correlation analysis to diesel engine fault identification. The public document states: "The feature weights are determined by grey correlation analysis, so as to complete the automatic diagnosis of electromechanical equipment faults. Finally, a specific diesel engine was used as the research object of fault diagnosis, and a simulation comparison test was carried out. The experimental results show that the diagnostic results of this method are the same as the actual fault output, and the diagnostic results are effective." Some researchers have applied the improved grey correlation analysis to fault diagnosis tasks. The public document "Fault Diagnosis of Wind Turbine Bearings Based on Morphological Multifractals" (Vibration, Testing and Diagnosis, 2021) proposes an improved grey correlation analysis method. The public document states: "In the process of calculating the grey correlation degree, the deviation maximization is used to weight different features. This weight value reflects the influence of each characteristic index on the overall correlation degree, making the grey correlation result more accurate, and the description of the relationship between the feature vector to be identified and the reference feature vector more balanced and objective, thereby improving the accuracy of fault diagnosis". The above two solve the fault diagnosis problem in the field of mechanical equipment through grey correlation analysis, but both have shortcomings: there are feature vectors with high trend similarity and approximately parallel in the feature vectors to be detected and each reference feature vector. Traditional grey correlation analysis can only analyze the closeness of the connection between different feature vectors from the trend similarity, but cannot make full use of the distance characteristics between different feature vectors to be detected. Although the improved grey correlation analysis increases the weight of important characteristic indicators and achieves good diagnostic results, it is easy to affect the diagnostic effect when the feature vector to be detected fluctuates greatly and extreme values ​​appear, and it also fails to make full use of the distance characteristics between feature vectors. Summary of the invention

[0004] The present invention provides a method for diagnosing faults of a marine diesel engine lubrication system based on improved grey correlation analysis. First, the operating data of the marine diesel engine lubrication system in various states are initialized, and then the reference characteristic vector of the diesel engine lubrication system operating state is calculated; then, the improved grey correlation analysis method is used to calculate the improved grey correlation between the characteristic vector of the lubrication system state to be checked and the reference characteristic vector of each state; then, each improved grey correlation is sorted, and the state to be checked is classified as the reference operating state corresponding to the maximum improved grey correlation; finally, the effectiveness of the method is verified by examples. The present invention is specifically implemented by the following technical solutions:

[0005] Step S1, collecting the operating data of the ship diesel engine lubrication system under different states and performing initialization processing to obtain initialization data;

[0006] Step S2, calculating reference feature vectors of different operating states and feature vectors of states to be tested based on the initialization data;

[0007] Step S3, using an improved grey relational analysis method to calculate an improved grey relational degree between the reference feature vector and the feature vector of the state to be detected;

[0008] Step S4, sorting the improved gray correlation values, and classifying the state to be inspected as the operating state corresponding to the minimum improved gray correlation value based on the sorting result, so as to realize the fault diagnosis of the diesel engine lubrication system.

[0009] Optionally, in step S1, the operating state includes a no-fault state, a filter blockage fault state, a bypass valve leakage fault state, a cooler fouling fault state and a filter leakage fault state.

[0010] Optionally, in step S1, the content of the initialization process specifically includes:

[0011]

[0012] Among them, f i represents the i-th running state, i∈[1,5], s j represents the jth operating data monitoring point, j = [1,9], xi,j(k) represents the time when the diesel engine lubrication system is at the fth i When the monitoring point s j The kth data collected; x1,j(1) is the first row of data samples when the diesel engine lubrication system is in a fault-free state, x i ' ,j Indicates that the diesel engine lubrication system is in operation i In each operating state, the detection point s j The normalized value of the kth data collected.

[0013] Optionally, the operating data monitoring points include the lubricating oil temperature after the lubricating oil pump, the lubricating oil temperature before the filter, the lubricating oil temperature after the filter, the cooling water outlet temperature of the cooler, the lubricating oil pressure after the lubricating oil pump, the lubricating oil pressure before the filter, the lubricating oil pressure after the filter, the lubricating oil inlet pressure and the cooling water outlet pressure of the cooler.

[0014] Optionally, in step S2, calculating reference feature vectors of different operating states based on the initialization data specifically includes:

[0015]

[0016] Where, t is the number of data samples collected under each diesel engine lubrication system operating state; v i (j) indicates the running state f i The running state reference feature vector v i The value of the j-th dimension.

[0017] Optionally, in step S3, the process of the improved grey relational analysis method specifically includes:

[0018] Constructing a state feature matrix to be checked based on the reference feature vector and the state feature vector to be checked;

[0019] Select a reference sequence from the state feature matrix to be tested, remove the reference sequence from the state feature matrix to be tested, and use the remaining row vector as a comparison sequence, and calculate the grey correlation coefficient between the reference sequence and the comparison sequence;

[0020] Calculate the grey correlation degree between the reference feature vector and the feature vector of the state to be tested based on the grey correlation coefficient;

[0021] Calculate the average Euclidean distance between the reference feature vector and the feature vector of the state to be tested;

[0022] An improved grey relational degree is calculated based on the grey relational degree and the average Euclidean distance.

[0023] Optionally, the grey correlation coefficient is used to calculate the grey correlation degree r between the reference feature vector and the feature vector of the state to be tested. i Specifically include:

[0024]

[0025] Among them, i (j) represents the grey correlation coefficient, and n represents the number of monitoring points.

[0026] Optionally, the content of calculating the average Euclidean distance between the reference feature vector and the feature vector of the state to be checked specifically includes:

[0027]

[0028] Among them, d i represents the average Euclidean distance, v p is the feature vector of the state to be tested.

[0029] Compared with the prior art, the present invention has the following beneficial effects:

[0030] 1. Make full use of the distance characteristics between the feature vector to be tested and the reference feature vector of the diesel engine lubrication system to reduce the adverse effects of the extreme value of the feature vector to be tested on the fault diagnosis effect;

[0031] 2. A ship diesel engine lubrication system fault diagnosis method based on improved grey correlation analysis is provided for ship diesel engine lubrication system fault diagnosis, which can comprehensively consider the trend similarity and distance characteristics between the feature vector to be tested and the reference feature vector of the lubrication system. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] In order to more clearly illustrate the technical solution of the present invention, the following briefly introduces the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0033] Figure 1 This is a flow chart of fault diagnosis of a marine diesel engine lubrication system based on improved grey correlation analysis according to an embodiment of the present invention;

[0034] Figure 2 This is a line graph of reference characteristic vectors of the operating status of a marine diesel engine lubrication system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0035] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0036] Example

[0037] A fault diagnosis method for lubrication system of marine diesel engine based on improved grey relational analysis Figure 1 As shown, the method includes:

[0038] Step S1, collecting the operating data of the ship diesel engine lubrication system under different states and performing initialization processing to obtain initialization data.

[0039] The diesel engine lubrication system operating state F includes five operating states, namely m=5, including: no fault state f1, filter blockage fault state f2, bypass valve leakage fault state f3, cooler fouling fault state f4, filter leakage fault state f5.

[0040] The diesel engine lubrication system operation data monitoring point S includes 9 monitoring points, that is, n=9, including: lubricating oil temperature after the lubricating oil pump s1, lubricating oil temperature before the filter s2, lubricating oil temperature after the filter s3, cooler cooling water outlet temperature s4, lubricating oil pressure after the lubricating oil pump s5, lubricating oil pressure before the filter s6, lubricating oil pressure after the filter s7, lubricating oil inlet pressure s8, cooler cooling water outlet pressure s9.

[0041] Diesel engine lubrication system operating status i The following data collection format is shown in Table 1:

[0042] Table 1

[0043] No. <![CDATA[s1]]> <![CDATA[s2]]> ... <![CDATA[s n ]]> 1 <![CDATA[x i，1 (1)]]> <![CDATA[x i，2 (1)]]> ... <![CDATA[x i，n (1)]]> 2 <![CDATA[x i，1 (2)]]> <![CDATA[x i，2 (2)]]> ... <![CDATA[x i，n (2)]]> ... ... ... ... ... t <![CDATA[x i，1 (t)]]> <![CDATA[x i，2 (t)]]> ... <![CDATA[x i，n (t)]]>

[0044] The initialization formula is:

[0045]

[0046] where i∈[1,5], j=[1,9]; x i，j (k) represents the state when the diesel engine lubrication system is in operation i When the monitoring point s j The kth data collected; x 1，j (1) is the first row of data samples when the diesel engine lubrication system is in a fault-free state, which is the initial value.

[0047] The data values ​​after normalization (initialization) are shown in Table 2:

[0048] Table 2

[0049] No. <![CDATA[s1]]> <![CDATA[s2]]> ... <![CDATA[s n ]]> 1 <![CDATA[x′ i1 (1)]]> <![CDATA[x′ i，2 (1)]]> ... <![CDATA[x′ i，n (1)]]> 2 <![CDATA[x′ i，1 (2)]]> <![CDATA[x′ i，2 (2)]]> ... <![CDATA[x′ i，n (2)]]> ... ... ... ... ... t <![CDATA[x′ i，1 (t)]]> <![CDATA[x′ i，2 (t)]]> ... <![CDATA[x′ i，n (t)]]>

[0050] Step S2: Calculate reference feature vectors of different operating states and feature vectors of states to be tested based on the initialization data.

[0051] Taking the operating state as an example, the operating state reference characteristic vector of the diesel engine lubrication system under the operating state is calculated according to the first tenth of the normalized data in Table 2. The calculation formula of the operating state reference characteristic vector of the diesel engine lubrication system is:

[0052]

[0053] Among them, in each diesel engine lubrication system operating state v i The following collection is t data samples; vi (j) indicates the running state f i In this case, the operating status refers to the value of the j-th dimension of the feature vector.

[0054] The reference characteristic vector v1 of the lubrication system fault-free state, the reference characteristic vector v2 of the filter blockage fault state, the reference characteristic vector v3 of the bypass valve leakage fault state, the reference characteristic vector v4 of the cooler fouling fault state, and the reference characteristic vector v5 of the filter leakage fault state are obtained.

[0055] The reference characteristic matrix V of the diesel engine lubrication system is:

[0056]

[0057] Step S3: using an improved grey relational analysis method to calculate an improved grey relational degree between the reference feature vector and the feature vector of the state to be detected.

[0058] The improved grey relational degree γ calculation method in step S3 is:

[0059] (1) Construct the characteristic matrix of the diesel engine lubrication system to be inspected.

[0060] The remaining samples in Table 2 are used as v p , combined with the diesel engine lubrication system reference feature matrix V to construct the state feature matrix V to be tested p :

[0061]

[0062] (2) Calculate the grey correlation between the reference feature vectors of each operating state of the lubrication system and the feature vectors of the state to be tested.

[0063] V p The last line v p is the reference sequence, and the other row vectors are the comparison sequence. Then the grey correlation coefficient ζ between the reference sequence and the corresponding dimension j in the comparison sequence is i (j) The calculation formula is as follows:

[0064]

[0065] Among them, Δ min =min i min j |v i (j)-v p (j)| and Δ max =max i max j |v i (j)-v p (j)| are the minimum difference between the two levels and the maximum difference between the two levels, Δ=|vi (j)-v p (j)|, ρ=0.5 is the resolution coefficient.

[0066] The reference characteristic vector v of the diesel engine lubrication system operating state is i and the state feature vector v p The grey relational degree r i The calculation formula is as follows:

[0067]

[0068] Where n represents the number of monitoring points.

[0069] (3) Calculate the average Euclidean distance between the reference feature vector of each operating state of the lubrication system and the feature vector of the state to be tested.

[0070] The reference characteristic vector of the diesel engine lubrication system operating state is v i , the feature vector of the state to be tested is v p , the average Euclidean distance d between the two i The calculation formula is as follows:

[0071]

[0072] (4) Calculate the improved grey correlation between the reference feature vectors of each operating state of the lubrication system and the feature vectors of the state to be tested.

[0073] The reference characteristic vector of the diesel engine lubrication system operating state is v i , the feature vector of the state to be tested is v p , the improved grey relational degree of two eigenvectors γ i The calculation formula is:

[0074] γ i =ω1(1-r i )+ω2d i

[0075] Among them, ω1 and ω2 are the grey relational degree r i The average Euclidean distance d i The weight is ω1=ω2=0.5.

[0076] The improved grey correlation degree {γ1,γ2,γ3,γ4,γ5} between the reference feature vectors of each operating state of the diesel engine lubrication system and the feature vectors of the state to be tested is obtained. The improved grey correlation degree γ i The smaller it is, the higher the trend similarity between the reference feature vector of the operating state and the feature vector of the state to be checked is, the closer the distance between the two feature vectors is, and the state to be checked is classified as the operating state corresponding to the minimum improved grey correlation degree.

[0077] Step S4, sorting the improved gray correlation values, and classifying the state to be inspected as the operating state corresponding to the minimum improved gray correlation value based on the sorting result, so as to realize the fault diagnosis of the diesel engine lubrication system.

[0078] The improved grey relational degree {γ1,γ2,γ3,γ4,γ5} is sorted according to the numerical value. For example, the sorted result is γ3<γ1<γ2<γ5<γ4, and the inspection state of the diesel engine lubrication system is determined to be fault f3.

[0079] The verification process is as follows:

[0080] (1) Collect the operating data of the lubrication system of 396 diesel engines in the fault-free state and fault state at 50% load, randomly select some data samples as the data samples to be tested, and the remaining data samples are used to calculate the reference feature vector of the operating state. (2) Take the first row of data samples of the diesel engine lubrication system in the fault-free state as the initial value x 1，j (1) Initialize the data sample to be tested and the remaining data samples. (3) Calculate the reference feature vectors {v1, v2, ..., v1, v2, ..., v2, v1, ..., v2 ... m}like Figure 2 As shown, the reference characteristic vector values ​​of each operating state are shown in Table 3, and the characteristic vector of the state to be tested is shown in Table 4.

[0081] Table 3

[0082] <![CDATA[s1]]> <![CDATA[s2]]> <![CDATA[s3]]> <![CDATA[s4]]> <![CDATA[s5]]> <![CDATA[s6]]> <![CDATA[s7]]> <![CDATA[s8]]> <![CDATA[s9]]> <![CDATA[v1]]> 0.9993 0.9990 1.0002 1.0004 1.0009 1.0087 1.0052 1.0141 0.9887 <![CDATA[v2]]> 1.0127 1.0103 1.0102 1.0078 1.0316 1.0565 0.9879 1.0061 0.8142 <![CDATA[v3]]> 1.0069 1.007 1.0035 1.0123 0.9494 0.9492 0.9855 1.0045 1.0129 <![CDATA[v4]]> 1.0440 1.0471 1.0485 1.0086 0.9883 1.0034 0.9978 1.0071 1.1236 <![CDATA[v5]]> 1.0080 1.0074 1.0084 1.0072 0.9939 1.0043 1.0011 1.0106 0.8147

[0083] Table 4

[0084] <![CDATA[s1]]> <![CDATA[s2]]> <![CDATA[s3]]> <![CDATA[s4]]> <![CDATA[s5]]> <![CDATA[s6]]> <![CDATA[s7]]> <![CDATA[s8]]> <![CDATA[s9]]> Reality <![CDATA[v p1 ]]> 0.9989 0.9982 0.9979 0.9996 0.9958 1.0063 1.0068 1.0360 0.9781 <![CDATA[f1]]> <![CDATA[v p2 ]]> 1.0004 0.9982 1.0006 1.0011 1.0036 1.0156 1.0078 1.0170 1.0196 <![CDATA[f1]]> <![CDATA[v p3 ]]> 1.0133 1.0100 1.0092 1.0086 1.0371 1.0640 0.9984 0.9962 0.8225 <![CDATA[f2]]> <![CDATA[v p4 ]]> 1.0130 1.0109 1.0114 1.0084 1.0282 1.0551 0.9856 0.9962 0.8066 <![CDATA[f2]]> <![CDATA[v p5 ]]> 1.0061 1.0069 1.0042 1.0118 0.9464 0.9452 0.9886 1.0068 0.9995 <![CDATA[f3]]> <![CDATA[v p6 ]]> 1.0073 1.0063 1.0026 1.0131 0.9473 0.9420 0.9847 1.0088 1.0217 <![CDATA[f3]]> <![CDATA[v p7 ]]> 1.0447 1.0474 1.0474 1.0072 0.9894 1.0046 0.9980 1.0150 1.1134 <![CDATA[f4]]> <![CDATA[v p8 ]]> 1.0431 1.0480 1.0501 1.0090 0.9795 1.0036 0.9941 1.0254 1.1479 <![CDATA[f4]]> <![CDATA[v p9 ]]> 1.0079 1.0087 1.0093 1.0083 0.9809 1.0120 1.0018 1.0277 0.8011 <![CDATA[f5]]> <![CDATA[v p10 ]]> 1.0079 1.0075 1.0087 1.0060 1.0002 1.0130 1.0096 1.0099 0.8328 <![CDATA[f5]]>

[0085] (4) Calculate the diesel engine lubrication system state feature vector vp and compare it with the reference feature vectors {v1, v2, ..., v m The improved grey relational degree {γ1,γ2,...,γ m The improved grey correlation degree and fault diagnosis results are shown in Table 5. It can be seen from Table 5 that the diesel engine lubrication system fault diagnosis effect based on the improved grey correlation analysis method is good, and the fault isolation rate in the current embodiment is 100%.

[0086] Table 5

[0087] Waiting for inspection <![CDATA[γ1]]> <![CDATA[γ2]]> <![CDATA[γ3]]> <![CDATA[γ4]]> <![CDATA[γ5]]> Diagnosis Reality <![CDATA[v p1 ]]> 0.0282 0.1128 0.1388 0.1378 0.0878 <![CDATA[f1]]> <![CDATA[f1]]> <![CDATA[v p2 ]]> 0.0241 0.0801 0.1143 0.1109 0.0809 <![CDATA[f1]]> <![CDATA[f1]]> <![CDATA[v p3 ]]> 0.0877 0.0166 0.1028 0.1201 0.0446 <![CDATA[f2]]> <![CDATA[f2]]> <![CDATA[v p4 ]]> 0.0876 0.0099 0.0959 0.1178 0.0436 <![CDATA[f2]]> <![CDATA[f2]]> <![CDATA[v p5 ]]> 0.0829 0.1191 0.0163 0.1374 0.1038 <![CDATA[f3]]> <![CDATA[f3]]> <![CDATA[v p6 ]]> 0.0862 0.1195 0.0139 0.1287 0.1028 <![CDATA[f3]]> <![CDATA[f3]]> <![CDATA[v p7 ]]> 0.0842 0.1197 0.1023 0.0093 0.0926 <![CDATA[f4]]> <![CDATA[f4]]> <![CDATA[v p8 ]]> 0.0918 0.1206 0.0992 0.0196 0.0996 <![CDATA[f4]]> <![CDATA[f4]]> <![CDATA[v p9 ]]> 0.0647 0.0472 0.0850 0.1001 0.0192 <![CDATA[f5]]> <![CDATA[f5]]> <![CDATA[v p10 ]]> 0.0538 0.0451 0.0878 0.1024 0.0168 <![CDATA[f5]]> <![CDATA[f5]]>

[0088] The embodiments described above are only descriptions of the preferred embodiments of the present invention and are not intended to limit the scope of the present invention. Without departing from the design spirit of the present invention, various modifications and improvements made to the technical solutions of the present invention by ordinary technicians in this field should all fall within the protection scope determined by the claims of the present invention.

Claims

1. A diesel engine lubrication system fault diagnosis method based on improved grey correlation analysis, characterized in that: Methods include: Step S1, collecting the operating data of the ship diesel engine lubrication system under different states and performing initialization processing to obtain initialization data; Step S2, calculating reference feature vectors of different operating states and feature vectors of states to be tested based on the initialization data; Step S3, using an improved grey relational analysis method to calculate an improved grey relational degree between the reference feature vector and the feature vector of the state to be detected; Step S4, sorting the improved gray correlation values, and classifying the state to be inspected as the operating state corresponding to the minimum improved gray correlation value based on the sorting result, so as to realize the fault diagnosis of the diesel engine lubrication system.

2. The diesel engine lubrication system fault diagnosis method based on improved grey relational analysis according to claim 1 is characterized in that: In the step S1, the operating state includes a no-fault state, a filter blockage fault state, a bypass valve leakage fault state, a cooler fouling fault state and a filter leakage fault state.

3. The diesel engine lubrication system fault diagnosis method based on improved grey relational analysis according to claim 2 is characterized in that: In step S1, the initialization process specifically includes: Among them, f i represents the i-th running state, i∈[1,5], s j represents the jth operating data monitoring point, j = [1,9], xi,j(k) represents the time when the diesel engine lubrication system is at the fth i When the monitoring point s j The kth data collected; x1,j(1) is the first row of data samples when the diesel engine lubrication system is in a fault-free state, x i ' ,j Indicates that the diesel engine lubrication system is in operation i In each operating state, the detection point s j The normalized value of the kth data collected.

4. The diesel engine lubrication system fault diagnosis method based on improved grey relational analysis according to claim 3 is characterized in that: The operating data monitoring points include the lubricating oil temperature after the lubricating oil pump, the lubricating oil temperature before the filter, the lubricating oil temperature after the filter, the cooling water outlet temperature of the cooler, the lubricating oil pressure after the lubricating oil pump, the lubricating oil pressure before the filter, the lubricating oil pressure after the filter, the lubricating oil inlet pressure and the cooling water outlet pressure of the cooler.

5. The diesel engine lubrication system fault diagnosis method based on improved grey relational analysis according to claim 3 is characterized in that: In the step S2, calculating the reference feature vectors of different operating states based on the initialization data specifically includes: Where, t is the number of data samples collected under each diesel engine lubrication system operating state; v i (j) indicates the running state f i The running state reference feature vector v i The value of the j-th dimension.

6. The diesel engine lubrication system fault diagnosis method based on improved grey relational analysis according to claim 5 is characterized in that: In step S3, the process of the improved grey relational analysis method specifically includes: Constructing a state feature matrix to be checked based on the reference feature vector and the state feature vector to be checked; Select a reference sequence from the state feature matrix to be tested, remove the reference sequence from the state feature matrix to be tested, and use the remaining row vector as a comparison sequence, and calculate the grey correlation coefficient between the reference sequence and the comparison sequence; Calculate the grey correlation degree between the reference feature vector and the feature vector of the state to be tested based on the grey correlation coefficient; Calculate the average Euclidean distance between the reference feature vector and the feature vector of the state to be tested; An improved grey relational degree is calculated based on the grey relational degree and the average Euclidean distance.

7. The diesel engine lubrication system fault diagnosis method based on improved grey relational analysis according to claim 6 is characterized in that: The grey relational degree r between the reference feature vector and the feature vector of the state to be tested is calculated based on the grey relational coefficient i Specifically include: Among them, i (j) represents the grey correlation coefficient, and n represents the number of monitoring points.

8. The diesel engine lubrication system fault diagnosis method based on improved grey relational analysis according to claim 7, characterized in that: The content of calculating the average Euclidean distance between the reference feature vector and the feature vector of the state to be tested specifically includes: Among them, d i represents the average Euclidean distance, v p is the feature vector of the state to be tested.

Images