Enhanced fault diagnosis method for physical field feedback of generator set

Abstract

The invention discloses an enhanced fault diagnosis method for physical field feedback of a generator set, and belongs to the field of intelligent operation and maintenance of power generation equipment. According to the method, physical field information inside and outside parts is obtained through establishment of a generator set mechanism model and simulation, and an enhanced fault diagnosis method is established through extraction and generation of new characteristic variables and updating of original fault reason logic criteria according to a generator set physical field digital twinborn model result. The method has the advantages of being scientific in principle, clear in process and remarkable in effect, the physical field digital twinning simulation result can be fed back to fault reason criterion logic, the accuracy and expansibility of a fault diagnosis model are effectively improved, and a beneficial method is provided for construction of the digital twinning generator set.

Description

Technical Field

[0001] The present invention relates to the field of intelligent operation and maintenance of power generation equipment, and in particular to an enhanced fault diagnosis method for physical field feedback of a generator set. Background Art

[0002] In the intelligent operation and maintenance system of generator sets, physical field digital twin is a very important research direction. Through physical field digital twin, the physical field information of generator set operation can be mastered, providing a basis for the operation and maintenance of generator sets.

[0003] However, the results of the physical field digital twin of the generator set are mainly displayed, and have not been fully and effectively utilized, and there is a lack of feedback from the results on fault diagnosis. The physical field digital twin can be understood as adding infinite measurement points to the key components of the unit. These "new measurement points" have added a large number of source bases for fault diagnosis. Therefore, the effective application of the physical field digital twin structure provides great possibilities for improving the accuracy of fault diagnosis results. Summary of the invention

[0004] The purpose of the present invention is to provide an enhanced fault diagnosis method for a generator set based on physical field feedback. The technical solution of the present invention comprises the following steps:

[0005] S1. According to the online monitoring point data of the generator set, the physical field digital twin simulation model of the generator set components is driven to obtain the whole domain physical field distribution result of the physical field digital twin simulation model. The physical field simulation value can be expressed as:

[0006] f=F(x,y,z)

[0007] In the formula, f is the physical field simulation value; x, y, z are the coordinate points corresponding to the domain space;

[0008] S2. Extract physical field characteristic parameters according to the component physical field simulation results. For the specified coordinate point (x k ,y k 、z k ) physical field results, and obtain f through the physical field simulation value of the domain space k :

[0009] f k =F(x k ,y k ,z k )

[0010] For the physical field characteristic value of the specified area Ω, the maximum value f of the specified area Ω is obtained through the physical field simulation value of the domain space max , mean value f men And the integral value f inv :

[0011] f max =max Ω (F(x,y,z))

[0012] f men =men Ω (F(x,y,z))

[0013]

[0014] S3. In the original fault diagnosis model criterion of the generator set, replace or add the fault cause logic criterion condition as follows:

[0015] The corresponding coordinate point (x k ,y k 、z k )'s physical field results f k , which is consistent with the physical quantity parameters of the component, replaces the logical criterion parameters corresponding to the physical quantity of the original component;

[0016] The corresponding other coordinate points (x k ,y k 、z k )'s physical field results f k , and the physical field eigenvalue f of the specified region Ω max 、f men and f inv , the physical quantity characteristic information of the specified area Ω of the component is newly obtained, and the logical judgment criteria of the fault cause are added;

[0017] S4. Supplement the conditional probability table of the fault cause corresponding to the fault cause logic criterion, including the probability values ​​corresponding to different results of the characteristic criterion. The method of supplementing the conditional probability table of the fault cause corresponding to the fault cause logic criterion is as follows:

[0018] The conditional probability table P of the fault cause to which the original fault cause logic criterion belongs is expressed as a matrix:

[0019]

[0020] In the formula, a ij Represents the element in the i-th row and j-th column of the conditional probability table P (1≤i≤m, 1≤j≤k);

[0021] When replacing the logical criterion parameters corresponding to the physical quantity of the original component in the fault diagnosis cause logical criterion, directly replace the corresponding element a ij Change to a' ij ;

[0022] When the fault diagnosis cause logic criterion is added to the fault diagnosis cause logic criterion and the corresponding conditional probability table Q, it is expressed as:

[0023]

[0024] Where b ij represents the i-th row and j-th column element of the conditional probability table Q (1≤i≤w, 1≤j≤k), then the conditional probability table of the corresponding fault cause is updated to P':

[0025]

[0026] S5. Drive the fault diagnosis module and physical field simulation model through online monitoring data, update the original fault diagnosis criteria and the physical field characteristic parameter criteria, enhance the fault diagnosis judgment logic, update the fault cause probability, and achieve fault isolation.

[0027] Judge the logic, update the probability of fault cause, and achieve fault isolation.

[0028] As a preferred implementation, in S1, the physical field simulation value can be expressed as:

[0029] f=F(x,y,z)

[0030] In the formula, f is the physical field simulation value; x, y, z are the coordinate points corresponding to the domain space;

[0031] As a preferred implementation, in S2, the method for extracting physical field characteristic parameters is as follows:

[0032] S31, for the specified coordinate point (x k ,y k 、z k ) physical field results, and obtain f through the physical field simulation value of the domain space k :

[0033] f k =F(x k ,y k ,z k )

[0034] S32, for the physical field characteristic value of the specified area Ω, obtain the maximum value f of the specified area Ω through the physical field simulation value of the domain space max , mean value f men And the integral value f inv :

[0035] f max =max Ω (F(x,y,z))

[0036] f men =men Ω (F(x,y,z))

[0037]

[0038] As a preferred implementation, in S3, the fault cause logic criterion condition is replaced or added as follows:

[0039] S41, the corresponding coordinate point (x k ,y k 、z k )'s physical field results f k , which is consistent with the physical quantity parameters of the component, replaces the logical criterion parameters corresponding to the physical quantity of the original component;

[0040] S42, the corresponding other coordinate points (x k ,y k 、z k )'s physical field results f k , and the physical field eigenvalue f of the specified region Ω max 、f men and f inv , the physical quantity characteristic information of the specified area Ω of the component is newly obtained, and the logical judgment criteria of the fault cause are added.

[0041] As a preferred implementation, in S4, the method for supplementing the conditional probability table of the fault cause corresponding to the fault cause logic criterion is as follows:

[0042] S51. The conditional probability table P of the fault cause to which the original fault cause logic criterion belongs is represented by the matrix:

[0043]

[0044] In the formula, a ij Represents the element in the i-th row and j-th column of the conditional probability table P (1≤i≤m, 1≤j≤k);

[0045] S52, when replacing the logical criterion parameter corresponding to the physical quantity of the original component in the fault diagnosis cause logical criterion, directly replace the corresponding element a ij Change to a' ij ;

[0046] S53, in the fault diagnosis cause logic criterion, a new fault cause logic criterion is added with the corresponding conditional probability table Q, which is expressed as:

[0047]

[0048] Where b ij represents the i-th row and j-th column element of the conditional probability table Q (1≤i≤w, 1≤j≤k), then the conditional probability table of the corresponding fault cause is updated to P':

[0049]

[0050] The beneficial effects of the present invention are:

[0051] The present invention proposes an enhanced fault diagnosis method for a generator set based on physical field feedback. Through this innovative method, the following technical effects can be achieved:

[0052] 1. Improve the accuracy of generator set fault diagnosis. Through the physical field digital twin results of generator set components, the data sources and criteria for fault diagnosis are greatly increased, and the analysis ability and effect of fault diagnosis are improved;

[0053] 2. Enhance the scalability of fault diagnosis cause criteria. New fault cause criteria are added through the physical field digital twin results, which can be combined with the conditional probability table of the original fault cause to establish a method for replacing and updating the conditional probability table, effectively enhancing the scalability of the fault diagnosis criteria. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 Flowchart of the enhanced fault diagnosis method based on the physical field feedback of the generator set. DETAILED DESCRIPTION

[0055] The present application is further described below in conjunction with the accompanying drawings. The following embodiments are only used to more clearly illustrate the technical solution of the present invention, and cannot be used to limit the protection scope of the present application.

[0056] An enhanced fault diagnosis method for a generator set based on physical field feedback, wherein a specific implementation method 1 comprises the following steps:

[0057] S1, such as Figure 1 As shown in the figure, according to the online monitoring point data of the generator set, the physical field digital twin simulation model of the generator set components is driven to obtain the full domain physical field distribution result of the physical field digital twin simulation model. The physical field simulation value can be expressed as:

[0058] f=F(x,y,z)

[0059] In the formula, f is the physical field simulation value; x, y, z are the coordinate points corresponding to the domain space;

[0060] S2, such as Figure 1 As shown, according to the physical field simulation results of the component, the physical field characteristic parameters are extracted, and the physical field characteristic parameters are extracted. For the specified coordinate point (x k ,y k 、z k ) physical field results, and obtain f through the physical field simulation value of the domain space k :

[0061] f k =F(x k ,y k ,z k)

[0062] For the physical field characteristic value of the specified area Ω, the maximum value f of the specified area Ω is obtained through the physical field simulation value of the domain space max , mean value f men And the integral value f inv :

[0063] f max =max Ω (F(x,y,z))

[0064] f men =men Ω (F(x,y,z))

[0065]

[0066] S3, such as Figure 1 As shown in the figure, in the original fault diagnosis model criterion of the generator set, the fault cause logic criterion condition is replaced or added as follows:

[0067] The corresponding coordinate point (x k ,y k 、z k )'s physical field results f k , which is consistent with the physical quantity parameters of the component, replaces the logical criterion parameters corresponding to the physical quantity of the original component;

[0068] The corresponding other coordinate points (x k ,y k 、z k )'s physical field results f k , and the physical field eigenvalue f of the specified region Ω max 、f men and f inv , the physical quantity characteristic information of the specified area Ω of the component is newly obtained, and the logical judgment criteria of the fault cause are added;

[0069] S4, such as Figure 1 As shown, the conditional probability table of the fault cause corresponding to the fault cause logic criterion is supplemented, including the probability values ​​corresponding to different results of the characteristic criterion. The method of supplementing the conditional probability table of the fault cause corresponding to the fault cause logic criterion is as follows:

[0070] The conditional probability table P of the fault cause to which the original fault cause logic criterion belongs is expressed as a matrix:

[0071]

[0072] In the formula, a ij Represents the element in the i-th row and j-th column of the conditional probability table P (1≤i≤m, 1≤j≤k);

[0073] When replacing the logical criterion parameters corresponding to the physical quantity of the original component in the fault diagnosis cause logical criterion, directly replace the corresponding element a ij Change to a' ij ;

[0074] When the fault diagnosis cause logic criterion is added to the fault diagnosis cause logic criterion and the corresponding conditional probability table Q, it is expressed as:

[0075]

[0076] Where b ij represents the i-th row and j-th column element of the conditional probability table Q (1≤i≤w, 1≤j≤k), then the conditional probability table of the corresponding fault cause is updated to P':

[0077]

[0078] S5. Drive the fault diagnosis module and physical field simulation model through online monitoring data, update the original fault diagnosis criteria and the physical field characteristic parameter criteria, enhance the fault diagnosis judgment logic, update the fault cause probability, and achieve fault isolation.

[0079] In this embodiment, the physical field parameters and eigenvalue information are obtained through the physical field digital twin results of the generator set components, which are not limited to eigenvalues ​​such as maximum value, average value and integral value, but also include other eigenvalues ​​and eigenvalue combinations that can be extracted. At the same time, the fault diagnosis results will also have a relevant impact on the physical field digital twin, further improving the accuracy of the physical field digital twin results, and the two are iteratively updated with each other.

[0080] The present invention is only an exemplary description of the present invention and does not limit its protection scope. Those skilled in the art may also make partial changes thereto, which are within the protection scope of the present invention as long as they do not exceed the spirit of the present invention.

Claims

1. An enhanced fault diagnosis method based on physical field feedback of generator sets, Its characteristics include the following steps: S1. According to the online monitoring point data of the generator set, the physical field digital twin simulation model of the generator set components is driven to obtain the whole domain physical field distribution results of the physical field digital twin simulation model; S2. Extracting physical field characteristic parameters according to the physical field simulation results of the components; S3. Replace or add a fault cause logic criterion to the original fault diagnosis model criterion of the generator set; S4. Supplement the conditional probability table of the fault cause corresponding to the fault cause logic criterion, including the probability values ​​corresponding to different results of the characteristic criterion; S5. Drive the fault diagnosis module and physical field simulation model through online monitoring data, update the original fault diagnosis criteria and the physical field characteristic parameter criteria, enhance the fault diagnosis judgment logic, update the fault cause probability, and achieve fault isolation.

2. The enhanced fault diagnosis method of a generator set based on physical field feedback according to claim 1 is characterized in that: In S1, the physical field simulation value can be expressed as: f=F(x,y,z) Where f is the physical field simulation value; x, y, and z are the coordinate points corresponding to the domain space.

3. The enhanced fault diagnosis method of a generator set physical field feedback according to claim 1 is characterized by: In S2, the method for extracting the physical field characteristic parameters is as follows: S31, for the specified coordinate point (x k ,y k 、z k ) physical field results, and obtain f through the physical field simulation value of the domain space k : f k =F(x k ,y k ,z k ) S32, for the physical field characteristic value of the specified area Ω, obtain the maximum value f of the specified area Ω through the physical field simulation value of the domain space max , mean value f men And the integral value f inv : f max =max Ω (F(x,y,z)) f men =men Ω (F(x,y,z)) 4. The enhanced fault diagnosis method of a generator set physical field feedback according to claim 1 is characterized by: In S3, the logical criterion condition for replacing or adding the fault cause is: S41, the corresponding coordinate point (x k ,y k 、z k )'s physical field results f k , which is consistent with the physical quantity parameters of the component, replaces the logical criterion parameters corresponding to the physical quantity of the original component; S42, the corresponding other coordinate points (x k ,y k 、z k )'s physical field results f k , and the physical field eigenvalue f of the specified region Ω max 、f men and f inv , the physical quantity characteristic information of the specified area Ω of the component is newly obtained, and the logical judgment criteria of the fault cause are added.

5. The enhanced fault diagnosis method of a generator set physical field feedback according to claim 1 is characterized by: In S4, the method of supplementing the conditional probability table of the fault cause corresponding to the fault cause logic criterion is as follows: S51. The conditional probability table P of the fault cause to which the original fault cause logic criterion belongs is represented by the matrix: In the formula, a ij Represents the element in the i-th row and j-th column of the conditional probability table P (1≤i≤m, 1≤j≤k); S52, when replacing the logical criterion parameter corresponding to the physical quantity of the original component in the fault diagnosis cause logical criterion, directly replace the corresponding element a ij Change to a' ij ; S53, in the fault diagnosis cause logic criterion, a new fault cause logic criterion is added with the corresponding conditional probability table Q, which is expressed as: Where b ij represents the i-th row and j-th column element of the conditional probability table Q (1≤i≤w, 1≤j≤k), then the conditional probability table of the corresponding fault cause is updated to P':

Images