Diesel engine multivariate state parameter correlation analysis method and system based on hierarchical CCA

Abstract

The invention provides a hierarchical CCA-based diesel engine multivariate state parameter association analysis method and system. The method comprises the following steps: constructing a multi-dimensional time sequence matrix according to a diesel engine state type and each measuring point state parameter; dividing a multi-dimensional time sequence matrix into a main parameter state matrix and other parameter state matrixes according to whether a measuring point exceeds the limit or not; performing hierarchical decomposition on each measuring point state parameter in the state matrix to obtain a main parameter hierarchical state matrix and other parameter hierarchical state matrixes; carrying out canonical correlation analysis on the main parameter hierarchical state matrix and other parameter hierarchical state matrixes to obtain an optimal projection vector; calculating the maximum correlation according to the main parameter hierarchical state matrix, the other parameter hierarchical state matrix and the optimal projection vector; and taking the measuring points corresponding to other parameter hierarchical state matrixes with the maximum correlation values as associated measuring points. Therefore, fault analysis is carried out based on the accurate associated measuring points in combination with the main measuring point, and the efficiency and accuracy of fault diagnosis are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a method and system for correlation analysis of multivariate state parameters of a diesel engine based on hierarchical CCA. Background Art

[0002] At present, the diagnosis of diesel engine mechanical failures generally involves selecting possible fault causes that are consistent with the changing trends of characteristic parameters such as thermal, oil, vibration, and cylinder pressure when a fault occurs, based on the deviations of the characteristic parameters from the baseline range and their changing trends, from the knowledge base to achieve fault diagnosis.

[0003] For complex situations where it is difficult to locate the fault through a single measurement point, it is necessary to gradually narrow down the suspected fault set by combing the fault tree and introducing related measurement points. For example, with the help of some auxiliary decision suggestions from the diesel engine expert knowledge base, the cylinder fault can be located and the scope can be narrowed down by integrating vibration monitoring signals and thermal parameters.

[0004] Among them, in traditional fault tree analysis, relevant auxiliary decision suggestions can only be given when the status parameters of the main measuring point and the status parameters of the associated measuring points exceed the alarm limit. However, in most cases, the status parameters of the associated measuring points never exceed the alarm limit, making it difficult to give effective decision suggestions.

[0005] In order to alleviate the problem of inaccurate positioning of associated measuring points, the associated measuring points can be determined by analyzing the correlation between the state parameters of the main measuring point and the state parameters of other measuring points, thereby realizing rapid fault positioning. At present, principal component analysis only involves the interdependence of a set of variables, and it is difficult to describe the degree of correlation between multiple state parameters of the main measuring point and multiple state parameters of other measuring points. Although canonical correlation analysis (CCA) can analyze the interdependence between two sets of variables, it only considers the overall correlation information and ignores the local important correlation information, making it difficult to accurately estimate the correlation between the state parameters of the main measuring point and the state parameters of other measuring points. Summary of the invention

[0006] In order to solve the above problems, the present invention proposes a diesel engine multivariate state parameter association analysis method and system based on hierarchical CCA. The method hierarchically decomposes the parameter state matrix based on different operators, effectively extracts the detailed features of the measuring points, and thus can accurately identify the associated measuring points that are strongly correlated with the main parameter measuring points of the diesel engine; it not only improves the accuracy of state monitoring, but also realizes a comprehensive and detailed analysis of the operating status of the diesel engine.

[0007] In order to achieve the above object, the present invention adopts the following technical solution:

[0008] In a first aspect, the present invention provides a method for correlation analysis of multivariate state parameters of a diesel engine based on hierarchical CCA, comprising:

[0009] Taking the diesel engine state type as the unit and based on the state parameters of each measuring point, a multi-dimensional time series matrix is ​​constructed;

[0010] According to whether there are measuring points exceeding the alarm limit, the multidimensional time series matrix is ​​divided into the main parameter state matrix and other parameter state matrices, and the main parameter measuring points are located;

[0011] Performing hierarchical decomposition on the state parameters of each measuring point in the main parameter state matrix and other parameter state matrices to obtain the main parameter hierarchical state matrix and other parameter hierarchical state matrix;

[0012] Perform canonical correlation analysis on the main parameter level state matrix and other parameter level state matrices to obtain the optimal projection vector;

[0013] Based on the main parameter hierarchical state matrix, the main parameter measuring point hierarchical state matrix is ​​obtained; based on the optimal projection vector, the hierarchical state matrices of other measuring points in all other parameter hierarchical state matrices are traversed, and the maximum correlation between them and the main parameter measuring point hierarchical state matrix is ​​calculated, and the measuring points corresponding to the other parameter hierarchical state matrices with the largest maximum correlation value are taken as associated measuring points.

[0014] Preferably, the state types include vibration, cylinder pressure, oil, temperature and rotation speed.

[0015] Preferably, the multidimensional time series matrix is ​​specifically:

[0016]

[0017] Where, l = 1, 2, ..., L, L represents the state type; is the state parameter of the measuring point; R (l) The number of measurement points deployed to monitor the lth state type, N is the number of collected samples.

[0018] Preferably, according to whether there are measuring points exceeding the alarm limit, the multidimensional time series matrix is ​​divided into a main parameter state matrix and other parameter state matrices, and the main parameter measuring points are located; specifically:

[0019] The state parameters of the measuring points are collected, and the measuring points whose state parameters exceed the preset alarm limit are taken as the main parameter measuring points. The multidimensional time series matrix corresponding to the main parameter measuring points is the main parameter state matrix; the multidimensional time series matrix other than the main parameter state matrix is ​​the other parameter state matrix.

[0020] Preferably, the hierarchical decomposition specifically includes:

[0021] Define an average operator, a difference operator and an index vector, wherein the average operator and the difference operator are used to perform average and difference operations on the state parameters of the measuring points, and the index vector is used to perform binary expression on each node, and the node is the result of the operator processing at each layer;

[0022] Constructing an operator matrix based on the average operator, the difference operator and the index vector;

[0023] Based on the state vector of the measuring point, the decomposition level and the operator matrix, the measuring point level subcomponents of the main parameters and other parameters are obtained;

[0024] The hierarchical subcomponents of all measurement points in the state matrix are aggregated to obtain a hierarchical state matrix.

[0025] Preferably, the method of performing canonical correlation analysis on the main parameter hierarchical state matrix and other parameter hierarchical state matrices to obtain the optimal projection vector specifically includes:

[0026] For the main parameter hierarchy state matrix A in the hth node (h) and the i-th other parameter level state matrix Perform pre-processing;

[0027] Based on the two projection directions w (h) and v (h) , maximize A (h) w (h) and Correlation coefficient:

[0028]

[0029] in, The correlation coefficient formula is optimized as follows:

[0030]

[0031] The optimized correlation coefficient formula is solved by singular value decomposition to obtain the optimal projection vector w at all levels. (h)* and v (h)* .

[0032] Preferably, the maximum correlation is calculated as follows:

[0033]

[0034] Among them, A (h) The main parameter level state matrix is ​​obtained based on the main parameter measurement points; is the hierarchical state matrix of other parameters, from which the hierarchical state matrix of other measuring points is extracted; w (h)* and v (h)* is the optimal projection vector.

[0035] In a second aspect, the present invention provides a diesel engine multivariate state parameter correlation analysis system based on hierarchical CCA, comprising:

[0036] A sequence matrix construction module is used to construct a multi-dimensional time series matrix based on the state parameters of each measuring point and the state type of the diesel engine;

[0037] The state matrix partitioning module is used to divide the multi-dimensional time series matrix into the main parameter state matrix and other parameter state matrices according to whether there are measuring points exceeding the alarm limit, and locate the main parameter measuring points;

[0038] A hierarchical decomposition module is used to hierarchically decompose the state parameters of each measuring point in the main parameter state matrix and other parameter state matrices to obtain the main parameter hierarchical state matrix and other parameter hierarchical state matrices;

[0039] The optimal projection calculation module is used to perform typical correlation analysis on the main parameter level state matrix and other parameter level state matrices to obtain the optimal projection vector;

[0040] The correlation calculation module is used to obtain the main parameter measurement point hierarchy state matrix based on the main parameter hierarchy state matrix; based on the optimal projection vector, the hierarchy state matrices of other measurement points in all other parameter hierarchy state matrices are traversed to calculate the maximum correlation between them and the main parameter measurement point hierarchy state matrix, and the measurement points corresponding to the other parameter hierarchy state matrices with the largest maximum correlation value are taken as associated measurement points.

[0041] In a third aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of a diesel engine multivariate state parameter association analysis method based on hierarchical CCA as described in the first aspect.

[0042] In a fourth aspect, the present invention provides a computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps in the method for correlation analysis of multivariate state parameters of a diesel engine based on hierarchical CCA as described in the first aspect are implemented.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] In actual engineering, due to the influence of low-frequency changes in equipment working conditions, it is difficult for the typical CCA model to accurately analyze the correlation between the state parameters of the main measuring point and the state parameters of other measuring points. The present invention fully considers the important local correlation information in the frequency domain between the state parameters of the main measuring point and the state parameters of the associated measuring points, and adds hierarchical analysis on the basis of the typical correlation analysis model. The scheme of the present invention proposes a hierarchical canonical correlation analysis model (Hierarchical Canonical Correlation Analysis, HCCA), which more comprehensively quantifies the correlation between the main parameter measuring point and other parameter measuring points, quickly locates the suspected fault position, and thus improves the fault diagnosis efficiency of the subsequent expert system intelligent algorithm.

[0045] Advantages of additional aspects of the present invention will be given in part in the following description, and in part will become obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The accompanying drawings constituting a part of the present invention are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their description are used to explain the present invention but do not constitute a limitation of the present invention.

[0047] Figure 1 A main flow chart of a diesel engine multivariate state parameter association analysis method based on hierarchical CCA provided by an embodiment of the present invention;

[0048] Figure 2 A detailed flow chart of a diesel engine multivariate state parameter association analysis method based on hierarchical CCA provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0050] Embodiment 1

[0051] like Figure 1 As shown, this embodiment discloses a diesel engine multivariate state parameter association analysis method based on hierarchical CCA, comprising the following steps:

[0052] S1: Taking the diesel engine state type as the unit and based on the state parameters of each measuring point, a multi-dimensional time series matrix is ​​constructed;

[0053] S2: According to whether there are measuring points exceeding the alarm limit, the multidimensional time series matrix is ​​divided into the main parameter state matrix and other parameter state matrices, and the main parameter measuring points are located;

[0054] S3: hierarchically decompose the state parameters of each measuring point in the main parameter state matrix and other parameter state matrices to obtain the main parameter hierarchical state matrix and other parameter hierarchical state matrix;

[0055] S4: Perform canonical correlation analysis on the main parameter level state matrix and other parameter level state matrices to obtain the optimal projection vector;

[0056] S5: Based on the main parameter hierarchical state matrix, obtain the main parameter measuring point hierarchical state matrix; based on the optimal projection vector, traverse the hierarchical state matrices of other measuring points in all other parameter hierarchical state matrices, calculate the maximum correlation between them and the main parameter measuring point hierarchical state matrix, and take the measuring points corresponding to the other parameter hierarchical state matrices with the largest maximum correlation value as the associated measuring points.

[0057] Next, combine Figure 2 , a diesel engine multivariate state parameter association analysis method based on hierarchical CCA disclosed in this embodiment is described in detail.

[0058] In S1, the state monitoring modules such as vibration, cylinder pressure, oil, temperature and speed are used to collect monitoring data and construct a multidimensional time series matrix, namely:

[0059]

[0060] Where, l = 1, 2, ..., L, L is the number of state monitoring modules; R (l) is the number of measurement points of the lth state monitoring module, and N is the number of collected samples; multiple modules collect data at fixed sampling points in chronological order.

[0061] In S2, when the state parameter of the lth module measurement point exceeds the alarm limit, the measurement point is set as the main parameter measurement point, that is, the multidimensional time series matrix A of the main parameter measurement point is (l) , and the multidimensional time series matrix set of other parameter measurement points except the main parameter measurement point A is It should be noted that the number of main parameter measurement points R A And other parameter measurement points same.

[0062] In S3, the state parameters of each measuring point in the main parameter state matrix A are hierarchically decomposed to obtain the main parameter hierarchical state matrix A at different levels. (h) ,h=1,2,…,2 k , k is the number of levels.

[0063] S301: define an average operator, a difference operator and an index vector, wherein the average operator and the difference operator are used to perform average and difference operations on the state parameters of the measuring points, and the index vector is used to perform binary expression on each node, and the node is all possible nodes processed by operators at each layer;

[0064] S302: constructing an operator matrix based on the average operator, the difference operator and the index vector;

[0065] S303: Based on the measuring point state vector, the decomposition level and the operator matrix, obtaining the measuring point level sub-components of the main parameters and other parameters;

[0066] S304: Gather the hierarchical subcomponents of all the measurement points in the state matrix to obtain a hierarchical state matrix.

[0067] Specifically, for the state vector A(e,:) of the e-th measurement point of the main parameter state matrix A, its subcomponents A at each level (h) (e,:) can be obtained by the following steps:

[0068] 1) Define the average operator Q respectively 0 and the difference operator Q 1 as follows:

[0069]

[0070] 2) Based on the average operator and the difference operator, define the k-th layer operator matrix as follows:

[0071]

[0072] Among them, r k is a binary number, r k ∈{0,1}; when r k When it is 0, the operator is the average operator at the k level; when r k When it is 1, the operator is the difference operator at the k level.

[0073] It should be understood that the process of obtaining the operator matrix can be clearly understood by those skilled in the art.

[0074] 3) Define a k-dimensional index vector [r 1 ,r 2 ,…,r k ], the index vector is used to represent each node in binary form, for example, if h=4, its index vector is [1,0,0]. The h-th hierarchical subcomponent corresponds to a unique index vector:

[0075]

[0076] In this embodiment, the index vector presents different combinations of binary numbers, which can uniquely identify different levels. Through the index vector, it is possible to clearly know which level is currently being processed, thereby avoiding confusion. At the same time, by introducing the index vector, the calculation results of different levels can be more clearly described and recorded, and the complex hierarchical decomposition process can be converted into a series of calculation and selection processes based on binary numbers. This makes the entire method more intuitive, easy to understand and apply.

[0077] 4) Estimate the components of each level of the state vector A(e,:) of the e-th measurement point as follows:

[0078]

[0079] The hierarchical subcomponent A (h) (e,:) is the hierarchical component of the state vector A(e,:) of the e-th measurement point in the main parameter state matrix A at the h-th node in the k-th layer. The hierarchical subcomponents of all measurement points in the main parameter state matrix A are collected to obtain the main parameter hierarchical state matrix A (h) .

[0080] Similarly, for the i-th parameter state matrix B except the main parameter state matrix (i) The state parameters of each measuring point are decomposed hierarchically to obtain other parameter state matrices at different levels.

[0081] In S4, for the main parameter state matrix A in the hth node (h) and other i-th parameter state matrices Perform normalization and de-averaging to find two projection directions w (h) and v (h) , maximize A (h) w (h) and The correlation coefficient of: The projection direction is simply the linear combination of the vectors of each measuring point of the parameter state matrix. That is, through linear combination, the two parameter matrices are converted into parameter vectors. By solving the two linear combinations, their correlation coefficients are maximized.

[0082]

[0083] in, Using an optimization method similar to SVM, the above objective function is rewritten as:

[0084]

[0085] Use singular value decomposition to solve the above formula and get the optimal projection vector w under all nodes (h)* and v (h)* .

[0086] In S5, the main parameter measurement point hierarchy state matrix is ​​obtained based on the main parameter hierarchy state matrix; based on the optimal projection vector, the hierarchy state matrices of other measurement points in all other parameter hierarchy state matrices are traversed, and the maximum correlation between them and the main parameter measurement point hierarchy state matrix is ​​calculated, and the measurement points corresponding to the other parameter hierarchy state matrices with the largest maximum correlation value are taken as associated measurement points.

[0087] Specifically, the maximum correlation between the main parameter variable and other i-th parameter variables is estimated as follows:

[0088]

[0089] The correlation between the main parameter measurement point and other parameter measurement points is evaluated, and they are sorted from large to small according to the maximum correlation. The suspected fault locations are checked in turn to obtain the associated measurement points.

[0090] The associated measuring points are obtained, and the fault judgment is performed by integrating the main measuring points and the associated measuring points, thereby realizing rapid fault location.

[0091] This embodiment facilitates the evaluation of the correlation between the main measuring point state parameters and the associated measuring point state parameters in different frequency bands through hierarchical decomposition, thereby preventing weak local important correlation information from being covered up.

[0092] This specific embodiment first constructs a multidimensional time series matrix to comprehensively reflect the status of each measuring point by monitoring the diesel engine status. The key main parameters and non-key parameters are distinguished by alarm limits, and then the two are hierarchically decomposed to refine the status characteristics. The projection direction is optimized using CCA technology to ensure that the maximum correlation between the main parameters and other parameters is accurately captured. Finally, through traversal calculation, the measurement points most closely associated with the state changes of the main parameters are determined, providing a scientific basis for accurate monitoring, fault diagnosis and performance optimization of diesel engines.

[0093] Embodiment 2

[0094] This embodiment provides a diesel engine multivariate state parameter correlation analysis system based on hierarchical CCA, including:

[0095] A sequence matrix construction module is used to construct a multi-dimensional time series matrix based on the state parameters of each measuring point and the state type of the diesel engine;

[0096] The state matrix partitioning module is used to divide the multi-dimensional time series matrix into the main parameter state matrix and other parameter state matrices according to whether there are measuring points exceeding the alarm limit, and locate the main parameter measuring points;

[0097] A hierarchical decomposition module is used to hierarchically decompose the state parameters of each measuring point in the main parameter state matrix and other parameter state matrices to obtain the main parameter hierarchical state matrix and other parameter hierarchical state matrices;

[0098] The optimal projection calculation module is used to perform typical correlation analysis on the main parameter level state matrix and other parameter level state matrices to obtain the optimal projection vector;

[0099] The correlation calculation module is used to obtain the main parameter measurement point hierarchy state matrix based on the main parameter hierarchy state matrix; based on the optimal projection vector, the hierarchy state matrices of other measurement points in all other parameter hierarchy state matrices are traversed to calculate the maximum correlation between them and the main parameter measurement point hierarchy state matrix, and the measurement points corresponding to the other parameter hierarchy state matrices with the largest maximum correlation value are taken as associated measurement points.

[0100] Embodiment 3

[0101] This embodiment provides a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the steps in the method for correlation analysis of multivariate state parameters of a diesel engine based on hierarchical CCA as described in the first embodiment above are implemented.

[0102] Embodiment 4

[0103] This embodiment provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, the steps in the method for correlation analysis of multivariate state parameters of a diesel engine based on hierarchical CCA as described in the first embodiment above are implemented.

[0104] The steps or modules involved in the above embodiments 2 to 4 correspond to those in embodiment 1. For the specific implementation, please refer to the relevant description of embodiment 1. The term "computer-readable storage medium" should be understood as a single medium or multiple media including one or more instruction sets; it should also be understood to include any medium that can store, encode or carry an instruction set for execution by a processor and enable the processor to execute any method in the present invention.

[0105] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A diesel engine multivariate state parameter correlation analysis method based on hierarchical CCA, characterized in that: include: Taking the diesel engine state type as the unit and based on the state parameters of each measuring point, a multi-dimensional time series matrix is ​​constructed; According to whether there are measuring points exceeding the alarm limit, the multidimensional time series matrix is ​​divided into the main parameter state matrix and other parameter state matrices, and the main parameter measuring points are located; Performing hierarchical decomposition on the state parameters of each measuring point in the main parameter state matrix and other parameter state matrices to obtain the main parameter hierarchical state matrix and other parameter hierarchical state matrix; Perform canonical correlation analysis on the main parameter level state matrix and other parameter level state matrices to obtain the optimal projection vector; Based on the main parameter hierarchical state matrix, the main parameter measuring point hierarchical state matrix is ​​obtained; based on the optimal projection vector, the hierarchical state matrices of other measuring points in all other parameter hierarchical state matrices are traversed, and the maximum correlation between them and the main parameter measuring point hierarchical state matrix is ​​calculated, and the measuring points corresponding to the other parameter hierarchical state matrices with the largest maximum correlation value are taken as associated measuring points.

2. A diesel engine multivariate state parameter correlation analysis method based on hierarchical CCA as claimed in claim 1, characterized in that: The status types include vibration, cylinder pressure, oil, temperature and speed.

3. A diesel engine multivariate state parameter association analysis method based on hierarchical CCA as claimed in claim 1, characterized in that: The multidimensional time series matrix is ​​specifically: Where, l = 1, 2, ..., L, L represents the state type; is the state parameter of the measuring point; R (l) The number of measurement points deployed to monitor the lth state type, N is the number of collected samples.

4. A diesel engine multivariate state parameter association analysis method based on hierarchical CCA as claimed in claim 1, characterized in that: According to whether there are measuring points exceeding the alarm limit, the multidimensional time series matrix is ​​divided into a main parameter state matrix and other parameter state matrices, and the main parameter measuring points are located; specifically: The state parameters of the measuring points are collected, and the measuring points whose state parameters exceed the preset alarm limit are taken as the main parameter measuring points. The multidimensional time series matrix corresponding to the main parameter measuring points is the main parameter state matrix; the multidimensional time series matrix other than the main parameter state matrix is ​​the other parameter state matrix.

5. A diesel engine multivariate state parameter correlation analysis method based on hierarchical CCA as claimed in claim 1, characterized in that: The hierarchical decomposition specifically includes: Define an average operator, a difference operator and an index vector, wherein the average operator and the difference operator are used to perform average and difference operations on the state parameters of the measuring points, and the index vector is used to perform binary expression on each node, and the node is the result of the operator processing at each layer; Constructing an operator matrix based on the average operator, the difference operator and the index vector; Based on the state vector of the measuring point, the decomposition level and the operator matrix, the measuring point level subcomponents of the main parameters and other parameters are obtained; The hierarchical subcomponents of all measurement points in the state matrix are aggregated to obtain a hierarchical state matrix.

6. A diesel engine multivariate state parameter association analysis method based on hierarchical CCA as claimed in claim 5, characterized in that: The method of performing canonical correlation analysis on the main parameter hierarchical state matrix and other parameter hierarchical state matrices to obtain the optimal projection vector specifically includes: For the main parameter hierarchy state matrix A in the hth node (h) and the i-th other parameter level state matrix Perform pre-processing; Based on the two projection directions w (h) and v (h) , maximize A (h) w (h) and Correlation coefficient: in, The correlation coefficient formula is optimized as follows: The optimized correlation coefficient formula is solved by singular value decomposition to obtain the optimal projection vector w at all levels. (h)* and v (h)* .

7. A diesel engine multivariate state parameter correlation analysis method based on hierarchical CCA as claimed in claim 1, characterized in that: The maximum correlation is calculated as follows: Among them, A (h) The main parameter level state matrix is ​​obtained based on the main parameter measurement points; is the hierarchical state matrix of other parameters, from which the hierarchical state matrix of other measuring points is extracted; w (h)* and v (h)* is the optimal projection vector.

8. A diesel engine multivariate state parameter correlation analysis system based on hierarchical CCA, characterized in that: include: A sequence matrix construction module is used to construct a multi-dimensional time series matrix based on the state parameters of each measuring point and the state type of the diesel engine; The state matrix partitioning module is used to divide the multi-dimensional time series matrix into the main parameter state matrix and other parameter state matrices according to whether there are measuring points exceeding the alarm limit, and locate the main parameter measuring points; A hierarchical decomposition module is used to hierarchically decompose the state parameters of each measuring point in the main parameter state matrix and other parameter state matrices to obtain the main parameter hierarchical state matrix and other parameter hierarchical state matrices; The optimal projection calculation module is used to perform typical correlation analysis on the main parameter level state matrix and other parameter level state matrices to obtain the optimal projection vector; The correlation calculation module is used to obtain the main parameter measurement point hierarchy state matrix based on the main parameter hierarchy state matrix; based on the optimal projection vector, the hierarchy state matrices of other measurement points in all other parameter hierarchy state matrices are traversed to calculate the maximum correlation between them and the main parameter measurement point hierarchy state matrix, and the measurement points corresponding to the other parameter hierarchy state matrices with the largest maximum correlation value are taken as associated measurement points.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps in a diesel engine multivariate state parameter association analysis method based on hierarchical CCA as described in any one of claims 1 to 7 are implemented.

10. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps in the method for correlation analysis of multivariate state parameters of a diesel engine based on hierarchical CCA as described in any one of claims 1-7 are implemented.

Images