A multi-granularity variable-scale fuzzy neighborhood measure and the corresponding Choquet-like integral and its fault diagnosis method

By introducing multi-grained variable-scale fuzzy Θ-covering approximate space and Choquet-like integral in multi-grained variable-scale fuzzy scenarios, combined with intelligent classifiers, the attribute reduction method in the existing technology cannot delete redundant attributes and adapt to different thresholds. The problem of β achieves efficient fault diagnosis.

CN117851898BActive Publication Date: 2024-11-15SHAANXI UNIV OF SCI & TECH
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202311119694.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2023-09-01
Publication Date
2024-11-15
Estimated Expiration
2043-09-01

AI Technical Summary

Technical Problem

The existing attribute reduction method of the fuzzy rough set approximation space cannot effectively delete redundant attributes in the data, and has strong constraints on conditional entropy, and cannot adapt to the threshold β of different attributes, resulting in fault diagnosis in multi-grained variable-scale fuzzy scenarios. Inefficient.

Method used

A method of attribute reduction in multi-grained variable-scale fuzzy Θ-covered approximate space is proposed. Combined with Choquet-like integral, the attribute reduction and fault diagnosis of decision tables are realized through multi-grained variable-scale fuzzy neighborhood measurement and intelligent classifier. .

Benefits of technology

Effectively delete redundant fault characteristics, improve the accuracy and efficiency of fault diagnosis after dimensionality reduction, and is suitable for the diagnosis of mechanical faults such as rolling bearings, and overcomes the limitations of the attribute reduction method in the existing technology.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN117851898B_ABST
    Figure CN117851898B_ABST
Patent Text Reader

Abstract

The invention discloses a multi-granularity variable-scale fuzzy neighborhood measure and a corresponding Choquet-like integral and a fault diagnosis method thereof, and a fault diagnosis method based on "attribute reduction + intelligent classifier": input: a complete fault diagnosis decision information table; step 1, using a new attribute reduction method of a decision table based on a Choquet-like integral of a multi-granularity variable-scale fuzzy neighborhood measure, deleting redundant fault features, and realizing dimension reduction of the fault decision information table; the multi-granularity variable-scale fuzzy neighborhood measure and the corresponding Choquet-like integral and a fault diagnosis method thereof, after exporting the required data, for each collected data file, taking every 500 data points as an example, using an existing feature extraction method, all feature parameters are calculated as candidate features. The proposed reduction method of the Choquet-like integral based on the multi-granularity variable-scale fuzzy neighborhood measure is used to realize the fault diagnosis of rolling bearings.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of multi-granularity variable-scale fuzzy technology, and in particular to a multi-granularity variable-scale fuzzy neighborhood measure and a corresponding Choquet-like integral and a fault diagnosis method thereof. Background Art

[0002] In existing research, for multiple fuzzy β-cover approximate spaces, different attributes in the data set obtain fuzzy β-covers with the same threshold β. However, in real problems, different attributes will produce different thresholds β. In other words, it is necessary to use different attributes in the data set to induce different fuzzy β-covers with different thresholds β, that is, the present invention will establish a new multi-granularity variable-scale fuzzy Θ-cover approximate space. Moreover, attribute reduction methods (i.e., feature selection methods) based on existing fuzzy rough set approximate spaces all have certain defects. For example, Wang et al. [1] found that some existing attribute reduction methods [2-4] could not find simplification, that is, they could not delete redundant attributes in the data; the attribute reduction method proposed by Dai et al. [5] has strong constraints on conditional entropy because log will appear. 10 0. Therefore, the present invention proposes a new attribute reduction method in multi-granularity variable-scale fuzzy θ-covering approximate space, and successfully uses it in the fault diagnosis of rolling bearings. Summary of the invention

[0003] The purpose of the present invention is to provide a multi-granularity variable-scale fuzzy neighborhood measure and a corresponding Choquet-like integral and a fault diagnosis method thereof to solve the problems raised in the above background technology.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: a multi-granularity variable-scale fuzzy neighborhood measure and corresponding Choquet-like integral and fault diagnosis method thereof, a fault diagnosis method based on "attribute reduction + intelligent classifier":

[0005] Input: Complete fault diagnosis decision information table;

[0006] Step 1: Using a new method of attribute simplification of decision tables based on Choquet-like integral of multi-granularity variable-scale fuzzy neighborhood measure, redundant fault features are deleted to achieve dimensionality reduction of the fault decision information table;

[0007] Step 2: Select an intelligent classifier model and establish an intelligent classifier diagnosis model; construct a new fault data set based on the reduction determined in step 1; then, use it to train the diagnosis model;

[0008] Step 3: Use the trained intelligent classifier to classify the test sample symptom set to obtain the diagnosis result;

[0009] Output: fault diagnosis results and decisions;

[0010] The new method for reducing the attributes of the decision table of the multi-granularity variable-scale fuzzy neighborhood measure and the corresponding Choquet-like integral in step 1 includes:

[0011] S1. Propose a multi-granularity variable-scale fuzzy Θ-covering approximation space:

[0012] Definition 1. Let U be a domain, Θ = {β1,…,βi,…,βn}, where βi∈(0,1]; if is a fuzzy βi-cover of U, then It is called a multi-granularity variable-scale fuzzy Θ-cover on U; (U,Ω) is also called a multi-granularity variable-scale fuzzy Θ-cover approximation space;

[0013] S2. Propose a multi-granularity variable-scale fuzzy Θ-neighborhood:

[0014] Definition 2. Let (U,Ω) be a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ={β1,…,βi,…,βn}; for any x∈U, the multi-granularity variable-scale fuzzy Θ-neighborhood of x is defined as:

[0015]

[0016] in, For x The fuzzy βi-minimum description under , IT is the R-implication operator induced by the fuzzy logic operator t-norm;

[0017] S3. Propose upper and lower multi-granularity variable-scale fuzzy neighborhood measures:

[0018] Let (U,Ω) be a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ={β1,…,βi,…,βn}; for any The upper and lower multi-granularity variable-scale fuzzy neighborhood measures of X are defined as:

[0019]

[0020] S4. A Choquet-like integral based on upper and lower multi-granularity variable-scale fuzzy neighborhood measure is proposed:

[0021] Definition 4. Let (U,Ω) be a multi-granular variable-scale fuzzy Θ-covering approximation space, T be the t-norm, A∈F(U), where U = {x1,…,xn}; then Θ = {β1,…,βi,…,βn}, for any The upper and lower multi-granularity variable-scale fuzzy neighborhood measures of X are defined as:

[0022]

[0023] where {x(1),x(2),…,x(n)} is a permutation of {x1,x2,…,xn} such that A(x(1))≤A(x(2))≤…≤A(x(n)), X(i)={x(i),x(i+1)…,x(n)},

[0024] S5. Based on S1-S4, a new method for attribute reduction of decision tables based on Choquet-like integral of multi-granularity variable-scale fuzzy neighborhood measure is proposed:

[0025] Input: Input decision information table T = (U, A∪{d}), where U = {x1, x2, …, xn}, A = {a1, a2, …, am}, bij = aj(xi), i∈{1, 2, …, n}, j∈{1, 2, …, m};

[0026] Table 1 Decision information table T = (U, A ∪ {d})

[0027]

[0028] Output: attribute reduction of attribute set A.

[0029] Preferably, the S5 implementation process specifically includes:

[0030] Step 1: Induce a multi-granularity variable-scale fuzzy Θ-covering approximation space: For any aj∈A, define the fuzzy set in is the variance of attribute aj; then is a fuzzy βj-cover (βj∈(0,γj]), Therefore, (U,Ω) is a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ = {β1,…,βi,…,βn}, also known as (U,Ω,d) is a multi-granularity variable-scale fuzzy Θ-covering group decision information system;

[0031] Step 2: Calculate the multi-granularity variable-scale fuzzy Θ-neighborhood for all xk∈U

[0032]

[0033] Step 3: Calculation Where xk∈U, U / d={D1,…,Dt,…,Ds} is the partition of the domain U with respect to the decision attribute d;

[0034] Step 4: Calculate the minimum Choquet-like reduction Redt of (U,Ω,d). The specific steps are as follows:

[0035] (a) P←{};

[0036] (b) for any D∈U / d

[0037] (c)do

[0038] (d)Q←P;

[0039] (e)

[0040] (f)

[0041] (e) Q←P∪{C}

[0042] (f) P←Q

[0043] (h)return P%Suppose

[0044] (i) Redt = {a (1) , …, a (h)}.

[0045] A multi-granularity variable-scale fuzzy neighborhood measure and its corresponding Choquet-like integral fault diagnosis method: The bearing used in the test supports the rotating shaft of the motor. In order to obtain the bearing fault data, the electric spark machining technology is used to machine cracks on the rolling bearing, generating a series of fault bearings with notch diameters of 0.007 inches, 0.014 inches, 0.021 inches, 0.028 inches and 0.4 inches. Among them, for the outer ring fault, not only cracks of different sizes are generated, but also outer ring faults of the bearing are created at different positions relative to the load stress point, which are located at 3 o'clock, 6 o'clock and 12 o'clock of the load point respectively; in order to measure the vibration data of the faulty bearing, the experimental platform is equipped with acceleration sensors on the drive end, fan end and base of the motor respectively. During the data collection process, the experimental platform only sets a single fault type, so the fault types in the bearing data set used in the test are all single fault types; during the bearing vibration data collection process, the test platform is loaded with loads of different horsepower, and the data sampling frequency is 12KHz;

[0046] After exporting the required data, for each collected data file, every 500 data points are taken as an example, and all feature parameters are calculated as candidate features using the existing feature extraction method; then the proposed reduction method of Choquet-like integral based on multi-granularity variable-scale fuzzy neighborhood measure is used to realize the fault diagnosis of rolling bearings.

[0047] Compared with the prior art, the present invention has the following beneficial effects:

[0048] 1. The multi-granularity variable-scale fuzzy neighborhood measure and the corresponding Choquet-like integral and its fault diagnosis method. The bearing used in the test supports the rotating shaft of the motor. In order to obtain the bearing fault data, the electric spark machining technology is used to process cracks on the rolling bearing, resulting in a series of fault bearings with diameters of 0.007 inches, 0.014 inches, 0.021 inches, 0.028 inches and 0.4 inches. Among them, for the outer ring fault, not only cracks of different sizes are generated, but also outer ring faults of the bearing are created at different positions relative to the load stress point, which are located at 3 o'clock, 6 o'clock and 12 o'clock of the stress point respectively; in order to measure the vibration data of the faulty bearing, the experimental platform is equipped with acceleration sensors on the drive end, fan end and base of the motor respectively. During the data collection process, the experimental platform only sets a single fault type, so the fault types in the bearing data set used in the test are all single fault types; during the bearing vibration data collection process, the test platform is loaded with loads of different horsepower, and the data sampling frequency is 12KHz;

[0049] After exporting the required data, for each collected data file, every 500 data points are taken as an example, and all feature parameters are calculated as candidate features using the existing feature extraction method; then the proposed reduction method of Choquet-like integral based on multi-granularity variable-scale fuzzy neighborhood measure is used to realize the fault diagnosis of rolling bearings. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 It is a schematic diagram of the process of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0052] See also Figure 1 The present invention provides a technical solution: a multi-granularity variable-scale fuzzy neighborhood measure and a corresponding Choquet-like integral and a fault diagnosis method thereof, a fault diagnosis method based on "attribute reduction + intelligent classifier":

[0053] Input: Complete fault diagnosis decision information table;

[0054] Step 1: Using a new method of attribute simplification of decision tables based on Choquet-like integral of multi-granularity variable-scale fuzzy neighborhood measure, redundant fault features are deleted to achieve dimensionality reduction of the fault decision information table;

[0055] Step 2: Select an intelligent classifier model and establish an intelligent classifier diagnosis model; construct a new fault data set based on the reduction determined in step 1; then, use it to train the diagnosis model;

[0056] Step 3: Use the trained intelligent classifier to classify the test sample symptom set to obtain the diagnosis result;

[0057] Output: fault diagnosis results and decisions;

[0058] The new method for reducing the attributes of the decision table of the multi-granularity variable-scale fuzzy neighborhood measure and the corresponding Choquet-like integral in step 1 includes:

[0059] S1. Propose a multi-granularity variable-scale fuzzy Θ-covering approximation space:

[0060] Definition 1. Let U be a domain, Θ = {β1,…,βi,…,βn}, where βi∈(0,1]; if is a fuzzy βi-cover of U, then It is called a multi-granularity variable-scale fuzzy Θ-cover on U; (U,Ω) is also called a multi-granularity variable-scale fuzzy Θ-cover approximation space;

[0061] S2. Propose a multi-granularity variable-scale fuzzy Θ-neighborhood:

[0062] Definition 2. Let (U,Ω) be a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ={β1,…,βi,…,βn}; for any x∈U, the multi-granularity variable-scale fuzzy Θ-neighborhood of x is defined as:

[0063]

[0064] in, For x The fuzzy βi-minimum description under , IT is the R-implication operator induced by the fuzzy logic operator t-norm;

[0065] S3. Propose upper and lower multi-granularity variable-scale fuzzy neighborhood measures:

[0066] Let (U,Ω) be a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ={β1,…,βi,…,βn}; for any The upper and lower multi-granularity variable-scale fuzzy neighborhood measures of X are defined as:

[0067]

[0068] S4. A Choquet-like integral based on upper and lower multi-granularity variable-scale fuzzy neighborhood measure is proposed:

[0069] Definition 4. Let (U,Ω) be a multi-granular variable-scale fuzzy Θ-covering approximation space, T be the t-norm, A∈F(U), where U = {x1,…,xn}; then Θ = {β1,…,βi,…,βn}, for any The upper and lower multi-granularity variable-scale fuzzy neighborhood measures of X are defined as:

[0070]

[0071] where {x(1),x(2),…,x(n)} is a permutation of {x1,x2,…,xn} such that A(x(1))≤A(x(2))≤…≤A(x(n)), X(i)={x(i),x(i+1)…,x(n)},

[0072] S5. Based on S1-S4, a new method for attribute reduction of decision tables based on Choquet-like integral of multi-granularity variable-scale fuzzy neighborhood measure is proposed:

[0073] Input: Input decision information table T = (U, A∪{d}), where U = {x1, x2, …, xn}, A = {a1, a2, …, am}, bij = aj(xi), i∈{1, 2, …, n}, j∈{1, 2, …, m};

[0074] Table 1 Decision information table T = (U, A ∪ {d})

[0075]

[0076] Output: attribute reduction of attribute set A.

[0077] Preferably, the S5 implementation process specifically includes:

[0078] Step 1: Induce a multi-granularity variable-scale fuzzy Θ-covering approximation space: For any aj∈A, define the fuzzy set in is the variance of attribute aj; then is a fuzzy βj-cover (βj∈(0,γj]), Therefore, (U,Ω) is a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ = {β1,…,βi,…,βn}, also known as (U,Ω,d) is a multi-granularity variable-scale fuzzy Θ-covering group decision information system;

[0079] Step 2: Calculate the multi-granularity variable-scale fuzzy Θ-neighborhood for all xk∈U

[0080]

[0081] Step 3: Calculation Where xk∈U, U / d={D1,…,Dt,…,Ds} is the partition of the domain U with respect to the decision attribute d;

[0082] Step 4: Calculate the minimum Choquet-like reduction Redt of (U,Ω,d). The specific steps are as follows:

[0083] (a) P←{};

[0084] (b) for any D∈U / d

[0085] (c)do

[0086] (d)Q←P;

[0087] (e)

[0088] (f)

[0089] (e) Q←P∪{C}

[0090] (f) P←Q

[0091] (h)return P%Suppose

[0092] (i) Redt = {a (1) , …, a (h)}.

[0093] A multi-granularity variable-scale fuzzy neighborhood measure and its corresponding Choquet-like integral fault diagnosis method: The bearing used in the test supports the rotating shaft of the motor. In order to obtain the bearing fault data, the electric spark machining technology is used to machine cracks on the rolling bearing, generating a series of fault bearings with notch diameters of 0.007 inches, 0.014 inches, 0.021 inches, 0.028 inches and 0.4 inches. Among them, for the outer ring fault, not only cracks of different sizes are generated, but also outer ring faults of the bearing are created at different positions relative to the load stress point, which are located at 3 o'clock, 6 o'clock and 12 o'clock of the load point respectively; in order to measure the vibration data of the faulty bearing, the experimental platform is equipped with acceleration sensors on the drive end, fan end and base of the motor respectively. During the data collection process, the experimental platform only sets a single fault type, so the fault types in the bearing data set used in the test are all single fault types; during the bearing vibration data collection process, the test platform is loaded with loads of different horsepower, and the data sampling frequency is 12KHz;

[0094] After exporting the required data, for each collected data file, every 500 data points are taken as an example, and all feature parameters are calculated as candidate features using the existing feature extraction method; then the proposed reduction method of Choquet-like integral based on multi-granularity variable-scale fuzzy neighborhood measure is used to realize the fault diagnosis of rolling bearings.

[0095] In summary: Figure 1As shown, when using the multi-granularity variable-scale fuzzy neighborhood measure and the corresponding Choquet-like integral and its fault diagnosis method, in order to analyze the application of the decomposition algorithm in the vibration signal of rotating machinery, the present invention uses the bearing data of Case Western Reserve University for experiments [123,124]. The Case Western Reserve University bearing fault experimental platform consists of the following parts: a 2-horsepower motor, which is placed on the left side of the platform, a torque sensor and an encoder in the middle of the platform, and a dynamometer on the right side of the platform. The bearing used in the test supports the rotating shaft of the motor. In order to obtain the bearing fault data, cracks are machined on the rolling bearing using electrospark machining technology, resulting in a series of fault bearings with diameters of 0.007 inches, 0.014 inches, 0.021 inches, 0.028 inches and 0.4 inches. For the outer ring fault, not only cracks of different sizes are generated, but also outer ring faults of the bearing are created at different positions relative to the load point, which are located at 3 o'clock, 6 o'clock and 12 o'clock of the load point respectively. In order to measure the vibration data of the faulty bearing, the experimental platform installed acceleration sensors on the drive end, fan end and base of the motor. During the data collection process, the experimental platform only set a single fault type, so the fault types in the bearing data set used in the test were all single fault types. During the bearing vibration data collection process, the test platform was loaded with loads of different horsepower, and the data sampling frequency was 12KHz.

[0096] After exporting the required data, for each collected data file, take every 500 data points as an example, use the existing feature extraction method to calculate all the feature parameters as candidate features (here mainly three types of data are processed, the data information is shown in Table 1; the data information after feature extraction is shown in Table 2). Then, the proposed reduction method of Choquet-like integral based on multi-granularity variable-scale fuzzy neighborhood measure is used (the simplified data information is shown in Table 3) to realize the fault diagnosis of rolling bearings.

[0097] Table 1 Data of three types of rolling bearings

[0098]

[0099] Table 2 Data information after alternative feature extraction

[0100]

[0101] Table 3 Simplified data information

[0102]

[0103] Finally, on the one hand, the data in Table 2 are directly processed using intelligent fault diagnosis methods (support vector machine, adaptive enhancement and K nearest neighbor algorithm); on the other hand, the data is first optimized using the proposed reduction method of Choquet-like integral based on multi-granularity variable-scale fuzzy neighborhood measure, and then the intelligent fault diagnosis method is used to process the fault data (data in Table 2 after attribute simplification) to achieve the effect of improving accuracy and optimizing the model. The results are shown in Tables 4, 5 and 6.

[0104] Table 4. Comparison of fault diagnosis in Dataset 1

[0105]

[0106] Table 5. Comparison of fault diagnosis in Dataset 2

[0107]

[0108] Table 6 Comparison of fault diagnosis of Dataset 3

[0109]

[0110] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A multi-granularity variable-scale fuzzy neighborhood measure and corresponding Choquet-like integral fault diagnosis method, characterized by: Fault diagnosis method based on "attribute reduction + intelligent classifier": Input: Complete fault diagnosis decision information table; Step 1: Using a new method of attribute simplification of decision tables based on Choquet-like integral of multi-granularity variable-scale fuzzy neighborhood measure, redundant fault features are deleted to achieve dimensionality reduction of the fault decision information table; Step 2: Select an intelligent classifier model and establish an intelligent classifier diagnosis model; construct a new fault data set based on the reduction determined in step 1; then, use it to train the diagnosis model; Step 3: Use the trained intelligent classifier to classify the test sample symptom set to obtain the diagnosis result; Output: fault diagnosis results and decisions; The new method for reducing the attributes of the decision table of the multi-granularity variable-scale fuzzy neighborhood measure and the corresponding Choquet-like integral in step 1 includes: S1. Propose a multi-granularity variable-scale fuzzy Θ-covering approximation space: Definition 1. Let U be a domain, Θ = {β1,…,βi,…,βn}, where βi∈(0,1]; if is a fuzzy βi-cover of U, then It is called a multi-granularity variable-scale fuzzy Θ-cover on U; (U,Ω) is also called a multi-granularity variable-scale fuzzy Θ-cover approximation space; S2. Propose a multi-granularity variable-scale fuzzy Θ-neighborhood: Definition 2. Let (U,Ω) be a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ={β1,…,βi,…,βn}; for any x∈U, the multi-granularity variable-scale fuzzy Θ-neighborhood of x is defined as: in, For x The fuzzy βi-minimum description under , IT is the R-implication operator induced by the fuzzy logic operator t-norm; S3. Propose upper and lower multi-granularity variable-scale fuzzy neighborhood measures: Let (U,Ω) be a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ={β1,…,βi,…,βn}; for any The upper and lower multi-granularity variable-scale fuzzy neighborhood measures of X are defined as: S4. A Choquet-like integral based on upper and lower multi-granularity variable-scale fuzzy neighborhood measure is proposed: Definition 4. Let (U,Ω) be a multi-granular variable-scale fuzzy Θ-covering approximation space, T be the t-norm, A∈F(U), where U = {x1,…,xn}; then Θ = {β1,…,βi,…,βn}, for any The upper and lower multi-granularity variable-scale fuzzy neighborhood measures of X are defined as: where {x(1),x(2),…,x(n)} is a permutation of {x1,x2,…,xn} such that A(x(1))≤A(x(2))≤…≤A(x(n)), X(i)={x(i),x(i+1)…,x(n)}, S5. Based on S1-S4, a new method for attribute reduction of decision tables based on Choquet-like integral of multi-granularity variable-scale fuzzy neighborhood measure is proposed: Input: Input decision information table T = (U, A∪{d}), where U = {x1, x2, …, xn}, A = {a1,a2,…,am},bij=aj(xi),i∈{1,2,…,n},j∈{1,2,…,m}; Output: attribute reduction of attribute set A.

2. The multi-granularity variable-scale fuzzy neighborhood measure and corresponding Choquet-like integral fault diagnosis method according to claim 1, characterized in that: The S5 implementation process specifically includes: Step 1: Induce a multi-granularity variable-scale fuzzy Θ-covering approximation space: For any aj∈A, define the fuzzy set in is the variance of attribute aj; then is a fuzzy βj-cover (βj∈(0,γj]), Therefore, (U,Ω) is a multi-granularity variable-scale fuzzy Θ-covering approximation space, where and Θ = {β1,…,βi,…,βn}, also known as (U,Ω,d) is a multi-granularity variable-scale fuzzy Θ-covering group decision information system; Step 2: Calculate the multi-granularity variable-scale fuzzy Θ-neighborhood for all xk∈U Step 3: Calculation Where xk∈U, U / d={D1,…,Dt,…,Ds} is the partition of the domain U with respect to the decision attribute d; Step 4: Calculate the minimum Choquet-like reduction Redt of (U, Ω, d). The specific steps are as follows: (a) P←{}; (b) for any D∈U / d (c)do (d)Q←P; (e) (f) (e)Q←P∪{C} (f)P←Q (h)return P%Suppose (i)Redt={a (1) ,…,a (h) }。 3. A multi-granularity variable-scale fuzzy neighborhood measure and corresponding Choquet-like integral fault diagnosis method, comprising any one of claims 1-2, characterized in that: The bearing used in the test supports the rotating shaft of the motor. In order to obtain bearing fault data, cracks were machined on the rolling bearing using electrospark machining technology, resulting in a series of faulty bearings with notch diameters of 0.007 inches, 0.014 inches, 0.021 inches, 0.028 inches and 0.4 inches. For the outer ring fault, not only cracks of different sizes were generated, but also outer ring faults of the bearing were created at different positions relative to the load stress point, located at 3 o'clock, 6 o'clock and 12 o'clock respectively. In order to measure the vibration data of the faulty bearing, the experimental platform installed acceleration sensors on the drive end, fan end and base of the motor. During the data collection process, the experimental platform only set a single fault type, so the fault types in the bearing data set used in the test were all single fault types. During the bearing vibration data collection process, the test platform was loaded with loads of different horsepower, and the data sampling frequency was 12KHz. After exporting the required data, for each collected data file, take every 500 data points as an example, use the existing feature extraction method to calculate all feature parameters as candidate features; Then, the proposed reduction method of Choquet-like integral based on multi-granularity variable-scale fuzzy neighborhood measure is used to realize the fault diagnosis of rolling bearings.

Citation Information

Patent Citations

  • Method for constructing expert system based on knowledge discovery

    CN101093559A

  • Equipment fault diagnosis method and system of electric power system

    CN103630244A

  • System and process for a fusion classification for insurance underwriting suitable for use by an automated system

    US20040220837A1

  • System and process for multivariate adaptive regression splines classification for insurance underwriting suitable for use by an automated system

    US20040220840A1